TWI600988B - Machining expert system for CNC machine tool and controller processing parameters generation method - Google Patents

Description

Processing expert system for CNC machine tool and controller processing parameter generation method

The invention relates to a CNC machine tool, in particular to a machining expert system for a CNC machine tool and a controller processing parameter generating method.

Due to the different product characteristics, CNC machine tools usually use the performance indexes such as cutting speed, accuracy and surface quality as processing requirements. There are many factors that affect these performance indicators. The machining parameters used in the operation of the CNC controller (for example, maximum acceleration, jerk, corner speed, etc.) have a great influence on the machining results, so the CNC controller plays a role. Whether the processing result can meet the important role of the processing demand goal. The control parameters of the traditional CNC controller are usually selected from the factory standard settings, and the user rarely further considers the assembly, mechanical characteristics, and machining targets of the machine to dynamically adjust the controller processing parameters. With the evolution of society and technology, today's processed products have gradually developed toward a small number, diverse styles and exquisite workmanship. Customers have many customization requirements. The processing parameters of a single controller set by standard are usually not up to standard. Set processing demand targets. Although the user can adjust the controller processing parameters based on experience or using the information provided by the manufacturer, the user may need to repeat the test several times and the manufacturer's information is very small. In the end, it not only causes trouble and inconvenience to the user, but is more likely to affect the user. Processing efficiency and quality. It can be seen that how to provide better controller processing parameters to achieve the desired processing demand is one of the goals of the relevant industry.

In view of the above-mentioned deficiencies, an object of the present invention is to provide a processing expert system for a CNC machine tool and a controller processing parameter generating method, which are easy for a user to operate, can obtain more precise processing parameters, and contribute to improvement. CNC controller performance, processing efficiency and processing quality.

To achieve the above object, the present invention provides a processing expert system for a CNC machine tool that operates based on a plurality of controller processing parameters, the processing expert system including a display unit, an input unit, and a processing The display unit is configured to display a user interface having three processing demand indicators, wherein the three processing demand indicators include a speed indicator, a precision indicator, and a surface quality indicator; the input unit is configured to receive the user interface. Inputting information corresponding to the three processing demand indicators; and the processing unit is coupled to the display unit and the input unit, and the processing unit calculates the result of an objective function according to the input information and through an optimization algorithm, The controller processing parameters corresponding to the input information are derived, wherein the objective function is constructed based on a machine learning algorithm.

In addition, the present invention further provides a controller processing parameter generating method for a CNC machine tool, the CNC machine tool operating based on a plurality of processing parameters, and the controller processing parameter generating method comprises the following steps: displaying three processing requirements a user interface of the indicator, wherein the three processing demand indicators include a speed indicator, a precision indicator, and a surface quality indicator; receiving input information corresponding to the three processing demand indicators on the user interface; and receiving the information according to the input information And calculating an objective function result by an optimization algorithm to obtain the controller processing parameters corresponding to the input information, wherein the objective function is constructed based on a machine learning algorithm.

Thereby, the user can easily input the preference demand for the processing demand indicators on the user interface, and the processing expert system can find the result that best meets the preference demand among the plurality of sets of controller processing parameters. .

Referring to FIG. 1 , a processing expert system 1 of a preferred embodiment of the present invention includes a display unit 10 , an input unit 30 , a processing unit 50 coupled to the display unit 10 and the input unit 30 , and a storage unit The unit 70 is coupled to the processing unit 50. The processing expert system 1 can be built into a CNC machine tool (not shown) or operated as a stand-alone device.

The display unit 10 can be a video display unit of the type such as an LCD or an LED. The display unit 10 is used to display a user interface for user operation. The user interface is described in detail in the following embodiments.

The input unit 30 can be a touch panel, a button, a keyboard or a mouse. The input unit 30 is configured to receive input information obtained by a user for an input operation on the user interface.

The processing unit 50 can be a central processing unit, a chip or other programmable general purpose or special purpose microcontroller, or the like having a function of signal processing, parameter calculation, etc., or a combination of the above components. Control all functional functions of the processing expert system 1.

The storage unit 70 can be a storage medium of a type such as a random access memory, a flash memory, a read-only memory, an erasable programmable read-only memory, or an electronic erasable rewritable read-only memory. The unit 70 is configured to store various control parameters (for example, maximum acceleration, jerk, angular velocity, etc.) of the CNC controller (not shown), processing parameters, straight lines, squares, and circular machining paths. The processing path (ie, coordinate position information) obtained by the optical tool (not shown) of the CNC machine tool, the input information, the plurality of sets of actual working samples, and the like. The actual operating samples include a plurality of existing control parameters and corresponding plurality of processing characteristics (eg, a processing time, a contour error, and a tracking error, etc.), and the CNC machine tool of the CNC machine tool is based on each The controller processing parameters in the existing control parameters are actually operated to obtain the result with the corresponding processing characteristics.

The foregoing description is related to the structure and component design of the processing expert system 1. The controller processing parameter generation method of the processing expert system 1 will be further described below with reference to FIG.

Referring to FIG. 2, in step S21, the processing unit 50 displays the user interface having three processing demand indicators through the display unit 10. In this embodiment, the three processing demand indicators are a speed index, a precision index, and a surface quality index related to the processing performance index. The three processing demand indicators are in conflict with each other, and the actual processing results cannot simultaneously achieve the highest preference demand for the three processing demand indicators. For example, Figure 3 is a user interface 80 of a preferred embodiment having a triangular region 81 centered at the velocity index, the accuracy index, and the surface quality index. The user can select a target position P for the three processing demand indicators in the triangle area 81, and the target position P represents the user's preferred demand for each processing demand indicator.

It should be noted that the definition of the triangular area 81 is due to the corresponding relationship between the three processing demand indicators. However, the user interface 80 may have various changes, which may be an interface that can adjust the percentage of the preferred demand. , for the option to directly enter the interface of the preferred demand, and so on.

When the user selects the target position P in the triangular area 81, the processing unit 50 can receive an input information corresponding to the three processing demand indicators on the user interface 80 through the input unit 30 (step S23). The input information is the target position P in the triangular area 81.

In step S25, the processing unit 50 calculates a result of an objective function according to the input information and through an optimization algorithm to obtain the controller processing parameters corresponding to the input information. In this embodiment, the objective function is a corresponding link between the controller processing parameters and the three processing demand indicators. That is to say, the three processing demand indicators are input to the objective function to obtain corresponding controller processing parameters. The objective function of the optimization algorithm is defined as equation (1): …(1) For the objective function, r is the distance between the target position P and the position of the three vertices ( , , Reciprocal (ie , i is 1, 2, 3), Is a processing time function (average of processing time for straight lines, squares and circular machining tracks), Is a contour error function (the average of the contour errors of the square and circular machining trajectories), Is a tracking error function (average of tracking error between straight lines, squares and circular machining trajectories).

The processing unit 50 calculates a reciprocal of the distance between the target position P and the three vertex positions ( , , ), as the weight value of the three processing demand indicators (ie, the preferred demand). The optimization algorithm of the embodiment adopts a particle swarm optimization algorithm, and the processing unit 50 calculates the minimum value of the objective function through the particle swarm optimization algorithm according to the weight values (and The results of limiting the range of the plurality of machining parameters of the CNC controller are derived to derive the controller processing parameters (eg, the machining parameters of the CNC controller). For example, the processing unit 50 defines a solution space according to the objective function, initializes the position (ie, solution or result) and speed of the plurality of particles in the solution space, and evaluates the adaptation values of the particles to determine whether to update the The optimal position of the particles and the optimal position of the group of particles are moved to move the particles, and the iterative calculation seeks to find the optimal positions of the particles as the control parameters.

It should be noted that the particle swarm optimization algorithm is an algorithm developed by using group wisdom, which can obtain an optimal solution in the optimization problem and effectively save cost and time. However, there are actually many optimization algorithms. In other embodiments, the optimization algorithm can be changed to simulate degradation, gene, and the like according to requirements, and the objective function and the weight values are used. Appropriate adjustments are required to comply with the corresponding algorithm.

The processing unit 50 can display the controller processing parameters in a message display field 83 of the user interface 80 through the display unit 10, or through a communication unit (not shown, for example, using USB, WiFi, etc. Or wireless communication technology) transmitting and setting the controller processing parameters to the CNC controller of the CNC machine tool.

It is worth noting that the objective function of the present invention is constructed based on a machine learning algorithm. The machine learning algorithm of this embodiment adopts a reverse transfer type neural network algorithm. The inverse transfer type neural network algorithm is an algorithm developed by imitating a biological neural network architecture, based on a supervised learning, when the processing unit 50 (from the storage unit 70 or the input unit 30) obtains the CNC When the controller actually operates the actual working samples, the processing unit 50 may obtain the processing characteristics (output layer neurons) and the existing control parameters (input layer neurons) according to the actual operating samples. All of the connection weights (including hidden layer neurons), the processing unit 50 can reconstruct the objective function based on the connection rights. For the foregoing construction manner, refer to FIG. 4, the processing unit 50 initializes the connection weights, and obtains a plurality of estimation characteristics based on the existing control parameters and the connection weights (for example, an estimated processing time, An estimated contour error and an estimated tracking error, etc.). The processing unit 50 calculates the prediction errors between the processing characteristics and the estimated characteristics, and adjusts the connection weights by minimizing the estimation errors of the actual operating samples, and finally the processing unit 50 can obtain The objective function.

It should be noted that there are actually many kinds of machine learning algorithms, such as support vector machines, decision trees, linear regression and other algorithms can be applied to construct the objective function or form a target model, and the actual operation samples The content needs to be adjusted according to different algorithms.

Since the different positions in the triangular area 81 correspond to the different preference requirements of the three processing demand indicators, the user can select a combination of the preferred demand quantities, even if the user can easily infer through the rule of thumb or the manufacturer's data. Accurate corresponding controller processing parameters (ie, the three processing demand indicators are too different from the processing characteristics generated based on the processing parameters of the controllers). In contrast, the embodiment of the present invention combines the machine learning algorithm and the optimization algorithm, which can calculate different controller processing parameters and the three processing demand indicators in a specific numerical range based on some known actual operating samples. The corresponding link (ie, the objective function) is used to quickly derive the best result of the objective function as the corresponding controller processing parameter. Thereby, the user can easily select the target position P in the triangular area 81 according to the desired processing demand index, and the processing unit 50 can quickly obtain accurate controller processing parameters (ie, the three processing demand indicators are close to Based on the processing characteristics generated by the processing parameters of the controllers, the performance, processing efficiency and processing quality of the CNC controller can be further improved.

The above is only the description of the embodiments of the present invention, and is not intended to limit the scope of the present invention. Any simple structural retouching or variation that does not depart from the spirit of the present invention is still within the scope of the patent application scope of the present invention.

1‧‧‧Processing expert system

10‧‧‧Display unit

30‧‧‧Input unit

50‧‧‧Processing unit

70‧‧‧ storage unit

S21～S25‧‧‧Steps

80‧‧‧User interface

81‧‧‧Triangle area

83‧‧‧Message display field

P‧‧‧target location

‧‧‧distance

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a block diagram of the components of a processing expert system in accordance with a preferred embodiment of the present invention. 2 is a flow chart showing a method of generating a processing parameter of a controller according to a preferred embodiment of the present invention. Figure 3 is a schematic illustration of a user interface of a preferred embodiment of the present invention. Figure 4 is a schematic diagram of an inverted-transition neural network learning architecture in accordance with a preferred embodiment of the present invention.

80‧‧‧User interface

81‧‧‧Triangle area

83‧‧‧Message display field

P‧‧‧target location

d ₁ ~ d ₃ ‧‧‧ distance

Claims (12)

A machining expert system for a CNC machine tool, the CNC machine tool operating based on a plurality of controller processing parameters, the processing expert system comprising: a display unit for displaying a user having three processing demand indicators The interface, wherein the three processing demand indicators include a speed indicator, a precision indicator, and a surface quality indicator; an input unit for receiving input information corresponding to the three processing demand indicators on the user interface; and a processing unit The processing unit is coupled to the display unit and the input unit, and the processing unit obtains a plurality of sets of actual operation samples, and the processing unit constructs an objective function according to the actual operation samples and through a machine learning algorithm, and the processing unit is configured according to the input information. And calculating, by an optimization algorithm, the result of the objective function to obtain the controller processing parameters corresponding to the input information, wherein each of the actual working samples includes a plurality of existing control parameters and correspondingly Processing characteristics, and the CNC machine tool is based on the existing control parameters to obtain the processing characteristics As a result, the plurality of processing includes a processing time characteristic, a contour error and a tracking error. The processing expert system for a CNC machine tool according to claim 1, wherein the optimization algorithm is a particle group optimization algorithm, and the processing unit obtains the weight of the three processing demand indicators according to the input information. The processing unit obtains the controller processing parameters through the particle swarm optimization algorithm according to the weight values. The processing expert system for a CNC machine tool according to claim 2, wherein the user mask has a triangular region with the velocity index, the accuracy index, and the surface quality index as a vertex, and the input information is in the triangle region. a target position, and the processing unit obtains the weight value of the three processing demand indicators according to the distance between the target position and the three vertex positions. The processing expert system for a CNC machine tool according to claim 3, wherein the weight value of the three processing demand indicators is a reciprocal of the distance between the target position and the three vertex positions. The processing expert system for a CNC machine tool according to claim 1, wherein the machine learning algorithm is a reverse-transfer-like neural network algorithm, and the processing unit passes the inverted operational neural network according to the actual operational samples. The road algorithm constructs the objective function. The processing expert system for a CNC machine tool according to claim 5, wherein the processing unit obtains a plurality of estimated characteristics based on the existing control parameters, and the processing unit calculates the processing characteristics and the estimated characteristics. The prediction error between the processing units and the estimation errors of the actual operational samples are minimized to derive the objective function. A controller processing parameter generating method for a CNC machine tool, the CNC machine tool operating based on a plurality of controller processing parameters, and the controller processing parameter generating method comprises the following steps: obtaining a plurality of sets of actual working samples, each of which The actual working sample includes a plurality of existing control parameters and corresponding plurality of processing characteristics, and the CNC machine tool obtains the result of having the processing characteristics based on the existing control parameters, and the processing characteristics include a processing Time, a contour error, and a tracking error; constructing an objective function by using a machine learning algorithm according to the actual working samples; displaying a user interface having three processing demand indicators, wherein the three processing demand indicators include a speed indicator a precision indicator and a surface quality indicator; receiving input information corresponding to the three processing demand indicators on the user interface; and calculating a result of the objective function according to the input information and through an optimization algorithm The controller processing parameters corresponding to the input information are output. The controller processing parameter generation method for a CNC machine tool according to claim 7, wherein the optimization algorithm is a particle group optimization algorithm, and the calculation is performed according to the input information and the optimization algorithm The step of the result of the objective function comprises: obtaining weight values of the three processing demand indicators according to the input information; and obtaining the controller processing parameters by the particle group optimization algorithm according to the weight values. The method for generating a controller processing parameter for a CNC machine tool according to claim 8, wherein the user mask has a triangular region with the velocity index, the accuracy index, and the surface quality index as a vertex, and the input information is A target position in the triangular area, and determining the weight value of the three processing demand indicators according to the input information comprises: determining a weight value of the three processing demand indicators according to the distance between the target position and the three vertex positions. The controller processing parameter generation method for a CNC machine tool according to claim 9, wherein the weight value of the three processing demand indicators is a reciprocal of a distance between the target position and the three vertex positions. The method for generating a controller processing parameter for a CNC machine tool according to claim 7, wherein the machine learning algorithm is a reverse-transfer-like neural network algorithm, and according to the actual working samples and constructed by the machine learning algorithm The step of the objective function comprises: constructing the objective function according to the actual operational samples and by the inverse transfer-like neural network algorithm. The controller processing parameter generating method for a CNC machine tool according to claim 11, wherein the step of constructing the objective function according to the actual operating samples and transmitting the inverse neural network algorithm comprises: Deriving a plurality of estimated characteristics based on the existing control parameters; calculating the processing characteristics and prediction errors between the estimated characteristics; and minimizing the estimated errors of the actual operational samples to obtain the target function.