CN118862466B - An optimization method for digital design of excavators - Google Patents
An optimization method for digital design of excavators Download PDFInfo
- Publication number
- CN118862466B CN118862466B CN202410897444.3A CN202410897444A CN118862466B CN 118862466 B CN118862466 B CN 118862466B CN 202410897444 A CN202410897444 A CN 202410897444A CN 118862466 B CN118862466 B CN 118862466B
- Authority
- CN
- China
- Prior art keywords
- excavator
- undirected graph
- feature data
- model
- feature
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/042—Knowledge-based neural networks; Logical representations of neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Geometry (AREA)
- Computer Graphics (AREA)
- Computer Hardware Design (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an optimization method for digital design of an excavator, and particularly relates to the technical field of excavator design, and the technical scheme is characterized in that characteristic data of the excavator are obtained, a digital twin-based three-dimensional model of the excavator is constructed, and an excavator simulation working environment is created; the method comprises the steps of carrying out space fusion processing on characteristic data by using a multi-layer perceptron to obtain a characteristic data set, constructing a characteristic undirected graph by utilizing the characteristic data in the characteristic data set, inputting the characteristic undirected graph into a graph neural network, combining a multi-head attention mechanism to carry out processing to obtain importance scores of nodes of the characteristic undirected graph, wherein the characteristic data is used as the nodes of the undirected graph, the association relation among the characteristic data generates edges of the undirected graph, constructing an initial excavator relationship model based on a deep neural network, inputting the characteristic data and the importance scores into the initial excavator relationship model to carry out training, and obtaining the excavator relationship model finally used for generating the excavator structural association coefficient after training.
Description
Technical Field
The invention relates to the technical field of excavator design, in particular to an optimization method for digital design of an excavator.
Background
In recent years, under the continuous promotion of technological development, more advanced technologies are applied to large mining excavators aiming at specific tasks in the design and manufacturing processes, and the overall technical level of the excavators is greatly improved. The novel theory, the novel method, the novel technology and the like adopted in the aspects of overall size, structural parameters, structural design, mechanism design, performance parameters, electrical design and the like play a role in promoting the improvement of the reliability and performance of the excavator products to different degrees.
In the field of digitalization of the excavator, key technical researches on digitalization design of the excavator mainly surround the conditions of random, vibration and impact load of the excavator under the condition of complex natural working conditions and the use requirements of high reliability and long service life, and in the existing excavator design process, the mapping mechanism between dynamic performance and design parameters is fuzzy, and the association relation between the optimal dynamic performance and the design parameters is not established, so that the related problems of influencing the design of the excavator structure are solved.
Accordingly, the present invention is directed to a method of optimizing a digital design of an excavator, which solves the above-mentioned related problems.
Disclosure of Invention
The invention aims to solve the technical problems that in the existing excavator design process, the mapping mechanism between the dynamic performance and the design parameters is fuzzy, the association relation between the optimal dynamic performance and the design parameters is not established, thereby influencing the design of the excavator structure, and provides an optimization method for the digital design of the excavator, and constructing a feature undirected graph based on the feature data, inputting the feature undirected graph into a graph neural network, combining a multi-head attention mechanism, processing to obtain importance scores of nodes of the feature undirected graph, and finally constructing an excavator relationship model for generating excavator structure association coefficients by combining the feature data and the importance scores through a neural network model, thereby solving the related problem that the association relationship between optimal dynamics performance and design parameters is not established in the prior art, and further influencing the design of the excavator structure.
The invention is realized by the following technical scheme:
An optimization method for digital design of an excavator, comprising the following steps:
acquiring characteristic data of the excavator, constructing a digital twin-based three-dimensional model of the excavator, and creating an excavator simulation working environment;
performing space fusion processing on the characteristic data by using a multi-layer sensor to obtain a characteristic data set;
Constructing a feature undirected graph by utilizing feature data in a feature data set, inputting the feature undirected graph into a graph neural network, and processing the feature undirected graph by combining a multi-head attention mechanism to obtain importance scores of nodes of the feature undirected graph, wherein the feature data are used as nodes of the undirected graph, and the association relationship among the feature data generates edges of the undirected graph;
constructing an initial excavator relationship model based on a deep neural network, inputting the feature data and the importance score into the initial excavator relationship model for training, and obtaining an excavator relationship model finally used for generating the excavator structure association coefficient after training;
Inputting the feature data to be processed into an excavator relation model to output an excavator structure association coefficient of the association relation between the feature data and the excavator structural part, and correcting the excavator three-dimensional model by using the excavator structure association coefficient;
operating an excavator three-dimensional model in an excavator simulated working environment, and returning various performance indexes and rewarding values of the excavator according to the operation condition of the excavator three-dimensional model;
Correcting the three-dimensional model of the excavator by utilizing various performance indexes and rewarding values of the excavator, inputting the corrected three-dimensional model of the excavator into the excavator relation model to output the association coefficient of the excavator structure, and repeatedly and iteratively correcting until the rewarding values are stable to obtain the optimal three-dimensional model of the excavator.
Further, the multi-layer perceptron is utilized to carry out space fusion processing on the characteristic data to obtain a characteristic data set, which is specifically as follows:
Setting class weights for the feature data of the excavator based on preset feature classes, extracting pearson correlation coefficients between the feature data, marking the feature data with the pearson correlation coefficients larger than a preset coefficient threshold value, obtaining a plurality of marked feature data, and respectively setting association weights for the marked feature data according to the pearson correlation coefficients of the marked feature data;
Based on the category weights and the associated weights, the multi-layer perceptron is utilized to perform space fusion processing on the feature data, so as to obtain a feature data set.
Further, a feature undirected graph is constructed by utilizing feature data in a feature data set, the feature undirected graph is input into a graph neural network and is processed by combining a multi-head attention mechanism, so that importance scores of undirected graph nodes are obtained, wherein the feature data is used as nodes of the undirected graph, and the association relationship among the nodes generates edges of the undirected graph, and the method specifically comprises the following steps:
Constructing an undirected graph according to the feature data set, taking each feature data in the feature data set as a node of the undirected graph, acquiring an association relation between each node according to the pearson correlation coefficient, and generating an edge of the undirected graph based on the association relation between each node to obtain a feature undirected graph associated with each feature data;
Obtaining the maximum mutual information coefficient between different nodes of the feature undirected graph, and constructing an adjacent matrix of the feature undirected graph according to the maximum mutual information coefficient;
Inputting an adjacency matrix of the feature undirected graph into a graph neural network, outputting neighbor node vectors of a plurality of neighbor nodes by the graph neural network, and acquiring multi-head attention coefficients of the plurality of neighbor nodes in the adjacency matrix based on a multi-head attention mechanism;
and multiplying the neighbor node vectors of the plurality of neighbor nodes with the corresponding multi-head attention coefficients respectively, and averaging to obtain the importance scores of the feature undirected graph nodes.
Further, an initial excavator relationship model is built based on the deep neural network, the feature data and the importance score are input into the initial excavator relationship model for training, and the excavator relationship model which is finally used for generating the excavator structure association coefficient is obtained after training, specifically:
Acquiring a self-encoder network, and taking the self-encoder network as an input layer of a deep neural network model to obtain an initial excavator relationship model;
Optimizing a network structure of the initial excavator relationship model by utilizing a particle swarm algorithm to obtain the initial excavator relationship model with an optimal structure;
And inputting the characteristic data and the importance score into an initial excavator relationship model for training, and obtaining the excavator relationship model finally used for generating the excavator structure association coefficient after training.
Further, the network structure of the initial excavator relationship model is optimized by utilizing a particle swarm algorithm, and the initial excavator relationship model with the optimal structure is obtained, specifically:
initializing parameters of a particle swarm algorithm, particle position and speed information, taking the output precision of an initial excavator relation model as fitness, and acquiring an individual extremum and a global extremum of particles;
Performing iterative optimization based on a simulated annealing algorithm, and updating the particle position and speed by using a simulated annealing neighborhood search;
Stopping optimizing when a preset termination condition is met, acquiring optimal structure parameters of the encoder network according to the optimal particle positions, and constructing an initial excavator relationship model of the optimal structure based on the optimal structure parameters.
Further, the preset feature categories include excavator work environment, power performance and work performance.
Further, the simulated working environment of the excavator comprises farmlands, mountain lands, cities and river channels.
The invention also provides an optimization system for digital design of the excavator, which comprises:
the feature data acquisition module is used for acquiring feature data of the excavator, constructing a digital twin-based three-dimensional model of the excavator and creating an excavator simulation working environment;
The characteristic data set construction module is used for carrying out space fusion processing on the characteristic data by utilizing the multi-layer perceptron to obtain a characteristic data set;
The undirected graph construction module is used for constructing a characteristic undirected graph by utilizing characteristic data in a characteristic data set, inputting the characteristic undirected graph into a graph neural network, and processing the characteristic undirected graph by combining a multi-head attention mechanism to obtain importance scores of nodes of the characteristic undirected graph, wherein the characteristic data is used as nodes of the undirected graph, and the association relationship among the characteristic data generates edges of the undirected graph;
The model training module is used for constructing an initial excavator relationship model based on the deep neural network, inputting the feature data and the importance score into the initial excavator relationship model for training, and obtaining an excavator relationship model finally used for generating the excavator structure association coefficient after training;
The structure correction module is used for inputting the characteristic data to be processed into the excavator relation model so as to output an excavator structure association coefficient of the association relation between the characteristic data and the excavator structural part, and correcting the excavator three-dimensional model by utilizing the excavator structure association coefficient;
The simulation operation module is used for operating the three-dimensional model of the excavator in the simulated working environment of the excavator and returning various performance indexes and rewarding values of the excavator according to the operation condition of the three-dimensional model of the excavator;
the structure correction module is also used for correcting the three-dimensional model of the excavator by utilizing various performance indexes and rewarding values of the excavator, inputting the corrected three-dimensional model of the excavator into the excavator relation model to output the association coefficient of the excavator structure, and repeatedly and iteratively correcting until the rewarding values are stable so as to obtain the optimal three-dimensional model of the excavator.
The invention also provides a computer device comprising a memory storing a computer program and a processor implementing the steps of any of the methods described above when the processor executes the computer program.
The invention also provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of the method of any of the above.
Compared with the prior art, the invention has the following advantages and beneficial effects:
According to the invention, the characteristic data of the excavator are acquired, the spatial fusion processing is carried out on the characteristic data by utilizing the multi-layer perceptron, a characteristic undirected graph is constructed based on the characteristic data, the characteristic undirected graph is input into a graph neural network and is processed by combining a multi-head attention mechanism, the importance score of the characteristic undirected graph node is obtained, finally, the neural network model is combined with the characteristic data and the importance score, an excavator relationship model for generating the excavator structure association coefficient is constructed, and the problem that the association relation between the optimal dynamics performance and the design parameter is not established in the prior art, so that the influence on the design of the excavator structure is caused is solved.
Drawings
In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the drawings that are needed in the examples will be briefly described below, it being understood that the following drawings only illustrate some examples of the present invention and therefore should not be considered as limiting the scope, and that other related drawings may be obtained from these drawings without inventive effort for a person skilled in the art. In the drawings:
FIG. 1 is a flow chart of an optimization method for digital design of an excavator according to the present embodiment;
FIG. 2 is a flowchart of a method for optimizing the digitized design of the excavator in step S3;
FIG. 3 is a flowchart of a method for optimizing the digitized design of the excavator in the present embodiment at step S4;
FIG. 4 is a schematic block diagram of an optimization system for digital design of an excavator according to the present embodiment;
fig. 5 is a schematic structural diagram of a computer device.
Detailed Description
Exemplary embodiments of the present disclosure are described below in conjunction with the accompanying drawings, which include various details of the embodiments of the present disclosure to facilitate understanding, and should be considered as merely exemplary. Accordingly, one of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
In the present disclosure, the use of the terms "first," "second," and the like to describe various elements is not intended to limit the positional relationship, timing relationship, or importance relationship of the elements, unless otherwise indicated, and such terms are merely used to distinguish one element from another. In some examples, a first element and a second element may refer to the same instance of the element, and in some cases, they may also refer to different instances based on the description of the context.
The terminology used in the description of the various examples in this disclosure is for the purpose of describing particular examples only and is not intended to be limiting. Unless the context clearly indicates otherwise, the elements may be one or more if the number of the elements is not specifically limited. Furthermore, the term "and/or" as used in this disclosure encompasses any and all possible combinations of the listed items.
Example 1
An optimization method for digital design of an excavator, comprising the following steps:
s1, acquiring characteristic data of an excavator, constructing a digital twin-based three-dimensional model of the excavator, and creating an excavator simulation working environment, wherein the excavator simulation working environment comprises farmlands, mountain areas, cities and river channels in the embodiment;
In this embodiment, a basic module in a simulation software simulink model library is called according to structural parts of the excavator to construct a three-dimensional model of the excavator, so that the proportion, texture and material of the model are correct, the three-dimensional space positions of all parts and the coordination motion relation information among all parts are ensured, and the motion synchronization of the virtual three-dimensional model is realized. Automatically generating a PLC code by packaging the constructed three-dimensional model of the excavator, performing simulation operation on the constructed three-dimensional model of the excavator through a solver, performing error checking on the constructed three-dimensional model of the excavator, wherein parameter configuration comprises the steps of setting the step length of the solver, setting a system target file generated by the PLC code, setting annotation options and a report format, setting the solver to be a fixed step length, setting the basic sampling period of the three-dimensional model of the excavator to be consistent with the interrupt period of a hardware chip of a target controller, placing the generated PLC code in a selected compiling environment to integrate to generate an executable program file, planning an excavating action for the specific shape of an excavating material pile, and controlling related units of the PLC to move a stepping motor rotary encoder in a planning driving unit to control related parts in the action realizing unit to perform excavating operation according to a specified excavating track;
Meanwhile, the characteristic data of the excavator comprise characteristic parameters of different types of the excavator, the characteristic data are classified and arranged according to working environments, dynamic performances and working performances of the excavator, the simulated working environments of the excavator are created, and 3D rendering and establishment can be carried out on working scenes of the excavator in a mode of combining aerial modeling and manual modeling.
The dynamics and the working performance mainly comprise tonnage, engine power, bucket capacity, action speed, climbing capacity, excavation angle, hydraulic and electric parameters and the like.
Specifically, in the embodiment, by constructing the three-dimensional model of the excavator based on digital twin, the correction optimization change of the excavator structure can be displayed truly and intuitively in real time, and meanwhile, the correction process can be monitored so as to perform optimization regulation and control on the correction process, so that error information is reduced.
S2, performing space fusion processing on the characteristic data by using a multi-layer sensor to obtain a characteristic data set, wherein the method specifically comprises the following steps:
S201, setting class weights for feature data of an excavator based on preset feature classes, wherein the preset feature classes comprise working environments, power performances and working performances of the excavator, extracting pearson correlation coefficients among the feature data, marking feature data with pearson correlation coefficients larger than a preset coefficient threshold value, obtaining a plurality of marked feature data, and respectively setting association weights for the marked feature data according to the pearson correlation coefficients of the plurality of marked feature data;
s202, based on the category weight and the association weight, performing space fusion processing on the feature data by using the multi-layer perceptron to obtain a feature data set.
It should be noted that, the pearson correlation coefficient is used to measure the linear correlation degree between two feature data, and its value is between-1 and 1, and typically, the pearson correlation coefficient between two variables is defined as the quotient (or normalized covariance) of the product of the covariance and the standard deviation between the two variables.
Specifically, in this embodiment, based on the working environment, the power performance and the working performance of the excavator, a class weight is set for the feature data, for example, the class weight of the feature data belonging to the class of the working environment of the excavator is 20%, the class weight of the feature data belonging to the class of the power performance is 40%, the class weight of the feature data belonging to the class of the working performance is 40%, and the actual class weight is determined according to the actual situation.
And S3, constructing a feature undirected graph by utilizing feature data in a feature data set, inputting the feature undirected graph into a graph neural network, combining a multi-head attention mechanism, processing to obtain importance scores of feature undirected graph nodes, screening a preset number of feature data according to the importance scores, and constructing an excavator relationship model according to parameter indexes corresponding to the feature data as references. The feature data are used as nodes of the undirected graph, and the association relation among the feature data generates edges of the undirected graph, specifically:
s301, constructing an undirected graph according to a feature data set, taking each feature data in the feature data set as a node of the undirected graph, acquiring an association relation between each node according to a Pearson correlation coefficient, and generating an edge of the undirected graph based on the association relation between each node to obtain a feature undirected graph associated with each feature data;
S302, obtaining maximum mutual information coefficients among different nodes of the feature undirected graph, and constructing an adjacent matrix of the feature undirected graph according to the maximum mutual information coefficients;
s303, inputting an adjacency matrix of the feature undirected graph into a graph neural network, outputting neighbor node vectors of a plurality of neighbor nodes by the graph neural network, and acquiring multi-head attention coefficients of a plurality of neighbor nodes in the adjacency matrix based on a multi-head attention mechanism;
And S304, multiplying the neighbor node vectors of the plurality of neighbor nodes by the corresponding multi-head attention coefficients respectively, and averaging to obtain the importance scores of the feature undirected graph nodes.
It should be noted that in this embodiment, the Multi-head Attention mechanism is used to calculate the importance score of the undirected graph node as a conventional technical means, and the Multi-head Attention mechanism (Multi-head Attention) is a mechanism widely used in the transform model, and captures different feature representations through multiple independent Attention heads. In a graph neural network, a multi-headed attention mechanism is used to handle relationships and information transfer between nodes.
Specifically, in this embodiment, the maximum mutual information coefficient between different nodes of the undirected graph is used to measure the relevance between two nodes, where the range is between 0 and 1, when the maximum mutual information coefficient is 0, the two nodes are completely independent, when the maximum mutual information coefficient is 1, the two nodes have a complete dependency relationship, and an adjacent matrix of the undirected graph is constructed according to the maximum mutual information coefficient by creating an n×n matrix (N is the number of nodes), where the matrix is used as an adjacent matrix of the undirected graph, and for each pair of nodes (X, Y), if the maximum mutual information coefficient exceeds a certain threshold, the corresponding positions (i, j) and (j, i) in the adjacent matrix are set to 1, to indicate that an edge exists between the two nodes;
Meanwhile, an adjacency matrix of the feature undirected graph and a feature matrix of the nodes (the feature matrix of the nodes is an N multiplied by F matrix, F is the feature number of each node) are input into the graph neural network, the graph neural network outputs neighbor node vectors of a plurality of neighbor nodes, meanwhile, a multi-head attention mechanism is arranged in the graph neural network, multi-head attention operation is carried out on each node aiming at the neighbor nodes, and multi-head attention coefficients of the plurality of neighbor nodes in the adjacency matrix are obtained.
S4, constructing an initial excavator relationship model based on a deep neural network, inputting feature data and importance scores into the initial excavator relationship model for training, and obtaining a final excavator relationship model after training, wherein the method specifically comprises the following steps:
S401, acquiring a self-encoder network, and taking the self-encoder network as an input layer of a deep neural network model to obtain an initial excavator relationship model;
s402, optimizing a network structure of an initial excavator relationship model by utilizing a particle swarm algorithm to obtain the initial excavator relationship model with an optimal structure, wherein the initial excavator relationship model specifically comprises the following steps:
S403, initializing parameters of a particle swarm algorithm, particle position and speed information, taking the output precision of an initial excavator relation model as fitness, and acquiring individual extremum and global extremum of particles;
s404, performing iterative optimization based on a simulated annealing algorithm, and updating the position and the speed of particles by using a simulated annealing neighborhood search;
and S405, stopping optimizing when a preset termination condition is met, acquiring the optimal structure parameters of the self-encoder network according to the optimal particle positions, and constructing an initial excavator relationship model of the optimal structure based on the optimal structure parameters, wherein in the embodiment, the structure parameters of the self-encoder network comprise network layers, neural network weights and thresholds.
S406, inputting the feature data and the importance score into an initial excavator relationship model for training, and obtaining the excavator relationship model finally used for generating the excavator structure association coefficient after training.
Specifically, in this embodiment, the excavator structural association coefficient is used to characterize a corresponding association relationship between the excavator operating condition and the structural component.
S5, inputting the feature data to be processed into an excavator relation model to output an excavator structure association coefficient of the association relation between the feature data and the excavator structural part, and correcting the excavator three-dimensional model by using the excavator structure association coefficient;
S6, operating the three-dimensional model of the excavator in the simulated working environment of the excavator, and returning various performance indexes and rewarding values of the excavator according to the operation condition of the three-dimensional model of the excavator;
Specifically, in this embodiment, by running the constructed visual three-dimensional model of the excavator in the simulated working environment of the excavator, it is convenient to obtain whether each performance index of the excavator after preliminary correction after running reaches a preset target and meets the actual condition, and a reward value is output according to each performance index to correct the three-dimensional model of the excavator.
And S7, correcting the three-dimensional model of the excavator by utilizing the performance indexes and the rewarding values of the excavator, inputting the corrected three-dimensional model of the excavator into the relation model of the excavator to output the association coefficient of the excavator structure, and repeatedly and iteratively correcting until the rewarding values are stable to obtain the optimal three-dimensional model of the excavator.
In this embodiment, the reward value of the deep neural network is based on the judgment of various performance indexes generated by running the excavator in the simulated working environment of the excavator, for example, the movement speed is too slow in the working environment, and the corresponding production reward value optimizes the characteristic data and the component structure parameters affecting the running speed in the excavator.
Specifically, in this embodiment, feature data of an excavator is obtained, spatial fusion processing is performed on the feature data by using a multi-layer perceptron, a feature undirected graph is built based on the feature data, the feature undirected graph is input into a graph neural network and processed by combining a multi-head attention mechanism to obtain an importance score of a feature undirected graph node, finally, an excavator relationship model for generating an excavator structure association coefficient is built by combining the feature data and the importance score by using a neural network model, the problem that an association relationship between optimal dynamics performance and design parameters is not built in the prior art, so that the design of the excavator structure is affected is solved, meanwhile, the modified excavator twin structure is put into a built excavator simulation working environment to operate, performance indexes of the excavator can be obtained, and an optimized rewarding value for modifying the structure parameters and the feature data in the operation process is obtained based on judging deviation of the performance indexes.
Example 2
The invention also provides an optimization system for digital design of the excavator, which comprises:
The feature data acquisition module 100 is used for acquiring feature data of the excavator, constructing a digital twin-based three-dimensional model of the excavator, and creating an excavator simulation working environment;
The feature data set construction module 200 is configured to perform spatial fusion processing on feature data by using a multi-layer sensor to obtain a feature data set;
the undirected graph construction module 300 is configured to construct a feature undirected graph by using feature data in a feature data set, input the feature undirected graph into a graph neural network, and process the feature undirected graph in combination with a multi-head attention mechanism to obtain an importance score of a node of the feature undirected graph, wherein the feature data is used as a node of the undirected graph, and an association relationship between the feature data generates an edge of the undirected graph;
The model training module 400 is configured to construct an initial excavator relationship model based on the deep neural network, input the feature data and the importance score into the initial excavator relationship model for training, and obtain an excavator relationship model finally used for generating the excavator structural association coefficient after training;
The structure correction module 500 is configured to input feature data to be processed into the excavator relationship model, so as to output an excavator structure association coefficient of an association relationship between the feature data and an excavator structural component, and correct the excavator three-dimensional model by using the excavator structure association coefficient;
The simulation running module 600 is used for running the three-dimensional model of the excavator in the simulated working environment of the excavator and returning various performance indexes and rewarding values of the excavator according to the running condition of the three-dimensional model of the excavator;
The structure correction module 500 is further configured to correct the three-dimensional model of the excavator by using each performance index and the reward value of the excavator, input the corrected three-dimensional model of the excavator into the excavator relation model, output the association coefficient of the excavator structure, and repeatedly iterate the correction until the reward value is stable, so as to obtain an optimal three-dimensional model of the excavator.
It should be noted that the modules in the system of embodiment 2 correspond to the steps in the method of embodiment 1, and the steps in the method of embodiment 1 are described in detail in embodiment one, and the details of the modules in the system are not described in detail in embodiment 2.
Example 3
The present embodiment also provides a computer device comprising a memory 1005 and a processor 1001, the memory 1005 storing a computer program, the processor 1001 implementing the steps of the method of any one of the above when executing the computer program.
It should be noted that the processor 1001 is configured to execute the steps in the above method embodiments according to the instructions in the program code. Or the processor 1001, when executing the computer programs, performs the functions of the modules/units in the systems/system embodiments described above.
In particular, in the present embodiment, the computer program may be divided into one or more modules/units, and one or more modules/units are stored in the memory 1005 and executed by the processor 1001 to accomplish the present application. One or more of the modules/units may be a series of computer program instruction segments capable of performing specific functions for describing the execution of the computer program in the terminal device.
The terminal device may be a computing device such as a desktop computer, a notebook computer, a palm computer, a cloud server, etc. Terminal devices can include, but are not limited to, a processor 1001, a memory 1005. Those skilled in the art will appreciate that the terminal device is not limited and may include more or less components than illustrated, or may combine certain components, or different components, e.g., the terminal device may also include an input-output device 1003, a network access device 1002, a bus 1006, etc.
The Processor 1001 may be a central processing unit (Central Processing Unit, CPU), but may also be other general purpose processors 1001, digital signal processors 1001 (DIGITAL SIGNAL Processor, DSP), application SPECIFIC INTEGRATED Circuit (ASIC), off-the-shelf Programmable gate array (Field-Programmable GATE ARRAY, FPGA) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. The general purpose processor 1001 may be a microprocessor or the processor may be any conventional processor or the like.
The memory 1005 may be an internal storage unit of the terminal device, such as a hard disk or a memory of the terminal device. The memory 1005 may also be an external storage device 1004 of the terminal device, such as a plug-in hard disk provided on the terminal device, a smart memory card (SMARTMEDIA CARD, SMC), a Secure Digital (SD) card, a flash memory card (FLASH CARD), or the like. Further, the memory 1005 may also include both an internal storage unit of the terminal device and an external storage device 1004. The memory 1005 is used to store computer programs and other programs and data required for the terminal device. The memory 1005 may also be used to temporarily store data that has been output or is to be output.
It will be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, systems and units may refer to the corresponding processes in the foregoing method embodiments, which are not repeated herein.
Example 4
The present embodiment provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of the method of any of the above.
The computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium would include an electrical connection having one or more wires, a portable computer diskette, and a hard disk. Random access Memory (Random Access Memory, RAM), read-Only Memory (ROM), erasable programmable Read-Only Memory (Erasable Programmable Read Only Memory, EPROM), registers, hard disk, optical fiber, portable compact disc Read-Only Memory (CD-ROM), an optical storage device, a magnetic storage device, or any other form of computer-readable storage medium suitable for use by a person or persons of skill in the art.
An exemplary storage medium is coupled to the processor such the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium may be integral to the processor. The processor and the storage medium may reside in an Application SPECIFIC INTEGRATED Circuit (ASIC). In embodiments of the present invention, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, or device.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (10)
1. An optimization method for digital design of an excavator is characterized by comprising the following steps:
acquiring characteristic data of the excavator, constructing a digital twin-based three-dimensional model of the excavator, and creating an excavator simulation working environment;
performing space fusion processing on the characteristic data by using a multi-layer sensor to obtain a characteristic data set;
Constructing a feature undirected graph by utilizing feature data in a feature data set, acquiring maximum mutual information coefficients among different nodes of the feature undirected graph, constructing an adjacency matrix of the feature undirected graph, inputting the adjacency matrix of the feature undirected graph into a graph neural network, and processing the adjacency matrix by combining a multi-head attention mechanism to obtain an importance score of a undirected graph node, wherein the feature data is used as the nodes of the undirected graph, and the association relationship among the nodes generates edges of the undirected graph;
Constructing an initial excavator relationship model based on the self-encoder network serving as an input layer of the deep neural network model, optimizing a network structure of the initial excavator relationship model by utilizing a particle swarm algorithm to obtain an initial excavator relationship model with an optimal structure, inputting feature data and importance scores into the initial excavator relationship model for training, and obtaining the excavator relationship model finally used for generating the excavator structure association coefficient after training;
Inputting the feature data to be processed into an excavator relation model to output an excavator structure association coefficient of the association relation between the feature data and the excavator structural part, and correcting the excavator three-dimensional model by using the excavator structure association coefficient;
operating an excavator three-dimensional model in an excavator simulated working environment, and returning various performance indexes and rewarding values of the excavator according to the operation condition of the excavator three-dimensional model;
Correcting the three-dimensional model of the excavator by utilizing various performance indexes and rewarding values of the excavator, inputting the corrected three-dimensional model of the excavator into the excavator relation model to output the association coefficient of the excavator structure, and repeatedly and iteratively correcting until the rewarding values are stable to obtain the optimal three-dimensional model of the excavator.
2. The optimization method for digital design of excavator according to claim 1, wherein the spatial fusion processing is performed on the feature data by using a multi-layer sensor to obtain a feature data set, specifically:
Setting class weights for the feature data of the excavator based on preset feature classes, extracting pearson correlation coefficients between the feature data, marking the feature data with the pearson correlation coefficients larger than a preset coefficient threshold value, obtaining a plurality of marked feature data, and respectively setting association weights for the marked feature data according to the pearson correlation coefficients of the marked feature data;
Based on the category weights and the associated weights, the multi-layer perceptron is utilized to perform space fusion processing on the feature data, so as to obtain a feature data set.
3. The optimization method for the digital design of the excavator according to claim 2, wherein a feature undirected graph is constructed by using feature data in a feature data set, an adjacency matrix of the feature undirected graph is constructed by acquiring maximum mutual information coefficients among different nodes of the feature undirected graph, the adjacency matrix of the feature undirected graph is input into a graph neural network and is processed by combining a multi-head attention mechanism, so as to obtain importance scores of nodes of the undirected graph, wherein the feature data is used as the nodes of the undirected graph, and the association relation among the nodes generates edges of the undirected graph, specifically:
Constructing an undirected graph according to the feature data set, taking each feature data in the feature data set as a node of the undirected graph, acquiring an association relation between each node according to the pearson correlation coefficient, and generating an edge of the undirected graph based on the association relation between each node to obtain a feature undirected graph associated with each feature data;
Obtaining the maximum mutual information coefficient between different nodes of the feature undirected graph, and constructing an adjacent matrix of the feature undirected graph according to the maximum mutual information coefficient;
Inputting an adjacency matrix of the feature undirected graph into a graph neural network, outputting neighbor node vectors of a plurality of neighbor nodes by the graph neural network, and acquiring multi-head attention coefficients of the plurality of neighbor nodes in the adjacency matrix based on a multi-head attention mechanism;
and multiplying the neighbor node vectors of the plurality of neighbor nodes with the corresponding multi-head attention coefficients respectively, and averaging to obtain the importance scores of the feature undirected graph nodes.
4. The optimization method of digital design of an excavator according to claim 1, wherein an initial excavator relationship model is built based on an input layer taking a self-encoder network as a deep neural network model, a network structure of the initial excavator relationship model is optimized by utilizing a particle swarm algorithm to obtain an initial excavator relationship model with an optimal structure, feature data and importance scores are input into the initial excavator relationship model for training, and the excavator relationship model finally used for generating the excavator structure association coefficient is obtained after training, specifically:
Acquiring a self-encoder network, and taking the self-encoder network as an input layer of a deep neural network model to obtain an initial excavator relationship model;
Optimizing a network structure of the initial excavator relationship model by utilizing a particle swarm algorithm to obtain the initial excavator relationship model with an optimal structure;
And inputting the characteristic data and the importance score into an initial excavator relationship model for training, and obtaining the excavator relationship model finally used for generating the excavator structure association coefficient after training.
5. The optimization method for digitized design of an excavator according to claim 4, wherein the network structure of the initial excavator relationship model is optimized by using a particle swarm algorithm, and the initial excavator relationship model with the optimal structure is obtained specifically:
initializing parameters of a particle swarm algorithm, particle position and speed information, taking the output precision of an initial excavator relation model as fitness, and acquiring an individual extremum and a global extremum of particles;
Performing iterative optimization based on a simulated annealing algorithm, and updating the particle position and speed by using a simulated annealing neighborhood search;
Stopping optimizing when a preset termination condition is met, acquiring optimal structure parameters of the encoder network according to the optimal particle positions, and constructing an initial excavator relationship model of the optimal structure based on the optimal structure parameters.
6. The method of optimizing digital design of an excavator of claim 2 wherein the predetermined characteristic categories include excavator work environment, power performance and work performance.
7. The method of optimizing digital design of an excavator according to claim 1, wherein the simulated working environment of the excavator comprises farmland, mountain, city and river.
8. An optimization system for digital design of an excavator, the system comprising:
the feature data acquisition module is used for acquiring feature data of the excavator, constructing a digital twin-based three-dimensional model of the excavator and creating an excavator simulation working environment;
The characteristic data set construction module is used for carrying out space fusion processing on the characteristic data by utilizing the multi-layer perceptron to obtain a characteristic data set;
The undirected graph construction module is used for constructing a characteristic undirected graph by utilizing characteristic data in a characteristic data set, acquiring the maximum mutual information coefficient between different nodes of the characteristic undirected graph, constructing an adjacent matrix of the characteristic undirected graph, inputting the adjacent matrix of the characteristic undirected graph into a graph neural network, and processing the adjacent matrix by combining a multi-head attention mechanism to obtain an importance score of a undirected graph node, wherein the characteristic data is used as the undirected graph node, and the association relationship between the nodes generates an undirected graph side;
the model training module is used for constructing an initial excavator relationship model based on the self-encoder network serving as an input layer of the deep neural network model, optimizing a network structure of the initial excavator relationship model by utilizing a particle swarm algorithm to obtain an initial excavator relationship model with an optimal structure, inputting feature data and importance scores into the initial excavator relationship model for training, and obtaining the excavator relationship model finally used for generating the excavator structure association coefficient after training;
The structure correction module is used for inputting the characteristic data to be processed into the excavator relation model so as to output an excavator structure association coefficient of the association relation between the characteristic data and the excavator structural part, and correcting the excavator three-dimensional model by utilizing the excavator structure association coefficient;
The simulation operation module is used for operating the three-dimensional model of the excavator in the simulated working environment of the excavator and returning various performance indexes and rewarding values of the excavator according to the operation condition of the three-dimensional model of the excavator;
the structure correction module is also used for correcting the three-dimensional model of the excavator by utilizing various performance indexes and rewarding values of the excavator, inputting the corrected three-dimensional model of the excavator into the excavator relation model to output the association coefficient of the excavator structure, and repeatedly and iteratively correcting until the rewarding values are stable so as to obtain the optimal three-dimensional model of the excavator.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 7 when the computer program is executed.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410897444.3A CN118862466B (en) | 2024-07-05 | 2024-07-05 | An optimization method for digital design of excavators |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410897444.3A CN118862466B (en) | 2024-07-05 | 2024-07-05 | An optimization method for digital design of excavators |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118862466A CN118862466A (en) | 2024-10-29 |
| CN118862466B true CN118862466B (en) | 2025-03-18 |
Family
ID=93157564
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410897444.3A Active CN118862466B (en) | 2024-07-05 | 2024-07-05 | An optimization method for digital design of excavators |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118862466B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105653811A (en) * | 2016-01-11 | 2016-06-08 | 重庆市送变电工程有限公司 | Rocked depth determining method for high voltage transformer substation deep back filled region rotary excavation drilling |
| CN116881328A (en) * | 2023-07-10 | 2023-10-13 | 电子科技大学 | Key knowledge graph key mode mining method for complex geological structure |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20230007422A (en) * | 2020-04-28 | 2023-01-12 | 스트롱 포스 티피 포트폴리오 2022, 엘엘씨 | Digital twin systems and methods for transportation systems |
| US20230055488A1 (en) * | 2021-08-14 | 2023-02-23 | Hcl Technologies Limited | Method and system for extracting and classifying manufacturing features from three-dimensional model of product |
| CN116484740A (en) * | 2023-04-28 | 2023-07-25 | 南京信息工程大学 | A Line Parameter Identification Method Based on Mining Spatial Topological Features of Power Grid |
-
2024
- 2024-07-05 CN CN202410897444.3A patent/CN118862466B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105653811A (en) * | 2016-01-11 | 2016-06-08 | 重庆市送变电工程有限公司 | Rocked depth determining method for high voltage transformer substation deep back filled region rotary excavation drilling |
| CN116881328A (en) * | 2023-07-10 | 2023-10-13 | 电子科技大学 | Key knowledge graph key mode mining method for complex geological structure |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118862466A (en) | 2024-10-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108645411B (en) | Robot path planning method and device based on particle swarm algorithm and terminal equipment | |
| CN110969649A (en) | Matching evaluation method, medium, terminal and device of laser point cloud and map | |
| CN111185909B (en) | Robot operation condition acquisition method and device, robot and storage medium | |
| CN114139637A (en) | Multi-agent information fusion method and device, electronic equipment and readable storage medium | |
| CN114255616A (en) | Unpowered ship track prediction method, device, equipment and storage medium | |
| CN111707279A (en) | Matching evaluation method, medium, terminal and device of laser point cloud and map | |
| CN112597612B (en) | Robot optimization method, device, terminal equipment and computer readable storage medium | |
| CN115659791A (en) | High-performance virtual simulation method and system driven by digital twin data model | |
| CN116340315A (en) | Data archiving method and AI system based on digital factory | |
| CN112966433B (en) | Instant compiling-based neurodynamics simulation method and device | |
| CN113239646A (en) | Wind power plant modeling method, medium and equipment based on equivalent roughness | |
| CN116467877A (en) | Floating wind turbine generator system platform dynamic response determination method and device and electronic equipment | |
| CN118862466B (en) | An optimization method for digital design of excavators | |
| CN114967465A (en) | Trajectory planning method, device, electronic device and storage medium | |
| CN114330043A (en) | Multi-degree-of-freedom multifunctional high-load hydraulic mechanical arm rigid-flexible coupling analysis method | |
| CN112766462A (en) | Data processing method, device and computer readable storage medium | |
| CN115049767B (en) | Animation editing method and device, computer equipment and storage medium | |
| CN115577573B (en) | Method, device, equipment and storage medium for predicting output current of synchronous generator | |
| WO2024108532A1 (en) | Method and apparatus for operating deformable object, and electronic device and storage medium | |
| Miatliuk | Conceptual Design of Mechatronic Systems | |
| CN115972216A (en) | Parallel robot forward motion solution method, control method, equipment and storage medium | |
| CN116050257A (en) | Unmanned aerial vehicle digital twin body modularized simulation method oriented to self-awareness enhancement | |
| CN119476049B (en) | Simulation model construction method, device, equipment and medium based on particle swarm optimization | |
| Liu et al. | Design of Kalman Filter Based on FPGA | |
| CN118795916B (en) | Solution method for bow-swing attitude controller of underwater robot |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |