A method, system, equipment and medium for accurate inversion of ultrasonic grinding processing parameters

Abstract

The present application discloses a method, system, equipment and medium for accurate inversion of ultrasonic grinding processing parameters, which extracts an objective function from a preset prediction model; the prediction model is trained based on the ultrasonic grinding processing parameters of a basic workpiece and the surface quality parameters of the basic workpiece, and is used to characterize the corresponding relationship between the ultrasonic grinding processing parameters and the surface quality parameters; based on the ultrasonic grinding processing parameters of a randomly generated target workpiece, a Pareto solution set is calculated using a genetic algorithm and an objective function; the basic workpiece and the target workpiece are workpieces of the same type; the target ultrasonic grinding processing parameters of the target workpiece are inverted based on the Pareto solution set, the ultrasonic grinding processing parameters are effectively optimized, the processed surface has better quality and performance, complex physical modeling is avoided, and the efficiency and accuracy of calculating the processing parameters are improved.

Description

Technical Field

[0001] The present application relates to the field of grinding processing, and in particular to a method, system, equipment and medium for accurate inversion of ultrasonic grinding processing parameters. Background Art

[0002] In the field of mechanical manufacturing, the surface quality of key components affects the operating stability and service life of the entire mechanical equipment. In particular, finishing processes such as ultrasonic grinding are often used as the final processing process for key components, which determines the final surface quality.

[0003] In the prior art, the optimization of ultrasonic grinding is mainly achieved by establishing a physical field forward model of the ultrasonic grinding part surface with many simplified assumptions. The model requires mathematical modeling of the physical phenomena in the ultrasonic grinding process, including but not limited to fluid dynamics, heat conduction, material removal mechanism, etc., so as to form a theoretical framework that can predict the processing results. However, a single simplified assumption cannot take into account the complex environment in actual applications, resulting in low accuracy of the forward modeling results. Summary of the invention

[0004] The following is a summary of the subject matter described in detail herein. This summary is not intended to limit the scope of the claims.

[0005] The main purpose of the embodiments of the present disclosure is to propose a method, system, equipment and medium for accurate inversion of ultrasonic grinding process parameters, which can improve the efficiency and accuracy of inverting ultrasonic grinding process parameters.

[0006] A first aspect of an embodiment of the present application provides a method for accurate inversion of ultrasonic grinding process parameters, the method comprising:

[0007] Extracting an objective function from a preset prediction model; the prediction model is trained based on ultrasonic grinding processing parameters of a basic workpiece and surface quality parameters of the basic workpiece, and is used to characterize the corresponding relationship between the ultrasonic grinding processing parameters and the surface quality parameters;

[0008] According to the ultrasonic grinding processing parameters of the randomly generated target workpiece, a Pareto solution set is calculated using a genetic algorithm and the objective function; the basic workpiece and the target workpiece are workpieces of the same type;

[0009] The target ultrasonic grinding processing parameters of the target workpiece are inverted according to the Pareto solution set.

[0010] The embodiment of the present application provides a method for accurate inversion of ultrasonic grinding processing parameters, which extracts an objective function for characterizing the correlation between ultrasonic grinding processing parameters and surface quality parameters of a material from a preset prediction model; a Pareto solution set is calculated using a genetic algorithm and the objective function according to the ultrasonic grinding processing parameters of a randomly generated target workpiece; the base workpiece and the target workpiece are workpieces of the same type; the target ultrasonic grinding processing parameters of the target workpiece are inverted according to the Pareto solution set, and the ultrasonic grinding processing parameters are effectively optimized, so that the processed surface has better quality and performance, and the surface roughness parameters are predicted based on the prediction model based on the actual processing data, thereby avoiding complex physical modeling and improving the efficiency and accuracy of calculating the processing parameters.

[0011] In some embodiments of the present application, before extracting the target function from the preset prediction model, the method further includes:

[0012] Acquire a model training set, wherein the model training set includes ultrasonic grinding processing parameters and surface quality parameters of a basic workpiece;

[0013] Training a plurality of preset basic machine learning models based on the model training set;

[0014] Obtaining model performance parameters of the multiple basic machine learning models;

[0015] A basic machine learning model with the best model performance parameters is selected from the multiple basic machine learning models as the prediction model.

[0016] In some embodiments of the present application, the method of calculating the Pareto solution set using a genetic algorithm and the objective function according to the randomly generated ultrasonic grinding processing parameters of the target workpiece includes:

[0017] Determine the surface quality parameters of the corresponding target workpiece according to the target function and the ultrasonic grinding processing parameters of the target workpiece;

[0018] The Pareto solution set corresponding to the surface quality parameters of the target workpiece is calculated by a non-dominated genetic algorithm.

[0019] In some embodiments of the present application, the calculating of the Pareto solution set corresponding to the surface quality parameter of the target workpiece by a non-dominated genetic algorithm includes:

[0020] generating genetic parameters according to the surface quality parameters of the target workpiece;

[0021] Perform genetic iteration of the non-dominated genetic algorithm according to genetic parameters;

[0022] When the i-th generation non-dominated solution is obtained, the number of repetitions of the i-th generation non-dominated solution and the non-dominated solution in the previous i-1 generation non-dominated solution is determined; wherein i is an integer greater than 1;

[0023] When the number of repetitions of the non-dominated solution exceeds a preset number, a Pareto solution set is obtained according to the i-th generation non-dominated solution.

[0024] In some embodiments of the present application, the surface quality parameters include the arithmetic mean height of the surface, the solid volume of the peaks, and the void volume of the valleys.

[0025] In some embodiments of the present application, inverting the target ultrasonic grinding processing parameters of the target workpiece according to the Pareto solution set includes:

[0026] Using entropy weight method and superior-inferior solution distance method to confirm the optimal Pareto sub-solution in the Pareto solution set;

[0027] The corresponding target ultrasonic grinding processing parameters are obtained according to the optimal Pareto sub-solution.

[0028] A second aspect of an embodiment of the present invention provides a system for accurate inversion of ultrasonic grinding process parameters, the system comprising:

[0029] An acquisition module, used to extract the objective function from a preset prediction model; the prediction model is trained based on the ultrasonic grinding parameters of the basic workpiece and the surface quality parameters of the basic workpiece, and is used to characterize the corresponding relationship between the ultrasonic grinding parameters and the surface quality parameters;

[0030] A calculation module, used for calculating a Pareto solution set using a genetic algorithm and the objective function according to ultrasonic grinding processing parameters of a randomly generated target workpiece; the base workpiece and the target workpiece are made of the same material;

[0031] An inversion module is used to invert the target ultrasonic grinding processing parameters of the target workpiece according to the Pareto solution set.

[0032] A third aspect of an embodiment of the present invention provides an electronic device, comprising: at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor so that the at least one control processor can execute the above-mentioned method for precise inversion of ultrasonic grinding processing parameters.

[0033] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the above-mentioned method for accurate inversion of ultrasonic grinding processing parameters.

[0034] A fifth aspect of an embodiment of the present invention provides a computer program product. When instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the above-mentioned method for accurate inversion of ultrasonic grinding process parameters.

[0035] It can be understood that the beneficial effects of the second to fifth aspects compared with the related art are the same as the beneficial effects of the first aspect compared with the related art. Please refer to the relevant description in the first aspect, and no further details will be given here. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0037] Picture 1 It is a flow chart of a method for accurate inversion of ultrasonic grinding processing parameters provided in an embodiment of the present application;

[0038] Picture 2 In one embodiment provided in the present application Sa Schematic diagram of the Pareto solution set from 0.15 to 0.2;

[0039] Picture 3 In one embodiment provided in the present application Sa Schematic diagram of the Pareto solution set of 0.3~0.4;

[0040] Picture 4 In one embodiment provided in the present application Sa Schematic diagram of the Pareto solution set of 0.7~0.8;

[0041] Picture 5 It is a structural schematic diagram of a precise inversion system for ultrasonic grinding processing parameters provided in an embodiment of the present application;

[0042] Picture 6 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0043] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be understood as limiting the present application.

[0044] In the description of this application, if there is a description of first, second, etc., it is only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0045] In the description of the present application, it should be understood that the descriptions involving orientation, such as the orientation or positional relationship indicated as up, down, etc., are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0046] In the description of this application, it should be noted that, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meaning of the above terms in this application based on the specific content of the technical solution.

[0047] In the prior art, the optimization of ultrasonic grinding is mainly achieved by establishing a physical field forward model of the ultrasonic grinding part surface with many simplified assumptions. The model requires mathematical modeling of the physical phenomena in the ultrasonic grinding process, including but not limited to fluid dynamics, heat conduction, material removal mechanism, etc., so as to form a theoretical framework that can predict the processing results. However, a single simplified assumption cannot take into account the complex environment in actual applications, resulting in low accuracy of the forward modeling results.

[0048] For this purpose, refer to Picture 1 The embodiment of the present application provides a method for accurate inversion of ultrasonic grinding processing parameters. The method is applied to a controller. The controller can be a server, an electronic device, or a mobile terminal, etc., which is not specifically limited here. The method includes the following steps S101 to S103.

[0049] Step S101, extracting an objective function from a preset prediction model; the prediction model is trained based on ultrasonic grinding processing parameters of a basic workpiece and surface quality parameters of the basic workpiece, and is used to characterize the corresponding relationship between ultrasonic grinding processing parameters and surface quality parameters.

[0050] In this step, the prediction model is trained based on the ultrasonic grinding processing parameters of the basic workpiece and the surface quality parameters of the basic workpiece, wherein the ultrasonic grinding processing parameters include but are not limited to: amplitude, rotation speed and cutting depth; the surface quality parameters include but are not limited to: arithmetic mean height ( Sa ), the solid volume of the peak ( Vmp ) and the void volume of the valley ( Vvv ).

[0051] In the training stage, the ultrasonic grinding processing parameters in the model training set are used as input, and the surface quality parameters are used as output. The model is trained to predict the surface quality parameters when the processing parameters are given, so as to characterize the correlation between the ultrasonic grinding processing parameters and the surface quality parameters of the material. When the prediction model training is completed, the predicted surface quality parameters can be output according to the input ultrasonic grinding processing parameters, which will be used as the objective function for evaluating different processing parameter combinations in the subsequent optimization process, thereby maintaining or improving the surface quality (such as reducing Sa value), while improving the load-bearing and lubrication properties of the surface (such as increasing Vmp and Vvv value).

[0052] The following explains the specific training steps of the prediction model:

[0053] A model training set is obtained, wherein the model training set includes ultrasonic grinding processing parameters and surface quality parameters of a basic workpiece.

[0054] Train multiple preset basic machine learning models based on the model training set.

[0055] Get model performance parameters for multiple underlying machine learning models.

[0056] A basic machine learning model with the best model performance parameters is selected from multiple basic machine learning models as the prediction model.

[0057] In this embodiment, it is first necessary to obtain the ultrasonic grinding processing parameters and surface quality parameters of the basic workpiece, that is, the workpiece morphology data under actual processing conditions and its corresponding ultrasonic grinding processing parameters (such as amplitude, rotation speed, and cutting depth), and calculate the corresponding three-dimensional roughness parameters ( Sa , Vmp , Vvv ).

[0058] The three-dimensional roughness parameters are given below: Sa , Vmp , Vvv The calculation formula for .

[0059] For defined areas K The height matrix in z ( x , y ), 3D height parameters Sa as follows:

[0060] (1)

[0061] 3D volume parameters Vmp and Vvv The expression is as follows:

[0062] (2)

[0063] (3)

[0064] In the formula, A , B The default values ​​are 10% and 80% respectively. Q Represents the proportional constant, used to adjust the scale, generally 100%, Sdc ( a 1) Indicates the support area ratio curve a The value of 1 point, a 1 represents any point on the support area ratio curve.

[0065] Secondly, the above ultrasonic grinding processing parameters and their corresponding surface quality parameters are combined to form a model training set, and then a variety of machine learning models are selected, such as BP neural network, random forest and support vector machine, etc., and then the initial prediction model corresponding to each machine learning model is constructed according to the model training set based on the multiple machine learning models.

[0066] Then use the preset performance evaluation indicators (such as mean absolute error M AE , Root Mean Square Error R MSE and prediction accuracy P A ) is used to measure the performance of each initial prediction model, that is, to calculate the performance parameters of each initial prediction model, compare the performance parameters of each initial prediction model, and finally determine an optimal prediction model for the subsequent optimization process.

[0067] In one implementation of this embodiment, a machine learning model is used to establish ultrasonic grinding parameters (amplitude, speed, cutting depth) and optimization targets ( Sa , Vmp , Vvv ) and then the mean absolute error ( M AE ), Root Mean Square Error ( R MSE ) and prediction accuracy ( P A ) three parameters are used to comprehensively evaluate the performance of the constructed initial prediction model. When constructing the initial prediction model, all model training sets are randomly shuffled, and 70% of the data are extracted from them for training. According to the performance parameter comparison of the three initial prediction models in Table 2, it is found that the BP neural network has the best training effect. Therefore, the initial prediction model constructed by the BP neural network is used as the objective function in the subsequent optimization process.

[0068] (4)

[0069] (5)

[0070] (6)

[0071] In the formula, i is the measured value, y is the predicted value, u is the number of samples.

[0072] Table 1

[0073]

[0074] This makes it easier to accurately predict the relationship between ultrasonic grinding processing parameters and surface quality parameters according to the optimal prediction model in the subsequent optimization process, and find the optimal combination of processing parameters.

[0075] Step S102: Calculate the Pareto solution set using a genetic algorithm and an objective function according to the ultrasonic grinding processing parameters of the randomly generated target workpiece; the basic workpiece and the target workpiece are workpieces of the same type.

[0076] This embodiment provides a training process in which multiple iterations of a genetic algorithm are executed. The ultrasonic grinding processing parameters of the target workpiece are randomly determined values ​​based on actual processing conditions and engineering requirements. First, the surface quality parameters of the target workpiece are determined based on the objective function and the ultrasonic grinding processing parameters of the target workpiece, and then the Pareto solution set corresponding to the surface quality parameters of the target workpiece is calculated through a non-dominated genetic algorithm, wherein the basic workpiece and the target workpiece are workpieces of the same type.

[0077] In this embodiment, the number of iterations (T), the population size (P), the optimization target (surface quality parameters of the target workpiece) are first set, sav_int is initialized to an empty set for storing the non-dominated solution set of each generation, and then the population is initialized, including population initialization, non-dominated solution construction, and crowding calculation.

[0078] For each iteration i=1 to T, a new population is generated through elite selection, crossover and mutation, and then the new population is merged with the current population and the non-dominated sorting is performed again. After each iteration, the new set of non-dominated solutions is stored in the sav_int list. When the i-th generation non-dominated solution is obtained, the number of repetitions of the i-th generation non-dominated solution and the non-dominated solution in the previous i-1 generation non-dominated solution is determined, where i is an integer greater than 1.

[0079] By taking the number of repetitions of each non-dominated solution in the non-dominated solution as the iteration judgment condition, it is judged whether the iteration condition is met and the Pareto solution set is obtained. For example, when all non-dominated solutions appear repeatedly for a specified number of times k in multiple generations, the iteration is stopped. Specifically, if a solution has also appeared in the non-dominated solution sets of previous generations, it is considered that the solution set has been stable and the population may have converged. When the number of repetitions of the non-dominated solution meets the preset condition, that is, the changes of each non-dominated solution are no longer obvious, the corresponding Pareto solution set is obtained according to the non-dominated solution of the N+1th generation. If the preset condition is not met, the iteration is continued until the maximum number of iterations T is reached, so that the algorithm avoids repeated searches in the existing Pareto solution set, thereby improving the iteration efficiency.

[0080] In some embodiments, in step S102, according to the randomly generated ultrasonic grinding processing parameters of the target workpiece, a Pareto solution set is calculated using a genetic algorithm and an objective function, including the following steps:

[0081] Determine the surface quality parameters of the target workpiece according to the target function and the ultrasonic grinding processing parameters of the target workpiece;

[0082] The Pareto solution set corresponding to the surface quality parameters of the target workpiece is calculated using a non-dominated genetic algorithm.

[0083] In this step, the ultrasonic grinding processing parameters of the target workpiece are randomly generated according to the actual processing conditions and engineering requirements. The surface quality parameters of the target workpiece are determined according to the randomly generated ultrasonic grinding processing parameters of the target workpiece, and then the Pareto solution set corresponding to the surface quality parameters of the target workpiece is calculated by the non-dominated genetic algorithm. Then, the processing parameter combination that best meets the engineering requirements can be selected from the Pareto solution set to achieve different engineering requirements.

[0084] In this embodiment, the ultrasonic grinding processing parameters of the target workpiece are determined according to the actual processing conditions and engineering requirements, and the actual processing conditions are associated with specific processing equipment. Taking ultrasonic vibration assisted grinding as an example, its processing parameters are mainly: J A (amplitude), J D (cut depth) and J S (rotation speed), engineering requirements are mainly related to three-dimensional roughness parameters Sa Closely related, considering the above factors, the constraints of the ultrasonic grinding processing parameters of the target workpiece are as follows:

[0085] (7)

[0086] (8)

[0087] (9)

[0088] (10)

[0089] (11)

[0090] (12)

[0091] (13)

[0092] In the formula, Δ JA , Δ JD With Δ JS Derived from the actual accuracy of the processing parameters, Sa The value range of depends on the engineering requirements of the actual processing conditions, that is, the ultrasonic grinding processing parameters of the target workpiece randomly determined by the actual processing conditions and the engineering requirements are determined according to the objective function to determine the surface quality parameters of the corresponding target workpiece.

[0093] Among them, the optimization problem of calculating the Pareto solution set corresponding to the surface quality parameters of the target workpiece by the non-dominated genetic algorithm is summarized as follows:

[0094] (14)

[0095] The optimization problem in equation (14) is constrained by inequalities (7)-(13). Its optimization purpose is to ensure and optimize the surface quality and wear resistance while improving its load-bearing and lubrication performance as much as possible.

[0096] In one implementation of this embodiment, the number of iterations, the population size, the optimization target and the number of non-dominated solution repetitions k are first determined, and then the population is initialized, including population initialization, non-dominated solution construction and congestion calculation.

[0097] In each iteration, a new set of non-dominated solutions is calculated and stored in sav_int for subsequent comparison. Then, a new population is generated through elite selection, crossover and mutation operations, the new population is merged with the current population, and the non-dominated sorting is performed again to ensure that the optimal solution is retained in the population.

[0098] Then compare the current non-dominated solution set with the non-dominated solution sets of the previous generations to check the degree of repetition of the solution set. If a solution continues to appear in multiple generations of iterations, it means that the population may have converged. When all non-dominated solutions appear repeatedly for a specified number of times k in multiple generations, the iteration is stopped. Otherwise, the iteration is continued until the maximum number of iterations is reached or the convergence condition is met. Then, based on the constructed objective function, the Pareto solution search is performed according to the actual processing conditions listed in Table 1 and the following equations (15)-(18). The search results are shown in the following table. Picture 2 , Picture 3 as well as Picture 4 As shown, Picture 2 yes Sa The Pareto solution set is 0.15~0.2, Figure 3 is Sa The Pareto solution set is 0.3~0.4, Figure 4 is Sa It is a Pareto solution set of 0.7~0.8.

[0099] Table 2

[0100]

[0101] (15)

[0102] (16)

[0103] (17)

[0104] (18)

[0105] Among them, each solution in the Pareto solution set can be used to invert and obtain the corresponding ultrasonic grinding processing parameter value (amplitude, speed, cutting depth), so as to select the most suitable processing parameter combination from the Pareto solution set.

[0106] In some embodiments, determining the surface quality parameters of the target workpiece according to the target function and the ultrasonic grinding processing parameters of the target workpiece in step S301 includes the following steps:

[0107] Generate genetic parameters according to surface quality parameters of target workpiece;

[0108] Perform genetic iteration of non-dominated genetic algorithm according to genetic parameters;

[0109] When the i-th generation non-dominated solution is obtained, the number of repetitions of the i-th generation non-dominated solution and the non-dominated solution in the previous i-1 generation non-dominated solution is determined; wherein i is an integer greater than 1;

[0110] When the number of repetitions of non-dominated solutions exceeds the preset number, the Pareto solution set is obtained according to the i-th generation non-dominated solutions.

[0111] In this embodiment, based on an improved non-dominated genetic algorithm, a prediction model is used as an objective function to obtain a Pareto solution set.

[0112] The process of the improved non-dominated genetic algorithm is as follows.

[0113] (1) Input the number of iterations, population, optimization target, number of repetitions k, and initialize sav_int to an empty set.

[0114] (2) Multi-objective genetic algorithm initialization, including population initialization, non-dominated solution construction, and crowding calculation.

[0115] (3) Generate and store the initial non-dominated solution. In each iteration, calculate a new set of non-dominated solutions and store it in sav_int for subsequent comparison.

[0116] (4) Elite selection, crossover and mutation in the multi-objective genetic algorithm generate a new population. The new population is then merged with the current population and the non-dominated sorting is performed again to ensure that the optimal solution is retained in the population.

[0117] (5) Convergence check: By comparing the current non-dominated solution set with the non-dominated solution sets of previous generations, the degree of repetition of the solution set is checked. Specifically, the number of repetitions of each solution after multiple generations of iterations is calculated. If a solution continues to appear in multiple generations of iterations, it means that the population may have converged.

[0118] (6) Stop condition: When all non-dominated solutions appear repeatedly for a specified number of times k in multiple generations, the algorithm stops iterating. Otherwise, it returns to process (3) until the maximum number of iterations is reached or the convergence condition is met.

[0119] Step S103: inverting target ultrasonic grinding processing parameters of the target workpiece according to the Pareto solution set.

[0120] In this step, the surface quality parameters include the arithmetic mean height of the surface, the solid volume of the peaks, and the void volume of the valleys. Sa Related to surface quality and wear resistance, Vmp , the larger the value, the better the surface load-bearing and lubrication performance. Sa and Vmp , Vvv are positively correlated. From the perspective of actual engineering expectations, these three parameters can be considered monotonic, that is, Sa The smaller the better. Vmp , Vvv The larger the better, which is conducive to the realization of parameters in genetic algorithms Sa In the Pareto solution search within the specified interval, if Sa is directly optimized for a single objective, the optimization will only make it smaller. Therefore, the three-dimensional roughness parameter Sa (arithmetic mean height), Vmp (solid volume of the peak) and Vvv (Void volume in valleys) is used as a surface quality parameter to evaluate surface quality and performance.

[0121] In this embodiment, the entropy weight method and the superior-inferior solution distance method are used to assign weights and sort the Pareto solution set to determine the optimal Pareto sub-solution, and then the corresponding target ultrasonic grinding processing parameters (amplitude, rotation speed, and cutting depth) are inverted according to the optimal Pareto sub-solution.

[0122] In some embodiments, in step S103, inverting the target ultrasonic grinding processing parameters of the target workpiece according to the Pareto solution set includes the following steps:

[0123] Using entropy weight method and superior-inferior solution distance method to confirm the optimal Pareto sub-solution in the Pareto solution set;

[0124] The corresponding target ultrasonic grinding processing parameters are obtained according to the optimal Pareto sub-solution.

[0125] In this step, the original matrix is ​​constructed according to the Pareto solution set. According to the optimization direction (positive or negative) of each Pareto sub-solution in the Pareto solution set, the matrix data is standardized to eliminate the dimension effect. Then, the entropy weight method is used to calculate the weight of each solution, and the positive ideal solution and the negative ideal solution are calculated. The positive ideal solution and the negative ideal solution are derived from the maximum and minimum values ​​in the sample, respectively. Then, the distance between each evaluation index (sub-solution) and the positive ideal solution and the negative ideal solution is calculated, and the entropy weight is introduced. The negative rational distance is used to calculate the relative closeness between the sample and the optimal solution, and normalized. Finally, based on the degree of closeness, the optimal Pareto sub-solution in the Pareto solution set is determined, and then the target ultrasonic grinding processing parameters of the target workpiece are confirmed according to the optimal Pareto sub-solution.

[0126] In this step, the entropy weight method is used to assign weights to the Pareto solution set, and the entropy weights of each Pareto sub-solution in the Pareto solution set are obtained, which can more accurately reflect the importance of different objectives in the optimization process, thereby improving the accuracy of the optimization results. The superior and inferior solution distance method is used to sort the superior and inferior solutions of each Pareto solution according to the entropy weight of each Pareto sub-solution in the Pareto solution set, and the sorting results of each Pareto sub-solution in the Pareto solution set are obtained, which can clearly distinguish the superiority and inferiority of each Pareto sub-solution, help select the optimal Pareto sub-solution, and then confirm the optimal target ultrasonic grinding processing parameters of the corresponding target workpiece according to the optimal Pareto sub-solution, which can provide guidance for precision manufacturing in actual engineering and ensure that the processed surface has better quality and performance.

[0127] In this embodiment, the entropy weight method and the superior and inferior solution distance method are used to assign weights and sort the superior and inferior solutions of each Pareto solution in the Pareto solution set, specifically including:

[0128] The entropy weight method can objectively assign weights to each target from a statistical perspective. Based on the weight assignment results, the samples can be ranked by using the superior and inferior solution distance method. The basic principle of combining the entropy weight method with the superior and inferior solution distance method is as follows:

[0129] Assume that there are E evaluation samples and F optimization targets. First, we need to construct the original matrix X =( x ij ) E×F .

[0130] (19)

[0131] Next, the matrix data needs to be normalized in the positive direction. The processing method needs to be based on the optimization direction of each target. If the optimization direction is positive, it is processed according to formula (20), and if it is negative, it is processed according to formula (21). At the same time, this processing method can also eliminate the influence of dimension.

[0132] (20)

[0133] (twenty one),

[0134] The entropy weight method and the superior and inferior solution distance method have the same preprocessing method for the above data. j The goal is i The proportion of the sample p .

[0135] (twenty two),

[0136] Based on specific gravity p Calculate the entropy value of each indicator e .

[0137] (twenty three),

[0138] By entropy e Calculate the variation index of each indicator d .

[0139] (twenty four),

[0140] Reference Variation Index d Calculate the j Entropy weight of an indicator w .

[0141] (25),

[0142] The weight of each target can be calculated based on equations (19)-(25). The data matrix constructed based on equations (19)-(21) is X =( x ij ) E×F And the entropy weight obtained by formula (25) w .

[0143] The following is an introduction to the superior and inferior solution distance method:

[0144] First, we need to determine the matrix X The positive ideal solutions of each index in X + With negative ideal solution X - , which come from the maximum and minimum values ​​in the sample respectively.

[0145] (26),

[0146] (27),

[0147] Then calculate the distance between the evaluation index and the positive ideal solution and the negative ideal solution, and introduce the entropy weight obtained by formula (25) w .

[0148] (28)

[0149] (29),

[0150] The negative rational distance is used to calculate the relative closeness of the sample to the optimal solution and normalize it again.

[0151] (30),

[0152] (31),

[0153] The measured data is analyzed by entropy weight method to obtain the three-dimensional roughness parameters Sa , Vmp , Vvv The weights are 0.4, 0.3 and 0.3 respectively. Based on this weight, the Pareto solution set of different engineering requirements and the measured data closest to the requirements are scored and sorted using the superior and inferior solution distance method. The results are shown in Table 3.

[0154] It can be seen from Table 3 that for different engineering requirements, the Pareto optimal sub-solution given in this embodiment is better than the measured data. Sa The Pareto optimal sub-solutions for the engineering requirements of [0.15, 0.2], [0.3, 0.4], and [0.7 0.8] are 28.8%, 11.0%, and 39.9% higher than the measured data, respectively. This shows that the optimization effect is significant and can provide guidance for precise manufacturing in actual engineering.

[0155] Table 3

[0156]

[0157] In this way, the corresponding processing parameters can be inverted according to the surface quality requirements of ultrasonic grinding in actual engineering.

[0158] This embodiment extracts an objective function for characterizing the correlation between ultrasonic grinding processing parameters and surface quality parameters of a material from a preset prediction model; uses a genetic algorithm and an objective function to calculate a Pareto solution set based on the ultrasonic grinding processing parameters of a randomly generated target workpiece; a base workpiece and a target workpiece are workpieces of the same type; and inverts the target ultrasonic grinding processing parameters of the target workpiece based on the Pareto solution set, thereby effectively optimizing the ultrasonic grinding processing parameters so that the processed surface has better quality and performance. In addition, the surface roughness parameters are predicted using a prediction model based on actual processing data, thereby avoiding complex physical modeling and improving the efficiency and accuracy of calculating the processing parameters.

[0159] like Picture 5 As shown, some embodiments of the present application provide a system for accurate inversion of ultrasonic grinding process parameters, the system comprising an acquisition module 701, a calculation module 702, and an inversion module 703, specifically:

[0160] The acquisition module 701 is used to extract the target function from a preset prediction model; the prediction model is trained based on the ultrasonic grinding parameters of the basic workpiece and the surface quality parameters of the basic workpiece, and is used to characterize the corresponding relationship between the ultrasonic grinding parameters and the surface quality parameters;

[0161] A calculation module 702 is used to calculate a Pareto solution set using a genetic algorithm and an objective function according to the ultrasonic grinding processing parameters of the randomly generated target workpiece; the base workpiece and the target workpiece are workpieces of the same type;

[0162] The inversion module 703 is used to invert the target ultrasonic grinding processing parameters of the target workpiece according to the Pareto solution set.

[0163] In one embodiment, the acquisition module 701 is further used to: acquire a model training set, wherein the model training set includes ultrasonic grinding processing parameters and surface quality parameters of the basic workpiece.

[0164] In one embodiment, the acquisition module 701 is also used to: train multiple preset basic machine learning models based on the model training set.

[0165] In one embodiment, the acquisition module 701 is also used to: obtain model performance parameters of multiple basic machine learning models.

[0166] In one embodiment, the calculation module 702 is also used to select a basic machine learning model with the best model performance parameters from multiple basic machine learning models as a prediction model.

[0167] In one implementation, the calculation module 702 is further configured to determine the surface quality parameters of the target workpiece according to the target function and the ultrasonic grinding processing parameters of the target workpiece.

[0168] In one embodiment, the calculation module 702 is further configured to calculate a Pareto solution set corresponding to the surface quality parameter of the target workpiece by using a non-dominated genetic algorithm.

[0169] In one embodiment, the calculation module 702 is further configured to generate genetic parameters according to the surface quality parameters of the target workpiece.

[0170] In one embodiment, the calculation module 702 is further used to perform genetic iteration of the non-dominated genetic algorithm according to the genetic parameters.

[0171] In one embodiment, the inversion module 703 is further used to: when the i-th generation non-dominated solution is obtained, determine the number of repetitions of the i-th generation non-dominated solution and the non-dominated solution in the previous i-1 generation non-dominated solution; wherein i is an integer greater than 1.

[0172] In one embodiment, the inversion module 703 is further configured to: when the number of repetitions of the non-dominated solutions exceeds a preset number, obtain a Pareto solution set according to the i-th generation non-dominated solutions.

[0173] In one embodiment, the inversion module 703 is further used to: confirm the optimal Pareto sub-solution in the Pareto solution set by using an entropy weight method and a superior-inferior solution distance method.

[0174] In one embodiment, the inversion module 703 is further used to obtain corresponding target ultrasonic grinding processing parameters according to the optimal Pareto sub-solution.

[0175] It should be noted that the ultrasonic grinding process parameter precise inversion system provided in this embodiment and the above-mentioned ultrasonic grinding process parameter precise inversion method are based on the same inventive concept, so the relevant content of the above-mentioned ultrasonic grinding process parameter precise inversion method is also applicable to the content of the ultrasonic grinding process parameter precise inversion system, so it will not be repeated here.

[0176] like Picture 6 The embodiment of the present application further provides an electronic device, the electronic device comprising:

[0177] At least one hydrogen fuel cell;

[0178] at least one memory;

[0179] at least one processor;

[0180] at least one program;

[0181] The program is stored in the memory, and the processor executes at least one program to implement the above-mentioned method for accurate inversion of ultrasonic grinding processing parameters in the present disclosure.

[0182] The electronic device may be any intelligent terminal including a mobile phone, a tablet computer, a personal digital assistant (PDA), a vehicle-mounted computer, etc.

[0183] The electronic device according to the embodiment of the present application is described in detail below.

[0184] The processor 1600 may be implemented by a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present disclosure;

[0185] The memory 1700 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1700 can store an operating system and other application programs. When the technical solution provided in the embodiments of this specification is implemented by software or firmware, the relevant program code is stored in the memory 1700, and the processor 1600 calls and executes a method for accurate inversion of ultrasonic grinding processing parameters in the embodiment of the present disclosure.

[0186] Input / output interface 1800, used to implement information input and output;

[0187] Communication interface 1900, used to realize communication interaction between the device and other devices, which can be realized through wired mode (such as USB, network cable, etc.) or wireless mode (such as mobile network, WIFI, Bluetooth, etc.);

[0188] Bus 2000 , which transmits information between various components of the device (e.g., processor 1600 , memory 1700 , input / output interface 1800 , and communication interface 1900 );

[0189] The processor 1600 , the memory 1700 , the input / output interface 1800 , and the communication interface 1900 are connected to each other in communication within the device via the bus 2000 .

[0190] The embodiment of the present disclosure further provides a storage medium, which is a computer-readable storage medium. The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the above-mentioned method for accurate inversion of ultrasonic grinding processing parameters.

[0191] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0192] The embodiment of the present disclosure also provides a computer program product. When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device is used to execute the above-mentioned method for accurate inversion of ultrasonic grinding processing parameters.

[0193] The embodiments described in the embodiments of the present disclosure are intended to more clearly illustrate the technical solutions of the embodiments of the present disclosure and do not constitute a limitation on the technical solutions provided by the embodiments of the present disclosure. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of the present disclosure are also applicable to similar technical problems.

[0194] Those skilled in the art will appreciate that the technical solutions shown in the figures do not limit the embodiments of the present disclosure and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0195] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0196] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0197] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0198] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0199] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0200] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0201] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0202] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions to enable an electronic device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of each embodiment method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store programs.

[0203] The above is a specific description of the preferred implementation of the embodiments of the present application, but the embodiments of the present application are not limited to the above-mentioned implementation methods. Technical personnel familiar with the field can also make various equivalent modifications or substitutions without violating the spirit of the embodiments of the present application. These equivalent modifications or substitutions are all included in the scope defined by the claims of the embodiments of the present application.

[0204] The embodiments of the present application are described in detail above in conjunction with the accompanying drawings, but the present application is not limited to the above embodiments. Various changes can be made within the knowledge scope of ordinary technicians in the relevant technical field without departing from the purpose of the present application.

Claims

1. A method for accurate inversion of ultrasonic grinding process parameters, characterized in that: The method comprises: Extracting an objective function from a preset prediction model; the prediction model is trained based on ultrasonic grinding processing parameters of a basic workpiece and surface quality parameters of the basic workpiece, and is used to characterize the corresponding relationship between the ultrasonic grinding processing parameters and the surface quality parameters; According to the ultrasonic grinding processing parameters of the randomly generated target workpiece, a Pareto solution set is calculated using a genetic algorithm and the objective function; the basic workpiece and the target workpiece are workpieces of the same type; Inversely analyzing target ultrasonic grinding processing parameters of the target workpiece according to the Pareto solution set; The method of calculating the Pareto solution set by using a genetic algorithm and the objective function according to the ultrasonic grinding processing parameters of the randomly generated target workpiece includes: Determine the surface quality parameters of the corresponding target workpiece according to the target function and the ultrasonic grinding processing parameters of the target workpiece; Calculating the Pareto solution set corresponding to the surface quality parameters of the target workpiece by a non-dominated genetic algorithm; The surface quality parameters include the arithmetic mean height of the surface, the solid volume of the peaks and the void volume of the valleys.

2. The ultrasonic grinding process parameter accurate inversion method according to claim 1, characterized in that: Before extracting the target function from the preset prediction model, the method further includes: Acquire a model training set, wherein the model training set includes ultrasonic grinding processing parameters and surface quality parameters of a basic workpiece; Training a plurality of preset basic machine learning models based on the model training set; Obtaining model performance parameters of the multiple basic machine learning models; A basic machine learning model with the best model performance parameters is selected from the multiple basic machine learning models as the prediction model.

3. The method for accurate inversion of ultrasonic grinding process parameters according to claim 1, characterized in that: The Pareto solution set corresponding to the surface quality parameter of the target workpiece is calculated by a non-dominated genetic algorithm, comprising: generating genetic parameters according to the surface quality parameters of the target workpiece; Perform genetic iteration of the non-dominated genetic algorithm according to genetic parameters; When the i-th generation non-dominated solution is obtained, the number of repetitions of the i-th generation non-dominated solution and the non-dominated solution in the previous i-1 generation non-dominated solution is determined; wherein i is an integer greater than 1; When the number of repetitions of the non-dominated solution exceeds a preset number, a Pareto solution set is obtained according to the i-th generation non-dominated solution.

4. The method for accurate inversion of ultrasonic grinding process parameters according to claim 1, characterized in that: The step of inverting the target ultrasonic grinding processing parameters of the target workpiece according to the Pareto solution set includes: Using entropy weight method and superior-inferior solution distance method to confirm the optimal Pareto sub-solution in the Pareto solution set; The corresponding target ultrasonic grinding processing parameters are obtained according to the optimal Pareto sub-solution.

5. An ultrasonic grinding process parameter accurate inversion system, characterized in that: The system comprises: An acquisition module, used to extract the objective function from a preset prediction model; the prediction model is trained based on the ultrasonic grinding parameters of the basic workpiece and the surface quality parameters of the basic workpiece, and is used to characterize the corresponding relationship between the ultrasonic grinding parameters and the surface quality parameters; A calculation module, used to calculate a Pareto solution set using a genetic algorithm and the objective function according to the ultrasonic grinding processing parameters of the randomly generated target workpiece; the basic workpiece and the target workpiece are workpieces of the same type; An inversion module, used for inverting target ultrasonic grinding processing parameters of the target workpiece according to the Pareto solution set; The method of calculating the Pareto solution set by using a genetic algorithm and the objective function according to the ultrasonic grinding processing parameters of the randomly generated target workpiece includes: Determine the surface quality parameters of the corresponding target workpiece according to the target function and the ultrasonic grinding processing parameters of the target workpiece; Calculating the Pareto solution set corresponding to the surface quality parameters of the target workpiece by a non-dominated genetic algorithm; The surface quality parameters include the arithmetic mean height of the surface, the solid volume of the peaks and the void volume of the valleys.

6. An electronic device, characterized in that: It includes at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor so that the at least one control processor can execute the method for accurate inversion of ultrasonic grinding processing parameters as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the method for accurate inversion of ultrasonic grinding processing parameters according to any one of claims 1 to 4.

8. A computer program product, characterized in that When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the method for accurate inversion of ultrasonic grinding process parameters as described in any one of claims 1 to 4.

Images