CN103235556B - The complex parts digital control processing manufacture method of feature based - Google Patents

Abstract

A feature-based NC machining and manufacturing method for complex parts, which expresses features based on ontology and object-oriented methods, and then expresses the geometric and process information of parts in the entire manufacturing process based on features, using features as the carrier of manufacturing knowledge and experience. Effective integration and closed-loop control of manufacturing stage information such as design-processing-testing, based on automatic feature recognition, automatic process decision-making, automatic NC programming, post-processing, process control, online testing, process optimization and man-hour prediction based on features . The feature-based manufacturing method realizes the effective connection of the information chain in the processing process, makes the manufacturing process automatic and intelligent, reduces the dependence of the manufacturing process on human experience, facilitates the production management of the manufacturing process, and improves the manufacturing efficiency. Reduced production costs.

Description

NC Machining and Manufacturing Method of Complex Parts Based on Feature

technical field

The present invention relates to an integrated manufacturing method, especially one based on CAD (Computer Aided Design)/CAPP (Computer Aided Process Planning)/CAM (Computer Aided Manufacturing)/CNC (Computer Numerical Control)/CAI (Computer Aided Inspection) The invention relates to a manufacturing method capable of automatically generating NC machining programs for complex parts, specifically a NC machining and manufacturing method for complex parts based on aircraft component features.

Background technique

High precision, high efficiency, and high reliability are the unremitting goals of manufacturing technology. The processing of high-precision, cutting-edge complex products in the fields of aviation, aerospace, automobiles, ships, energy microelectronics, etc. continues to pose new challenges to the ultimate performance of modern manufacturing processes, equipment and systems, and urgently needs methods that can solve the essence of the manufacturing process. Digital manufacturing combines advanced manufacturing technology with digital technology, and profound changes have taken place in the description, planning, operation and control of manufacturing processes, equipment and systems. Digital manufacturing realizes digital modeling and information integration, collaboration between manufacturing process and manufacturing system, virtual simulation, etc. through digital simulation and optimization of manufacturing process, as well as digital expression, organization and storage of manufacturing knowledge and information. But so far, there is no method or means that can realize the integration and expression of the above-mentioned digital manufacturing knowledge, so as to realize the integration of the entire manufacturing process.

As a means of digital manufacturing, feature technology has played an important role in information integration. In the application of traditional research and commercial software, feature technology is widely used in the field of CAD/CAM, but it is still in the research stage in terms of integration. , are basically independent applications in CAD or CAM. In recent years, many scholars have studied feature technology in STEP-NC, and proposed a series of international standards, but they are all in the exploratory stage, and have poor interface with current industrial equipment, making it difficult to popularize and apply.

Searches for published or public literature are basically related to CAD/CAM integration or CAM/CNC integration, or research on a single application field. At the same time, in terms of information expression and methods, it cannot be applied to complex parts and cannot effectively express the information of the entire manufacturing process.

To sum up, in view of the important role and challenges of digital manufacturing in the entire manufacturing field, the entire manufacturing system is relatively loose; and feature technology, as a means of realizing digitalization, has not yet realized the integration of manufacturing systems.

Invention content:

The purpose of the present invention is to solve the problem that the existing entire manufacturing system is relatively loose and it is difficult to realize digital integrated manufacturing, use features as the carrier of manufacturing knowledge and experience, realize effective integration of information, and automatically generate the NC machining program of the entire part based on feature CNC machining and manufacturing methods for complex parts.

Technical scheme of the present invention is:

A feature-based NC machining and manufacturing method for complex parts, which is characterized by first reading the 3D digital model of the part to be processed, and secondly performing automatic feature recognition to extract the geometric information and process of the part to be processed from the read 3D digital model Information, the information not included in the 3D digital model is obtained through the relevant process documents matched with the digital model, and the extracted geometry and process information are characterized and marked, and all features are uniquely identified according to the feature category and serial number, that is, feature identification = feature category + serial number, using this identification as a tool for feature association between different applications; if the feature is decomposed during the manufacturing process, then feature identification = decomposition identification + feature category + serial number + decomposition sub-serial number, so as to realize the use of features as manufacturing knowledge and the carrier of experience to realize the effective integration of information; thirdly, in order to maintain the validity of features and the continuity of feature evolution throughout the manufacturing process, object-oriented and ontology-based methods are used to represent features, that is, to define the ontology of features and each application For the class of domain features, the method of holographic attribute surface and edge graph is used for feature recognition; finally, the NC machining program of complex parts is automatically generated after process optimization, and the NC machining equipment is driven to complete the entire machining process.

After the automatic feature recognition is completed, the processing technology optimization model is constructed based on the feature, automatic process decision-making based on the feature, automatic NC programming based on the feature, post-processing based on the feature, process control based on the feature, online detection based on the feature, Based on the characteristics, construct the man-hour sample library of the processed parts and construct the mapping model of the feature geometric information and process information to the machine tool performance parameters;

(1) The feature-based construction of the processing technology optimization model should include the geometric information of the feature, the tool information, the machine tool information and the material information of the workpiece. The constraints of the optimization model are cutting force and vibration, and the optimization target of the optimization model is the machining Efficiency, processing quality and processing cost, the optimized items are depth of cut, width of cut and feed speed, and a feature-based cutting parameter library is established; a 3D model of cutting test specimens based on features is built for the application field, including typical features and typical Combination, as a standard sample to verify the processing technology;

(2) The feature-based automatic process decision-making is to provide complete information support for automatic process decision-making, express process rules and process schemes based on features, and then perform automatic process decision-making based on features, and express the results of automatic process decision-making based on features;

(3) The feature-based automatic NC programming refers to using the results of automatic process decision-making as the basis for automatic programming, and then automatically generating machining tool paths based on features, and expressing the results of automatic programming based on features;

(4) The feature-based post-processing is based on the results of automatic programming, considering the dynamic characteristics of the numerical control system, performing feature-based post-processing, and expressing the results of the post-processing based on the features to form a CNC machining program;

(5) The processing process control based on features is a processing strategy for expressing problems in the processing process, including state monitoring of the processing process based on features, intelligent self-adaptive adjustment during the processing process, and feature-based recording during the processing process. data for process optimization;

(6) The feature-based online detection refers to the formation of detection point generation specifications based on the geometric information and process information of the features, the optimization of the probe attitude vector and detection trajectory during online detection based on the geometric information of the features, and the formation of detection and analysis results based on the features , as the basis for optimizing the processing technology;

(7) The man-hour sample library of the feature-based construction of processed parts contains geometric information of features related to man-hours and part man-hours. The feature geometric information and process information are used as input to construct a neural network. Predict the working hours of the parts to be processed under the premise, as the basis for parts processing external quotations and production planning arrangements; add feature geometric information and process information to the feature-based NC program, combined with the dynamic characteristics of the CNC machine tool and the cutting force during the cutting process , to accurately predict the NC machining hours of each process, as the basis for machine tool scheduling;

(9) The construction of the mapping model of feature geometric information and process information to machine tool performance parameters is based on feature analysis to meet the performance parameters of machine tools required for processing parts, and provides application-oriented design references for machine tool related parameters.

The steps of performing automatic process decision-making based on features are:

(1) Based on the geometric information and process information related to feature expression and process decision-making, the geometric information is: the geometric size, position, geometric structure, and maximum tool diameter allowed by the feature, and the process information is: material, blank form, processing Allowance and machining accuracy;

(2) Process decision-making based on features is carried out in the order of machine tool decision-making, processing sequence decision-making, processing method decision-making and cutting parameter decision-making, and automatic process decision-making results are formed according to process rules and process plans;

(3) The form of process decision-making results based on feature expression is: each process contains several steps, each step contains several operations, each operation corresponds to a processing feature, and the feature contains the operation type of the feature, the processing strategy and cutting parameters.

The described feature-based automatic numerical control programming method is:

(1) Geometric information and process information related to automatic programming based on feature expression, where the geometric information includes: the geometric size of the feature, the driving geometry required to generate the tool path, the starting point and end point of the tool path, the avoidance geometry of the tool path, and the process The information includes: tool information for machining features, machining accuracy of features and tool path strategy;

(2) Extract processing features through feature recognition, and output feature recognition results including geometric information and process information;

(3) Based on the geometric information of the feature recognition result and the process information of the process decision result, the tool path of each processing feature is automatically generated;

(4) Generate the tool location file of the machining tool path for each feature in units of features, and mark the corresponding information according to different application requirements.

The described feature-based post-processing steps include:

(1) Geometric information and process information related to post-processing are expressed based on features in the NC program, where geometric information refers to: the type of line in the geometry of the feature: straight line, circular arc, conic section, free curve, and surface Type: plane, ruled surface, free-form surface, the process information refers to the type of CNC machining operation when machining the corresponding feature;

(2) The characteristics of the CNC system considered in post-processing, including acceleration characteristics, interpolation accuracy characteristics and tool holder compensation control parameter types;

(3) Based on the feature geometry and process information, combined with the dynamic characteristics of the CNC system, the tool position points are fitted to form a polynomial interpolation method supported by the CNC system, and the feed rate is optimized to adapt to complex CNC machining conditions.

The steps of the feature-based process control include:

(1) Geometric information and process information related to machining process control based on feature expression, where geometric information is: corner information and driving geometric information, and process information is: machining tools, cutting parameters, monitoring strategies and detection strategies;

(2) According to the geometric information and process information of the feature and the processing strategy, carry out online monitoring, and realize the processing of the monitoring results based on artificial intelligence methods;

(3) The data recorded during the processing refers to the actual cutting parameters during the processing, the operating power of the machine tool, the tool wear and chatter during the processing; the above data are organized in units of features.

The described feature-based online detection comprises the following steps:

(1) Based on the geometric information and process information related to feature expression and detection, the geometric information is: the surface type, curvature change, area, normal direction and theoretical geometric model of the intermediate processing state of the detection feature; the process information is: processing accuracy;

(2) Based on the geometric information and process information of the feature, the generation rule of the detection point of the detection feature is constructed, the detection point of the detection feature is generated based on the feature, and the detection track is optimized considering the intermediate state of the detection feature, and the detection file is generated;

(3) Install probes for online detection, record the detection results and analysis results, organize based on characteristics, and feed back to the process department as the basis for process optimization.

The feature-based processing technology optimization model considers the geometric information and process information of the feature, wherein the geometric information refers to: the geometric size of the feature as an important consideration of rigidity, the thickness and height of the geometric elements constituting the feature.

The geometric information based on the feature-based construction of the man-hour sample library of the processed parts includes the position of the corner, the size of the corner, etc.; the process information includes the cutting allowance, depth of cut, width of cut and main feed speed; the dynamic characteristics of the machine tool include the Acceleration, jerk and turning performance, based on geometric information, process information and dynamic characteristics of the machine tool, predict the actual feed speed during the machine tool movement, and predict the change of the feed speed due to the change of the cutting force in combination with the cutting force, In turn, the feed rate can be accurately predicted, and the machining man-hours can be calculated based on the tool path.

The object-oriented and ontology-based feature representation method is:

(1) First define the feature ontology. The feature ontology includes the unique identification of the feature, the identification surface, and the seed identification surface that may become the identification surface. The definition of the seed identification surface is due to the evolution of the feature during the manufacturing process, and the feature may be decomposed;

(2) The feature-based ontology derives feature classes in various application fields. The class inherits the attributes of the ontology. The class contains geometric information and process information. According to the needs of different application fields, different geometric information and process information are defined, so as to realize Information integration and transmission throughout the manufacturing process.

Beneficial effects of the present invention:

(1) Ontology-based and object-oriented methods are used to express information under different application views at different stages of features, so that the information of different application views at different stages can be effectively integrated.

(2) The present invention effectively integrates manufacturing knowledge and experience, reducing the dependence of the manufacturing process on human experience.

(3) The present invention can effectively digitally express the information of the manufacturing process, effectively integrate the information of the manufacturing process, and form a closed-loop control of the entire manufacturing process.

(4) The present invention is beneficial to the management of production information, and provides information guarantee for enterprises to carry out knowledge engineering.

(5) The present invention is beneficial to the automation and intelligence of the production process, improves the production efficiency and reduces the production cost.

Description of drawings

Figure 1. Flowchart of feature-based fabrication method.

Figure 2. (a) Front side of a typical double-sided aircraft structure; (b) Reverse side of a typical double-sided aircraft structure.

Figure 3. Screenshot of information file required for process decision on slot characterization.

Figure 4. Screenshot of the groove feature process decision result information file.

Figure 5. A screenshot of the information file required for automatic programming of slot features.

In the figure, 1 denotes a groove, 2 denotes a rib, 3 denotes a hole, and 4 denotes an outline.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

A feature-based NC machining and manufacturing method for complex parts, expressing the geometry and process information of parts in the entire manufacturing process based on features, using features as the carrier of manufacturing knowledge and experience to achieve effective integration of information, in order to achieve effective integration of information in the entire manufacturing process Maintain the validity of features and the consistency of feature evolution, and use object-oriented and ontology-based methods to represent features:

(1) Define the ontology of features and the classes of features in each application domain;

(2) Feature recognition is performed based on the method of holographic attribute surface and edge graph, and the geometric information and process information of the part are extracted from the 3D digital model. The information not included in the 3D digital model is obtained through relevant files. Carry out unique identification, that is, feature identification = feature category + serial number, and use this identification as a tool for feature association between different applications; if the feature is decomposed during the manufacturing process, then feature identification = decomposition identification + feature category + serial number + decomposition sub-serial number.

The present invention expresses the geometry and process information of parts in the entire manufacturing process based on features. In order to adapt to different application fields and maintain the validity of features and the continuity of feature evolution in the entire manufacturing process, object-oriented and ontology-based methods are used to express features. Based on this, automatic process decision-making, automatic NC programming, post-processing, process control, online inspection, process optimization and man-hour prediction are carried out based on features. This method is realized through the following steps:

(1) Define the ontology of features and the classes of features in each application domain;

(2) Feature recognition is performed based on the method of holographic attribute surface and edge graph, and the geometric information and process information of the part are extracted from the 3D digital model. The information not included in the 3D digital model is obtained through relevant files. Carry out unique identification, that is, feature identification = feature category + serial number, and use this identification as a tool for feature association between different applications; if the feature is decomposed during the manufacturing process, then feature identification = decomposition identification + feature category + serial number + decomposition sub-serial number;

(3) Build a processing technology optimization model based on features. The factors of the optimization model include geometric information of features, tool information, machine tool information and workpiece material information. The constraints of the optimization model are cutting force and vibration, and the optimization target of the optimization model is machining Efficiency, processing quality and processing cost, the optimized items are depth of cut, width of cut and feed speed, and a feature-based cutting parameter library is established; a 3D model of cutting test specimens based on features is built for the application field, including typical features and typical Combination, as a standard sample to verify the processing technology;

(4) Based on feature expression and process decision-related geometric information and process information, provide complete information support for automatic process decision-making, express process rules and process schemes based on features, and then make automatic process decisions based on features, and automatically process decisions based on feature expression the result of;

(5) Geometric information and process information related to NC programming are expressed based on features, and the results of automatic process decision-making are also used as the basis for automatic programming, and then machining tool paths are automatically generated based on features, and the results of automatic programming are expressed based on features;

(6) Based on the geometry and process information related to feature expression and post-processing, based on the results of automatic programming, consider the dynamic characteristics of the CNC system, perform feature-based post-processing, and based on the results of feature expression post-processing, form NC program;

(7) Express the geometric information and process information related to the control of the machining process based on the features, express the processing strategies for various problems in the machining process, monitor the status of the machining process based on the features, perform intelligent adaptive adjustment during the machining process, and record the machining based on the features Process data for process optimization;

(8) Form the detection point generation specification based on the geometric information and process information of the feature, optimize the probe attitude vector and detection trajectory of the detection point for online detection based on the geometric information of the feature, and form the detection analysis result based on the feature as the basis for optimizing the processing technology ;

(9) Construct the man-hour sample library of the processed parts based on the features. The man-hour sample library contains the geometric information of the features related to the man-hour and the part man-hours. The feature geometric information and process information are used as input to construct a neural network. When there is no part NC program Under the premise of predicting the working hours of the parts to be processed, it is used as the basis for external quotations for parts processing and production planning arrangements;

(10) Add feature-based geometric information and process information to the feature-based NC program, combine the dynamic characteristics of the NC machine tool and the cutting force during the cutting process, and accurately predict the NC machining hours of each process as the basis for machine tool scheduling;

(11) Construct a mapping model of feature geometric information and process information to machine tool performance parameters, meet the performance parameters of machine tools required for processing parts based on feature analysis, and provide application-oriented machine tool design;

The object-oriented and ontology-based feature representation method is:

(1) First define the feature ontology. The feature ontology includes the unique identification of the feature, the identification surface, and the seed identification surface that may become the identification surface. The definition of the seed identification surface is due to the evolution of the feature during the manufacturing process, and the feature may be decomposed;

(2) The feature-based ontology derives feature classes in various application fields. The class inherits the attributes of the ontology. The class contains geometric information and process information. According to the needs of different application fields, different geometric information and process information are defined, so as to realize Information integration and transmission throughout the manufacturing process.

The characteristic-based process decision-making steps are:

(1) Based on the geometric information and process information related to feature expression and process decision-making, the geometric information is: the geometric size, position, geometric structure, and maximum tool diameter allowed by the feature, and the process information is: material, blank form, processing Allowance, machining accuracy, etc.;

(2) Process decision-making based on features is carried out in the order of machine tool decision-making, processing sequence decision-making, processing method decision-making and cutting parameter decision-making, and automatic process decision-making results are formed according to process rules and process plans;

(3) The form of process decision-making results based on feature expression is: each process contains several steps, each step contains several operations, each operation corresponds to a processing feature, and the feature contains the operation type of the feature, the processing strategy and cutting parameters.

The feature-based automatic programming method is:

(1) Based on the geometric information and process information related to automatic programming based on feature expression, the geometric information includes: the geometric size of the feature, the driving geometry required to generate the tool path, the starting point and end point of the tool path, the avoidance geometry of the tool path, etc., Process information includes: tool information for machining features, machining accuracy of features, tool path strategy, etc.;

(2) Extract processing features through feature recognition, and output feature recognition results including geometric information and process information;

(3) Based on the geometric information of the feature recognition result and the process information of the process decision result, the tool path of each processing feature is automatically generated;

(4) Generate the tool location file of the machining tool path for each feature in units of features, and mark the corresponding information according to different application requirements.

The feature-based post-processing steps are:

(1) Geometric information and process information related to post-processing are expressed based on features in the NC program, where geometric information refers to: the type of line in the geometry of the feature: straight line, circular arc, conic section, free curve, and surface Type: plane, ruled surface, free-form surface, the process information refers to the type of CNC machining operation when machining the corresponding feature;

(2) The characteristics of the CNC system considered in post-processing, including acceleration characteristics, interpolation accuracy characteristics, tool holder compensation control and other parameter types;

(3) Based on the feature geometry and process information, combined with the dynamic characteristics of the CNC system, the tool position points are fitted to form a polynomial interpolation method supported by the CNC system, and the feed rate is optimized to adapt to complex CNC machining conditions.

The steps of the feature-based process control are:

(1) Geometric information and process information related to machining process control based on feature expression, where geometric information includes: corner information, driving geometric information, etc., and process information includes: machining tools, cutting parameters, monitoring strategies, detection strategies, etc.;

(2) According to the geometric information and process information of the feature and the processing strategy, carry out online monitoring, and realize the processing of the monitoring results based on artificial intelligence methods;

(3) The data recorded during the processing refers to the actual cutting parameters during the processing, the operating power of the machine tool, and various problems during the processing such as tool wear, chatter, etc.; the above data are organized in units of features.

The feature-based online detection includes the following steps:

(1) Based on the geometric information and process information related to feature expression and detection, the geometric information is: the surface type, curvature change, area, normal direction, theoretical geometric model of the intermediate processing state of the detection feature, etc.; the process information is: machining accuracy ;

(2) Based on the geometric information and process information of the feature, the generation rule of the detection point of the detection feature is constructed, the detection point of the detection feature is generated based on the feature, and the detection track is optimized considering the intermediate state of the detection feature, and the detection file is generated;

(3) Install probes for online detection, record the detection results and analysis results, organize based on characteristics, and feed back to the process department as the basis for process optimization.

The feature-based processing technology optimization model considers the geometric information and process information of the feature, wherein the geometric information refers to: the geometric size of the feature, the thickness and height of the geometric elements constituting the feature, and these factors are important considerations for rigidity.

In the feature-based precise machining man-hour prediction method, geometric information refers to the corner position and corner size of the feature; process information refers to cutting allowance, depth of cut, width of cut, and feed rate; dynamic characteristics of the machine tool include acceleration, acceleration, and Acceleration, turning performance, based on geometric information, process information and the dynamic characteristics of the machine tool, predict the actual feed speed during the machine tool movement, and combine the cutting force to predict the change of the feed speed caused by the change of the cutting force, and then accurately predict The feed rate, and then the geometric tool path to calculate the processing time.

The details are as follows:

As shown in Figure 1-5.

A feature-based numerical control manufacturing method and system for complex parts. The overall process is shown in Figure 1. A typical double-sided aircraft structural part is taken as an example to illustrate the specific implementation process of the feature-based manufacturing method.

As shown in Figure 2, the features of aircraft structural parts can be divided into slot 1, rib 2, hole 3, and profile 4. The ID of slot feature is P0001, the ID of rib feature is R0001, the ID of hole feature is H0001, and It is identified as F0001, and the number after the letter is determined according to the serial number of each feature in the feature category. After each feature is decomposed, the sub-features are identified as follows: the identification of the slot decomposition feature is ZP0001-001, the identification of the rib decomposition feature is ZR0001-001, the identification of the hole decomposition feature is ZH0001-001, and the identification of the contour decomposition feature is ZF0001-001; The identification surface of the groove feature is the web surface, the seed identification surface is the side surface, the identification surface of the rib feature is the rib top surface, the seed identification surface is the side surface of the rib, the identification surface of the hole feature is the hole wall surface, and the seed identification surface is the upper and lower top surfaces of the hole , for contour features, mark the largest contour surface as the identification surface, and the rest of the surfaces are used as the seed identification surface.

The following mainly uses the groove feature as an example to illustrate the whole process, and the rest of the features are similar.

Define the ontology of the feature and the class of the feature in each application field: first define the feature ontology, the feature ontology includes the unique identification of the feature, the identification surface, and the seed identification surface that may become the identification surface. The definition of the seed identification surface is due to the evolution of the feature during the manufacturing process , the features may be decomposed; all features are uniquely identified according to the feature category and serial number, that is, feature identification = feature category + serial number, and this identification is used as a tool for feature association between different applications; if the feature is decomposed during the manufacturing process, Then feature identification = decomposition identification + feature category + serial number + decomposition sub-serial number; for example, the identification of slot 1 of a typical double-sided aircraft structure in Figure 2 is P0001, and the identification surface of the slot feature is the web surface of the slot feature;

The feature-based ontology derives feature classes in various application fields. The classes inherit the attributes of the ontology. The classes contain geometric information and process information. According to the needs of different application fields, different geometric information and process information are defined to realize the entire manufacturing process. information integration and delivery. The main characteristics of several application areas are defined as follows:

The geometric information of the groove features related to the process decision is: the center position of the groove web, the area, depth, corner radius, the maximum allowable tool diameter, the maximum inclination angle of the processing surface; the process information is: the material, the blank form is Precision forgings, rough machining allowance, machining accuracy;

The geometric information of the groove feature related to automatic programming is: the depth of the groove feature, the driving surface of the groove feature machining operation, such as the side wall surface of the side wall processing, the starting point and the end point of the tool track, the geometry of the boss avoidance in the middle, etc. The process information includes: machining tools for groove features, precision, tool path advance and retreat strategies, cutting strategies, etc.;

Based on the method of holographic attribute surface edge graph for feature recognition, the geometric information and process information of parts are extracted from the 3D digital model, and the information not included in the 3D digital model is obtained through related files; Extensible Markup Language) files are saved to facilitate subsequent calls.

Taking groove feature 1 of a typical aircraft structural part in Figure 2 as an example, the recognition result is shown in the XML file in Figure 5, which includes the serial number of the groove, the axial direction of the groove, the radius of the corner of the groove, the radius of the bottom corner, all layers of the groove feature, Information such as the vertex of each layer, the height of the layer, the corner and bottom corner radius of the layer, the unique identification value of the web surface, side surface, top surface and bottom corner surface in the CAD model;

Based on geometric information and process information related to feature expression and process decision-making, it provides complete information support for automatic process decision-making. Since the processing features extracted from the part model are only static correspondences between specific geometric shapes and processing information, there is a dynamic relationship between each processing stage. In order to avoid only using static processing features and affecting the accuracy of process decision results, the state of the feature processing process is constructed, and the process decision method based on the state of feature processing process is used to realize the dynamic and overall optimization of the process plan. Finally, the results of automated process decisions are expressed based on features.

a. The state construction method of the feature processing process is based on the automatically recognized feature result file, and consists of three steps: constructing the structure of the feature processing process plan, extracting the state monitoring position and calculating the state of the feature processing process.

Structuring feature machining process plan structure refers to extracting the process specification structure and typical process parameters related to the feature from the typical part-level process plan.

Extracting state monitoring position information refers to extracting the corresponding processing area according to the structure of the characteristic processing technology plan, and performing geometric analysis to obtain the key cutting position points that can reflect the characteristic cutting state under the process plan, as IPS (Interim Process State, intermediate processing state ) of the processing state calculation input.

Calculating the state of the feature processing process is to calculate the cutting state of the key cutting position points after the calculation position is determined, combined with the process parameter information, and finally complete the construction of the state of the feature processing process.

b. The process decision based on the state of the feature processing process is carried out in the order of machine tool decision, processing feature sorting decision, processing method decision and cutting parameter decision.

On the basis of the state of the characteristic processing process, the information structure is firstly layered, and the two types of relationship, vertical constraint and horizontal, are summarized, and there are six constraint relationships of subordination, correspondence, reference, sequence, transmission, and restriction to describe the internal state of the processing process. complex process relationships. Establish the analysis method of process state process constraint relationship, transform the decision-making and optimization of process plan into the analysis and treatment of six relationships. The process decision-making method based on the processing state can take the actual processing state as the reasoning basis, and take into account the overall optimization of feature processing while pursuing local optimization.

The processing feature sorting decision in process decision-making is divided into machine tool level sorting, clamping level sorting and operation level sorting. By integrating machining knowledge and experience into the sorting process in the form of rules, the problem of sorting processing features caused by intersecting features including free-form surfaces is well solved.

c. The form of process decision-making based on feature expression is: each process contains several steps, each step contains several operations, each operation corresponds to a processing feature, and the feature contains the operation type, processing strategy and cutting of the feature parameter.

Take the front finishing of typical aircraft structural parts shown in Figure 2 as an example to illustrate the process decision results. Finishing includes four steps: fine milling rib top, groove web, inner shape, fine milling corner, drilling, and milling outer contour. Since the same tool is used for fine milling of the rib top, groove web and inner shape, they are placed in the same work step, and each work step corresponds to the corresponding processing feature. The processing sequence among features is arranged in the order of rib, groove web, groove shape, corner, hole, and contour. The sorting of the same feature is sorted according to the principle of the shortest empty tool path. Slot 1 is machined fourth of all slot features.

Take groove 1 as an example to illustrate the operation type and cutting parameter decision results. The operation type is Pocketing, located in step 1, the fourth pocketing operation. The processing strategy of this groove is from inside to outside, with a depth of cut of 2mm and a width of 10mm. , high-speed fillet of 0.5mm, feed rate of 2500mm/min, spindle speed of 9000r/min, and the number of cutting layers is 2 layers.

The machining process optimization model is constructed based on features to realize the optimization of multiple cutting parameter groups. The process optimization model includes geometric information of features, tool information, machine tool information and workpiece material information. In the cutting process, the cutting speed, cutting depth, and feed rate are set as the three elements of the cutting amount. Therefore, process optimization mainly refers to finding the optimal cutting speed, cutting depth and feed rate. In addition, it also includes the choice of tools and so on. The goals of process optimization are processing efficiency, processing quality and processing cost. According to the optimization objective, the corresponding objective functions are established, that is, the objective function of the highest production efficiency, the objective function of the highest processing quality and the objective function of the lowest processing cost.

The main influencing factors of the objective function of maximum production efficiency are: process cutting time, tool changing time, clamping time and other auxiliary time. Among them, the main factors affecting the cutting time of the process are: cutting depth, cutting width, cutting speed, feed speed, etc.; the factors affecting the tool change time are the time of one tool change and the number of tool changes, and the durability of the tool affects the number of tool changes. Establish corresponding influence coefficients for these factors, and establish the highest production efficiency objective function by integrating all factors;

The influencing factors of the objective function of the highest machining quality are: material properties of parts, cutting speed, feed speed, etc. Also establish the influence coefficient for the influencing factors and establish the objective function.

The influencing factors of the minimum machining cost objective function are: blank cost, cost required in cutting time, tool cost, cost required in clamping time and cost required in auxiliary time. The influence coefficient is established for the influencing factors, and finally the minimum cost objective function is established.

The cutting parameters and tools are optimized according to three optimization objective functions. The constraints that need to be considered during the optimization process are:

① Safety constraints, that is, cutting force, cutting torque, etc. cannot exceed the allowable range of machine tools, tools, fixtures and workpieces;

② Capacity constraints, that is, the spindle speed, feed speed, spindle acceleration, feed acceleration, etc. cannot exceed the performance indicators of the machine tool;

③Quality constraints, after process optimization, the workpiece must meet the given machining accuracy and surface quality;

④ Vibration constraints. During the processing, the cutting tool, workpiece and machine tool should not vibrate significantly.

Establish a feature-based cutting parameter library to store the optimized cutting parameters based on the feature-based process optimization model for reference in automatic process decision-making.

Finally, the three-dimensional model of the cutting test specimen based on the feature is constructed for the application field as a standard sample for verifying the processing technology. The three-dimensional model includes typical features and their typical combinations.

Geometric information and process information related to NC programming are expressed based on features, and the results of automatic process decision-making are also used as the basis for automatic programming, and then machining tool paths are automatically generated based on features, and the results of automatic programming are expressed based on features; feature-based automatic NC The programming algorithm mainly includes drive information acquisition, tool path generation, machining simulation and post-processing.

Driving information includes driving geometry information, processing mode and parameter information, feed mode and parameter information. Among them, the processing method and parameters, feed method and parameters are directly given in the form of numbers, letters and other data, while the driving geometry is implicit in the geometric information of the processing features, which needs to be processed and converted into a tool path generation algorithm that can be used directly. data form.

Different feature machining tool path generation algorithms require different driving geometries. Here, the processing of typical grooved webs is taken as an example to illustrate the process of processing feature reconstruction.

Processing methods and parameters mainly include: tool path type (Tool Path Style), cutting direction (Direction Of Cut), machining accuracy (Machining Tolerance), tool (Tool), tool diameter D, tool bottom tooth radius Rc, edge length Lc, Tool working length L, cutting width (Radial Distance Between Path), cutting depth (Axial Maximum Depth Of Cut), fillet radius (Corner Radius), fillet arc center angle (Limit Angle), fillet tangential extension length (Extra Segmentoverlap), transition arc radius (Transition Radius), transition angle (Transition Angle), transition line length (TransitionLength), fillet deceleration rate (Reduction Rate), fillet corner minimum center angle (Minimum Angle), fillet deceleration maximum Radius (Maximum Radius), rounded corner deceleration preset distance (Distance Before Corner), rounded corner deceleration cancellation distance (Distance After Corner).

Feed mode and parameters mainly include: Approach, Horizontal Safety Distance, Vertical Safety Distance, Ramping Angle, Approach Feed Speed, Retract ), the central angle of the retracting arc (Angular Sector), the orientation of the retracting arc (Orientation), the radius of the retracting arc (Radius), the retracting speed (Retract Feed Speed), the processing speed (Machining Feed Speed), the spindle speed ( Spindle Speed).

The drive geometry is generated using a ring analysis-based approach. This method solves the problem of many classifications of curve cutting in the creation of the web finishing area, and the complex and error-prone problem of line segment selection after cutting; it can process groove feature webs such as closed angles, open and closed angles coexisting, and common sides. Unified calculation is performed for area creation, and the algorithm has good versatility, which effectively improves the efficiency of web finishing programming. The specific steps of the method are as follows:

①Project the outer ring boundary of the generalized side surface and generalized bottom surface of the groove feature to the web surface;

② Obtain a plane closed contour including straight lines, arcs and splines through projection;

③Approximate the projected curve segment with a straight line segment, and intersect and divide all the straight line segments;

④ For the needs of ring analysis vector rotation, discretize the planar closed contour into a series of straight line segments, and intersect and segment them to generate a planar closed contour containing only straight line segments;

⑤Using the straight line segment rotation method to analyze the plane closed contour that only contains straight line segments, and obtain all closed loops;

⑥Search the target ring from all the rings;

⑦ Calculate the web finishing area according to the target ring to obtain the driving geometry.

Through the above driving information, use the mature tool path algorithm to generate the NC program.

In the NC program, insert the type of line, that is, straight line, circular arc, conic curve, free curve, and the type of surface, that is, plane, ruled surface, free-form surface, etc. geometric information for different tool positions, and insert and process at the same time Process information such as the type of NC machining operation when corresponding to the feature; the characteristics of the NC system considered in the post-processing, including acceleration characteristics, interpolation accuracy characteristics, tool holder compensation control and other parameter types; based on feature geometry and process information, combined According to the dynamic characteristics of the CNC system, the tool position point is fitted to form a polynomial interpolation method supported by the CNC system, and the feed rate is optimized to adapt to complex CNC machining conditions. Feature-based postprocessing consists of the following steps:

Step 1. Geometric and process information related to post-processing is expressed based on features. In the tool path generation process of CAM software, the ID (IDentification) of the feature of the tool path is obtained and identified. According to the driving geometry information corresponding to the tool path, the statement segment of the tool path is identified as a straight line or a curved path. Add the original processing operation type information in the sentence segment.

Step 2: Input the discrete tool path points in the post-processing, determine the geometric type of the tool path points that belong to the tool path segment according to the above information, and directly output the NC of linear interpolation for the discrete tool path points that belong to the linear tool path ( NumericalControl) program, which directly outputs the NC program of circular interpolation for the discrete tool track points belonging to the circular arc tool track.

Step 3. For the tool path points belonging to the curve (at least five points are required for fitting), the fitting is performed by substituting the conventional expression equation Ax2+Bxy+Cy2+Dx+Ey+1=0 of the conic curve equation, the formula Among them, x and y are parameters, and the coefficients A, B, C, D, and E are all real numbers and none of A, B, and C are zero.

Step 4. Compare the fitted curve with the original point position to calculate the fitting error. If the fitting error exceeds the specified value, the fitted curve segment ends and can be directly substituted into the corresponding pre-programmed standard curve program Generate the corresponding NC program, and then re-input the discrete tool path points for fitting, that is, repeat the second to fourth steps until the entire curve segment has a corresponding NC program output.

Step 2: All the point files are discrete tool path points, and the algorithm reads the points sequentially to generate NC programs continuously. Every time a discrete point is read, the system will make a judgment. If the current input point is Pi, the system will judge whether the tool path at the Pi point is a straight line or a curve according to the driving geometry keyword where Pi is located.

Step 3. After judgment, if Pi is on a straight line or arc tool path, and the point before Pi is on a curved (non-arc) tool path, then output the curve interpolation NC program before Pi, and then output the corresponding to Pi NC program for linear or circular interpolation, and then go to step 2. The program example of linear and circular interpolation in NC program is as follows:

Step 4. If the point belongs to the part of the curved tool path, you need to continue to input the point Pi until there are more than 5 points to be fitted. Fit the discrete tool path points belonging to the curved segment to obtain the conic curve Regular form equations.

The conventional expression of the conic section equation is Ax2+Bxy+Cy2+Dx+Ey+1=0, x, y are parameters and A, B, C, D, E are real numbers, and A, B, C are not zero. This form includes three forms of ellipse, hyperbola and parabola.

Use this equation to perform least square fitting on the points to obtain the coefficients in the equation.

Step 5. Since the input requirements of NC programs of different numerical control systems are different, it is necessary to convert the parameters of the equation.

Step 6: Carry out error judgment on the fitting curve. When the number of fitting points increases, the error between the fitting curve and the point may increase. At this time, it is necessary to start fitting the curve with a new point. The error judgment is represented by the minimum distance from the discrete tool path point to the fitted curve. If the fitting error is greater than the set error, output the conic curve interpolation NC program corresponding to the fitting curve within the last generated error, and return to step two.

Step 7. If the fitting error is smaller than the set error, return to Step 2 and add new points. If the new point belongs to the linear tool path, go to step 3, judge whether the last fitted conic curve NC program has been output, and output the linear interpolation NC program; if the new point belongs to the curved tool path, go to step Fourth, continue to fit the curve.

Generate process control files based on feature expression, including geometric information: corner information, driving geometric information, etc., and process information: machining tools, cutting parameters, monitoring strategies, detection strategies, etc.; geometric information and process information and processing strategies based on features , carry out online monitoring, and realize the processing of monitoring results based on artificial intelligence methods; record the data in the processing process, which refers to the actual cutting parameters in the processing process, the operating power of the machine tool, and various problems in the processing process such as tool wear and chatter etc.; organize the above data with features as units. Step 1. Based on the feature organization, geometric process information related to the machining process is formed to form machining process control information. Geometric process information based on feature organization includes feature type, tool used, tooling strategy, width of cut, depth of cut, spindle speed,

Step 2. On-line monitoring of the processing process is carried out according to the processing process control information, and the processing of the monitoring results is realized based on the artificial intelligence method.

Step 3. Record the data during the processing, including the actual cutting parameters during the processing, the operating power of the machine tool, and various problems during the processing such as tool wear, chatter, etc.; the final process monitoring needs to be based on features. Organization of information, actual processing feed speed, depth of cut, width of cut, operating power, etc. are all represented by curves, and divided into units based on features. Finally, the organized processing data will be fed back for process optimization.

Based on the geometric information and process information related to feature expression and detection, the geometric information is: the surface type, curvature change, area, normal direction, theoretical geometric model of the intermediate processing state of the detection feature, etc.; the process information is: processing accuracy; feature-based The geometric information and process information construct the generation rules of the detection points of the detection features, generate the detection points of the detection features based on the features, and consider the intermediate state of the detection features to optimize the detection trajectory and generate detection files; install the probe for online detection, record the detection results and The results are analyzed, organized based on characteristics, and fed back to the process department as a basis for process optimization. Described online detection method based on characteristic expression, this method comprises the following steps:

Step 1. Input the overall information and define some information that cannot be obtained from the 3D model of the part, including the name and attribute information of the part, and the default tolerance information of each type of detection feature, etc., to support detection feature recognition and detection data generation. And output and save in the form of XML neutral file;

Step 2. According to the defined detection features, feature recognition is performed on the part, and the obtained feature information is stored in the form of a linked list. For different features, the data structure is different, but they have common attribute information, such as tolerance, feature number, feature orientation, etc.;

Step 3. Read in the detection feature information. You can read in the feature list file saved by feature recognition, or you can read in the existing detection data list file, and continue to edit the last detection point editing result; the read-in feature should include complete Feature information, providing feature information for the generation of detection data;

Step 4. According to the geometric information and process information of the input feature, under the framework of the online detection point creation specification, determine the type of detection feature. If it is an "axis" or a plane regular shape, follow the method of determining the number and distribution of detection points at equal intervals; If it is a curved surface, use the method of automatically creating surface detection points based on the curvature change matrix to automatically generate online detection detection points; for detection features with special detection requirements, it is necessary to manually intervene in the automatically generated detection data points to obtain satisfactory results. The detection point data;

Step 5. Based on the detection state, carry out online detection path planning with the feature as the unit according to the three levels of detection operation level, detection feature level, and feature level;

Step 6. Perform post-processing on the detection path planned in the previous step to obtain the detection feature detection data. Through coordinate transformation, the coordinates of the detection point data in the design coordinate system are transformed into the detection coordinate system. At the same time, according to the path information and machine tool configuration The information generates and detects NC codes and detection process files, and transmits them to CNC machine tools for online detection;

Step 7. Use the contact electric touch probe to measure the actual coordinates of the detection points, and perform error compensation on the obtained actual coordinates to obtain the final required correction data;

Step 8. Compare the error-compensated data with the theoretical value in the online inspection guidance document, analyze and process according to the requirements in the online inspection guidance document, obtain complete inspection results and final quality evaluation conclusions, and then evaluate the processing Whether the parts meet the requirements and the adjustment strategy, the test results can also be used as the basis for process optimization.

In the feature-based neural network man-hour prediction method, geometric information and process information are respectively: geometric information refers to the overall area, part height, number of grooves, ribs, holes, and contour features; process information refers to the material of the part and the machine tool required for processing. The feature-based neural network man-hour prediction method includes the following steps:

Step 1, read the automatic feature recognition results of the aircraft structural parts shown in Figure 2;

Step 2. Extracting typical processing characteristic parameters, typical processing characteristic parameters include:

a. Regarding rib features: for horizontal rib roofs, the characteristic parameters that affect the processing time are the degree of the processing drive line of the horizontal rib roof; for oblique rib roofs, the characteristic parameters that affect the processing time are the length of the The inclination angle of the top of the rib; for the top of the arc, the characteristic parameter that affects the processing time is the length of the processing drive line of the top of the arc; Figure 2 shows a schematic diagram of the characteristics of a typical top of the rib;

b. Regarding groove features: For the groove web feature, the characteristic parameter that affects its processing time is the area of the web, and if the groove contains a boss, it is also treated as the web surface; for the internal groove type, the characteristic parameter that affects its processing time is the length of the inner drive line and the height of the groove; for the groove corner, the characteristic parameters that affect the processing time are corner radius, corner angle, and corner depth;

c. About the contour: the feature parameters that affect its processing time are the circumference of the part contour and the height of the contour;

d. About the round hole: the characteristic parameters that affect its processing time are the radius of the round hole and the depth of the round hole;

e. Regarding non-circular holes: the characteristic parameters that affect the processing time are the circumference of the non-circular hole and the depth of the non-circular hole;

The above features and their parameters are implemented using feature recognition and saved as feature information.

Step 3. Extract the cutting parameters that affect the working hours. When the workpiece, tool, machine tool and other processing resources are determined, the main factors affecting production efficiency are: cutting speed, feed rate, cutting depth and cutting width. The cutting parameters are as follows:

For rib top features, groove internal features, and hole features, the cutting parameters that affect the processing time are: speed, feed rate, depth of cut;

For groove web features, groove corner features, contour features, and non-circular hole features, the cutting parameters that affect the processing time are: speed, feed rate, width of cut, depth of cut;

Step 4. Carry out cutting simulation experiments, with features as the research object, based on typical parts processing technology and refined cutting parameters, by continuously changing the size features of features, recording the processing time of features of different sizes during the test, forming a feature-based processing time library. The selection principles of feature parameters are as follows:

a. Rib roof: 1) For the horizontal rib roof, the length of its processing driving line is set at 10-1500mm. The length of the processing driving line of the horizontal rib roof samples selected from the experimental results is in the above range, and the data selected by Penguin meet the requirements. Requirements for normal distribution; 2) For the oblique rib roof, the length of its processing driving line is set at 10-200 mm, and the range of inclination angle is set at 10-80°. The processing driving line length and The inclination range is within the above range, and the selected data meet the requirements of normal distribution; 3) For the arc rib top, its processing drive line length range is set at 10-100 mm. The processing drive of the arc rib top sample selected from the experimental results The line length is within the above range, and the selected data meet the requirements of normal distribution; 50 sets of sample data are taken for each type of rib top feature;

b. Groove: 1) For the groove web, its area range is positioned as 0.001-0.1m ² , and the selected groove web sample area parameters are distributed in this range with reference to the normal distribution; 2) For the groove type, the groove type The length of the driving line is set at 100-1200mm, and the depth of the groove is set at 3-100mm. The length and depth of the selected driving line are distributed in the range according to the normal distribution, and a total of 60 sets of data are taken; 3) For the groove corner, the corner radius data is set to 6mm, 10mm, 12mm, 16mm, 20mm, the included angle data is set to 30°, 45°, 60°, 75°, 90°, 120°, and the corner depth range is set to 3-100mm , the depth data meets the requirements of normal distribution, and a total of 160 sets of groove corner sample data are taken;

c. Outline: the perimeter range is 1000-6000mm, and the outline height is set at 20-180mm. The perimeter and outline height data of the selected outline are distributed in the above range according to the normal distribution, and 40 sets of sample data are selected;

d. Round hole: the aperture range is set to 8-80mm, the hole depth range is set to 3-50mm, the diameter data of the selected hole is set to: 8mm, 10mm, 20mm, 30mm, 60mm, 80mm, and the depth data is normally distributed in In the above range, 50 sets of data are taken under different machining accuracy conditions;

e. Non-circular holes: The perimeter range of non-circular holes is set at 10-1000mm, and the depth of non-circular holes is set at 3-50mm. The selected perimeter and depth data of non-circular holes are distributed in this range according to the normal distribution. Select 50 sets of sample data;

Step 5: Establish the corresponding relationship between features and feature processing time, and establish a certain number of feature instances for each type of feature. Form a typical feature sample library.

Step 6: Based on the established typical feature sample library, a BP network corresponding to the feature is established. The input nodes of the BP network are the geometric feature parameters of each type of feature, related cutting parameters and other possible influencing factors, and the output of the BP network is the processing time of each type of feature. The BP network of each type of feature is trained through the samples of the feature sample library and is marked and saved.

In order to predict the processing time of a new part instance, first obtain the features of the part, and use the above-mentioned feature geometric parameters and cutting parameters as the input of the corresponding trained BP neural network to predict, and the output of the BP network is the processing time of the feature , and finally the total processing time of the part is obtained by accumulating the processing time of all features of the part.

In the feature-based precise machining man-hour prediction method, the geometric information refers to the corner position and corner size of the feature; the process information refers to the cutting allowance, depth of cut, width of cut, and feed rate; the dynamic characteristics of the machine tool include the acceleration, jerk, Turning performance, based on geometric information, process information and the dynamic characteristics of the machine tool, predict the actual feed speed during the machine tool movement, and combine the cutting force to predict the change of the feed speed caused by the change of the cutting force, and then accurately predict the feed rate Speed, and then the geometric tool path to calculate the processing time. The feature-based precise machining man-hour prediction method includes the following steps:

Step 1. Read the NC program obtained by performing automatic NC programming on the machined parts;

Step 2, organizing NC programs and machine tool characteristic information. In the NC program, the data is organized in units of features, and each feature is coded, and each feature contains the feature code and the NC data for processing the feature. Through feature coding, NC program information can be accessed and called. The processing data corresponding to the feature under the feature code maintains the data format of the existing NC program. Each NC program file has the machine tools required to complete the program.

Machine tool characteristic information is stored in the form of knowledge base, which is divided into three components: machine tool library, CNC system library and speed control method library. The machine tool library includes the name of the machine tool, the CNC system used by the machine tool, and the performance parameters of the machine tool such as maximum power, maximum speed, and maximum acceleration.

Step 3, organizing geometric process information based on features. Geometric process information organizes data as a result of process decision-making, that is, organizes data according to the three-level structure of process, process step and operation. Machine tool and tool information are stored under process and step respectively. The machine tool information here only contains a brief description of the machine tool, including the name and category of the machine tool. Tool information mainly describes the tool code, material, structural parameters, etc. The operation information includes feature code, cutting speed, machining speed, feature position, machining allowance, depth of cut and width of cut. The feature code includes the feature ID (Identification, identification) and the ID of the operation to be performed. The speed of the advance and retreat tool and the processing speed can be used to distinguish between the machining tool path and the advance and retreat tool path. The machining allowance, depth of cut and width of cut are used to determine the cutting status basis.

Step 4. Due to the complexity of setting the local value of the cutting speed in NC programming, the current NC programming only sets the local cutting speed at key positions such as corners, and other positions use the global cutting speed. Inhomogeneous, intelligent adjustment modules of modern machine tools such as Artis will optimize the cutting speed during machining.

The power required for cutting is determined by the cutting force and cutting speed, namely:

p=fv

In the formula, p is the power required for cutting, f is the cutting force, and v is the cutting speed.

Cutting force is related to cutting speed, cutting depth, cutting width, cutting material, tool material and processing method, and its calculation can be indexed to the cutting force calculation model.

In the prediction process, the cutting speed is the speed calculated according to the kinematic characteristics of the machine tool, and the depth of cut needs to be extracted from the feature information. In the actual cutting process, the actual cutting width will be different from the set value in the feature, and it needs to be calculated according to the part’s Machining status calculation. In order to improve the calculation efficiency, the calculation of cutting power is only calculated where there may be problems. According to the characteristics of feature processing, since the top surface of the rib feature and the feature web of the groove are processed after rough machining, the margin is large, and the The probability that the required power is large is relatively high, and it is used as the calculation object. Find the cutting power in the cutting process, and compare the calculated power with the power that the machine tool can provide. If the calculated power is smaller than the power that the machine tool can provide, you can run according to the value set by the NC program. Otherwise, it needs to be adjusted. The adjustment method is : According to the power calculation formula, gradually reduce the speed until the requirement is met.

Step 5: Carry out precise process-level man-hour calculation based on features, and the processing time of the NC process is composed of the processing time of each work step.

The tool path of the feature includes the rapid positioning tool path, the advance and retreat tool path and the cutting tool path, so the processing time of the feature is also composed of the rapid positioning time, the advance and retreat time and the cutting time.

The characteristic tool path is composed of rapid positioning tool path, advancing and retreating tool path and cutting tool path. The rapid positioning tool path is simple and only needs to calculate the theoretical time. The change of the tool advance and retreat is more complicated. If the analysis function of the machine tool is used to calculate, the efficiency is low. Since the tool advance and retreat is part of the feature tool path strategy, it belongs to the knowledge and experience carried by the feature. Each feature has a corresponding tool advance and retreat strategy. Through Through the experimental analysis, the proportional coefficient between the actual processing time and the theoretical time of the tool advance and retreat for each feature can be obtained, and the time for the tool advance and retreat can be obtained.

The parts not involved in the present invention are the same as the prior art or can be realized by adopting the prior art.

Claims (9)

1. A feature-based NC machining and manufacturing method for complex parts, characterized in that it first reads in the three-dimensional digital model of the part to be processed, and secondly, performs automatic feature recognition to extract the geometric information of the part to be processed from the read-in three-dimensional digital model and process information, the information not included in the 3D digital model is obtained through the relevant process documents matched with the digital model, and the extracted geometry and process information are characterized and marked, and all features are uniquely identified according to the feature category and serial number, that is, Feature identification = feature category + serial number, using this identification as a tool for feature association between different applications; if the feature is decomposed during the manufacturing process, then feature identification = decomposition identification + feature category + serial number + decomposition sub-serial number, so as to realize the use of features as The carrier of manufacturing knowledge and experience realizes the effective integration of information; thirdly, in order to maintain the validity of features and the continuity of feature evolution throughout the manufacturing process, object-oriented and ontology-based methods are used to represent features, that is, to define the ontology of features and For the class of features in each application field, the method of holographic attribute surface and edge graph is used for feature recognition; finally, the NC machining program for complex parts is automatically generated after process optimization, and the NC machining equipment is driven to complete the entire processing process; after the automatic feature recognition is completed, the Construction of processing technology optimization model based on features, automatic process decision-making based on features, automatic NC programming based on features, post-processing based on features, process control based on features, online detection based on features, man-hours of processed parts based on features The sample library and the mapping model of building feature geometric information and process information to machine tool performance parameters; (1) The feature-based construction of the processing technology optimization model should include the geometric information of the feature, the information of the tool, the information of the machine tool and the material information of the workpiece. The constraints of the optimization model are cutting force and vibration, and the optimization target of the optimization model is the processing Efficiency, processing quality and processing cost, the optimized items are depth of cut, width of cut and feed speed, and a feature-based cutting parameter library is established; a 3D model of cutting test specimens based on features is built for the application field, including typical features and typical Combination, as a standard sample to verify the processing technology; (2) The feature-based automatic process decision-making is to provide complete information support for automatic process decision-making, express process rules and process schemes based on features, and then perform automatic process decision-making based on features, and express the results of automatic process decision-making based on features; (3) The automatic NC programming based on features refers to using the results of automatic process decision-making as the basis of automatic programming, and then automatically generating machining tool paths based on features, and expressing the results of automatic programming based on features; (4) The post-processing based on the feature is based on the results of automatic programming, considering the dynamic characteristics of the numerical control system, performing post-processing based on the feature, and forming a numerical control machining program based on the result of the feature expression post-processing; (5) The feature-based processing process control is a processing strategy for expressing problems in the processing process, including state monitoring of the processing process based on features, intelligent self-adaptive adjustment during processing, and feature-based recording during processing. data for process optimization; (6) The feature-based online detection refers to the formation of detection point generation specifications based on the geometric information and process information of the features, the optimization of the probe attitude vector and detection trajectory during online detection based on the geometric information of the features, and the formation of detection and analysis results based on the features , as the basis for optimizing the processing technology; (7) The feature-based construction of the man-hour sample library of the processed parts contains geometric information of features related to man-hours and part man-hours. The feature geometric information and process information are used as input to construct a neural network. In the absence of part CNC programs Predict the working hours of the parts to be processed under the premise, as the basis for parts processing external quotations and production planning arrangements; add feature geometric information and process information to the feature-based NC program, combined with the dynamic characteristics of the CNC machine tool and the cutting force during the cutting process , to accurately predict the NC machining hours of each process, as the basis for machine tool scheduling; (9) The mapping model of building feature geometric information and process information to machine tool performance parameters is based on feature analysis to meet the performance parameters of the machine tool required for processing parts, and provides application-oriented design references for machine tool related parameters. 2. The method according to claim 1, wherein the step of performing automatic process decision-making based on features is: (1) expressing geometric information and process information related to process decision-making based on features, wherein the geometric information is: feature The maximum tool diameter allowed by the geometric size, position, geometric structure, and features of the tool, and the process information is: material, blank form, machining allowance and machining accuracy; (2) Process decision-making based on features is carried out in the order of machine tool decision-making, processing sequence decision-making, processing method decision-making and cutting parameter decision-making, and automatic process decision-making results are formed according to process rules and process plans; (3) The form of process decision-making results based on feature expression is: each process contains several steps, each step contains several operations, each operation corresponds to a processing feature, and the feature contains the operation type, processing strategy and cutting parameters. 3. method according to claim 1, it is characterized in that described automatic numerical control programming method based on feature is: (1) Geometric information and process information related to automatic programming are expressed based on features. The geometric information includes: the geometric size of the feature, the driving geometry required to generate the tool path, the starting point and the end point of the tool path, and the avoidance geometry of the tool path. The information includes: tool information for machining features, machining accuracy of features and tool path strategy; (2) Extract processing features through feature recognition, and output feature recognition results including geometric information and process information; (3) Based on the geometric information of the feature recognition result and the process information of the process decision result, the tool path of each processing feature is automatically generated; (4) Generating the tool position file of the machining tool path for each feature in units of features, and marking the corresponding information according to different application requirements. 4. The method according to claim 1, wherein said post-processing step based on features comprises: (1) Geometric information and process information related to post-processing are expressed based on features in the NC program, where geometric information refers to: the type of line in the geometry of the feature: straight line, circular arc, conic section, free curve, and surface Type: plane, ruled surface, free-form surface, the process information refers to the type of CNC machining operation when machining the corresponding feature; (2) The characteristics of the numerical control system considered in post-processing, including acceleration characteristics, interpolation accuracy characteristics and tool holder compensation control parameter types; (3) Based on the feature geometry and process information, combined with the dynamic characteristics of the CNC system, the tool position points are fitted to form a polynomial interpolation method supported by the CNC system, and the feed rate is optimized to adapt to complex CNC machining conditions. 5. The method according to claim 1, wherein the step of said feature-based process control comprises: (1) Geometric information and process information related to machining process control based on feature expression, where geometric information is: corner information and driving geometric information, and process information is: machining tool, cutting parameters, monitoring strategy and detection strategy; (2) According to the geometric information and process information of the feature and the processing strategy, online monitoring is carried out, and the processing of the monitoring results is realized based on artificial intelligence methods; (3) The data recorded during the processing refers to the actual cutting parameters during the processing, the operating power of the machine tool, the tool wear and chatter during the processing; the above data are organized in units of features. 6. The method according to claim 1, characterized in that said online detection based on features comprises the following steps: (1) Based on the geometric information and process information related to feature expression and detection, the geometric information is: the surface type, curvature change, area, normal direction and theoretical geometric model of the intermediate processing state of the detection feature; the process information is: processing accuracy; (2) Construct the generation rule of the detection point of the detection feature based on the geometric information and process information of the feature, generate the detection point of the detection feature based on the feature, and consider the intermediate state of the detection feature to optimize the detection track and generate the detection file; (3) Install probes for on-line detection, record the detection results and analysis results, organize based on characteristics, and feed back to the process department as the basis for process optimization. 7. The method according to claim 1, characterized in that the feature-based processing technology optimization model considers the geometric information and process information of the feature, wherein the geometric information refers to: the geometric size, feature The thickness and height of the constituent geometric elements. 8. The method according to claim 1, wherein the geometric information based on the feature-based construction of the man-hour sample library of the machined parts includes corner position and corner size; process information includes cutting allowance, depth of cut, The main feed speed of cutting width; the dynamic characteristics of the machine tool include the acceleration, jerk and turning performance of the machine tool. Based on the geometric information, process information and the dynamic characteristics of the machine tool, the actual feed speed during the machine tool movement is predicted, combined with the cutting force prediction The change in feed rate due to the change in cutting force can be accurately predicted, and the machining man-hour can be calculated based on the tool path. 9. The method according to claim 1, wherein the object-oriented and ontology-based feature representation method is: (1) First define the feature ontology. The feature ontology includes the unique identification of the feature, the identification surface, and the seed identification surface that may become the identification surface. The definition of the seed identification surface is due to the evolution of the feature during the manufacturing process, and the feature may be decomposed; (2) The feature-based ontology derives feature classes in various application fields. The classes inherit the attributes of the ontology, and the class contains geometric information and process information. According to the needs of different application fields, different geometric information and process information are defined, so as to realize Information integration and transmission throughout the manufacturing process.