Cutter feature screening method and system, electronic equipment and medium

By creating three-dimensional arrays and improving the dung beetle optimization algorithm, combining the chaos mapping algorithm and flight strategy, the optimal feature set of tools is selected, which solves the problems of information redundancy and feature correlation in tool wear monitoring, and improves classification accuracy and screening Accuracy.

CN118916612APending Publication Date: 2024-11-08CHINA JILIANG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410963521.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-18
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

In tool wear monitoring, some features are not related to wear conditions, resulting in information redundancy and unnecessary computing features, affecting the accuracy of model classification.

Method used

By creating three-dimensional arrays and improving the dung beetle optimization algorithm, combining chaotic mapping algorithms and flight strategies, optimizing the combination of feature indicator weights, filtering out the optimal feature set to avoid local optimal solutions and information redundancy.

Benefits of technology

It improves the accuracy and classification accuracy of tool feature screening, reduces unnecessary calculations, enhances global search capabilities, and avoids information redundancy and local optimal solutions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118916612A_ABST
    Figure CN118916612A_ABST
Patent Text Reader

Abstract

The invention discloses a tool feature screening method and system, electronic equipment and a medium, and the method comprises the steps: obtaining a plurality of original tool signals, and obtaining a tool feature data set according to the original tool signals; initializing algorithm parameters of a preset improved dung beetle optimization algorithm, and creating a three-dimensional array; performing position initialization on the three-dimensional array based on a chaotic mapping algorithm and algorithm parameters, and determining a plurality of target dung beetle individuals; updating the position of the target dung beetle individual through the flight strategy and the inertia weight, and updating the breeding and foraging behaviors of the target dung beetle individual to obtain an updating result; when an updating result meets an iteration threshold value in the algorithm parameters, determining a feature index weight combination; and screening the tool feature data set according to the feature index weight combination to obtain target tool features. In the embodiment of the invention, interference of irrelevant features and redundant features can be avoided by creating three groups of arrays, the accuracy of model classification is improved, and the precision of signal screening is further improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of signal processing technology, and in particular to a tool feature screening method, system, electronic device and storage medium. Background Art

[0002] With the development of intelligent manufacturing technology, CNC machine tools account for an increasing proportion in the manufacturing industry and are increasingly widely used in the field of mechanical processing. Tools are important components of machine tools, and their condition directly affects the quality of workpieces, production efficiency, and processing costs. Therefore, it is necessary to monitor the wear of the tool and replace the tool reasonably according to the wear state of the tool. Among them, the monitoring feature set of the tool is mainly obtained by time domain, frequency domain, and time-frequency domain analysis. After feature extraction, a large amount of feature information will be obtained.

[0003] However, some features are not related to tool wear. Sampling only a single feature index as an indicator for calculating the correlation between features and labels may not fully capture the complex structure of high-order data. In addition, there is correlation between some features, which leads to information redundancy. Irrelevant and redundant features will cause a lot of unnecessary calculations, affecting the accuracy of model classification. Summary of the invention

[0004] The embodiments of the present application provide a tool feature screening method, system, electronic device and storage medium. By creating three groups of arrays, it is possible to avoid interference from irrelevant and redundant features, improve the accuracy of model classification, and further improve the accuracy of signal screening.

[0005] In a first aspect, an embodiment of the present application provides a method for screening tool features, the method comprising:

[0006] Acquire multiple original tool signals, and perform preprocessing operations on the original tool signals to obtain a tool feature data set, wherein the original tool signals are obtained by collecting the tool in different directions;

[0007] Initialize the algorithm parameters of the preset improved dung beetle optimization algorithm and create a three-dimensional array, wherein the three-dimensional array is used to store combinations of different indicator weights;

[0008] Initializing the position of the three-dimensional array based on a preset chaotic mapping algorithm and the algorithm parameters to determine a plurality of target dung beetle individuals;

[0009] The position of the target dung beetle individual is updated through a preset flight strategy and a preset inertia weight, and the reproduction and foraging behavior of the target dung beetle individual is updated based on the algorithm parameters to obtain an updated result;

[0010] When the update result satisfies the iteration threshold in the algorithm parameter, determining a feature index weight combination according to the update result;

[0011] The tool feature data set is screened according to the feature index weight combination to obtain target tool features.

[0012] The tool feature screening method provided in the embodiment of the present application has at least the following beneficial effects: first, a plurality of original tool signals are obtained, and the original tool signals are preprocessed to obtain a tool feature data set, thereby extracting a tool feature data set of a vibration signal with less interference, and then initializing the preset algorithm parameters of the improved dung beetle optimization algorithm, and creating a three-dimensional array, so that the combination of different indicator weights can be stored through the three-dimensional array to provide a search space for the subsequent optimization process, and at the same time, the parameter combination of the search space in the optimization process can be defined, and then, based on the preset chaotic mapping algorithm and algorithm parameters, the position of the three-dimensional array is initialized, and a plurality of target dung beetle individuals with relatively uniform position distribution in the search space of the three-dimensional array are determined, thereby increasing the number of species. The diversity of the group is determined, and the position of the target dung beetle individual is updated through a preset flight strategy and a preset inertia weight. Global exploration is performed through long-distance jumps to avoid local optimality, and the search step size is adaptively adjusted through the inertia weight to balance global exploration and local development. The reproduction and foraging behavior of the target dung beetle individual is updated based on the algorithm parameters to obtain an updated result. The position of the dung beetle individual is iteratively updated to gradually approach the optimal solution. When the updated result meets the iteration threshold in the algorithm parameters, the feature index weight combination is determined according to the updated result, that is, the optimal feature set is found. Finally, the tool feature data set is screened according to the feature index weight combination to obtain the target tool feature, thereby achieving accurate screening of the tool feature and improving classification accuracy. The embodiment of the present application creates a three-dimensional array to comprehensively consider the pros and cons of different feature combinations, so as to define the parameter combination of the search space in the optimization process, and update the position of the target dung beetle individual through different strategies. The optimal solution of the weight of each evaluation index is selected through the cooperation of the three-dimensional array and the improved dung beetle optimization algorithm to achieve the purpose of optimal feature selection, avoid falling into the local optimal solution and insufficient exploration ability, improve the global search capability, and improve the accuracy of feature classification.

[0013] In some embodiments, the preprocessing operation on the original tool signal to obtain the tool feature data set includes:

[0014] Performing filtering processing on the original tool signal;

[0015] Perform feature extraction on the tool signal after filtering to obtain multiple columns of features and tool wear columns;

[0016] A tool feature data set is generated according to the plurality of feature columns and the tool wear column.

[0017] In some embodiments, creating a three-dimensional array includes:

[0018] A three-dimensional array is created according to preset index requirements, wherein the index requirements include a Spearman correlation index, a robustness index, and a mutual information index.

[0019] In some embodiments, the algorithm parameters include population size; the position initialization of the three-dimensional array based on the preset chaotic mapping algorithm and the algorithm parameters to determine multiple target dung beetle individuals includes:

[0020] Initializing the position of the population size in the search space of the three-dimensional array by using a preset chaotic mapping algorithm to obtain a plurality of dung beetle individuals with uniform position distribution;

[0021] For each of the dung beetle individuals, determining a parameter combination of the dung beetle individual;

[0022] Performing index calculation on the parameter combination through a preset characteristic evaluation index function to obtain a comprehensive index value;

[0023] Sort the parameter combinations in descending order according to the comprehensive index values ​​to obtain a feature sequence;

[0024] Filtering a feature set at a preset position in the feature sequence, inputting the feature set into a preset support vector machine model, and outputting a fitness value;

[0025] All the dung beetle individuals are screened according to the fitness values ​​to determine a plurality of target dung beetle individuals.

[0026] In some embodiments, updating the position of the target dung beetle individual by using a preset flight strategy and a preset inertia weight includes:

[0027] Determining the individual type of the target dung beetle individual;

[0028] When the individual type is a ball-rolling dung beetle, updating the ball-rolling position of the ball-rolling dung beetle according to a preset flight strategy;

[0029] When the individual type is a thieving dung beetle, the stealing position of the thieving dung beetle is updated based on a preset inertia weight.

[0030] In some embodiments, the algorithm parameters include an upper bound of the search space and a lower bound of the search space; and updating the reproduction and foraging behavior of the target dung beetle individual based on the algorithm parameters to obtain an updated result includes:

[0031] When the individual type is a breeding dung beetle or a foraging dung beetle, a safe area of ​​the breeding dung beetle or the foraging dung beetle is simulated based on a preset boundary selection strategy, the search space upper bound, and the search space lower bound;

[0032] updating the breeding behavior of the breeding dung beetle based on the safety area, and updating the foraging behavior of the foraging dung beetle;

[0033] Calculate the first position of the rolling ball position of the rolling dung beetle after being updated, the second position of the stealing position of the thieving dung beetle after being updated, the third position of the breeding behavior of the breeding dung beetle after being updated, the fourth position of the foraging behavior of the foraging dung beetle after being updated, and the number of iterations of the position of the target dung beetle individual;

[0034] Calculating an error rate value of the target dung beetle individual according to the first position, the second position, the third position, and the fourth position;

[0035] When the error rate value is less than a preset global error rate value, recording the target fitness value and target position corresponding to the error rate value;

[0036] An update result is generated according to the target fitness value, the target position and the number of position iterations.

[0037] In some embodiments, determining a feature index weight combination according to the update result includes:

[0038] Determining a target fitness value and a target position corresponding to the update result;

[0039] Determine a plurality of candidate indicator weight combinations according to the target fitness value and the target position, and sort all the candidate indicator weight combinations in descending order to obtain a combination sequence;

[0040] Selecting a weight combination at a preset position in the combination sequence, inputting the weight combination into a preset support vector machine model, and outputting a classification accuracy rate;

[0041] The classification accuracy is compared with the preset feature indicator weights to determine the feature indicator weight combination.

[0042] In a second aspect, an embodiment of the present application further provides a tool feature screening system, the system comprising:

[0043] A data acquisition module, used for acquiring a plurality of original tool signals and performing preprocessing operations on the original tool signals to obtain a tool feature data set, wherein the original tool signals are obtained by collecting the tool in different directions;

[0044] An array creation module, used to initialize the algorithm parameters of the preset improved dung beetle optimization algorithm and create a three-dimensional array, wherein the three-dimensional array is used to store a combination of different indicator weights;

[0045] An individual determination module is used to initialize the position of the three-dimensional array based on a preset chaotic mapping algorithm and the algorithm parameters to determine a plurality of target dung beetle individuals;

[0046] A position updating module, used to update the position of the target dung beetle individual through a preset flight strategy and a preset inertia weight, and to update the reproduction and foraging behavior of the target dung beetle individual based on the algorithm parameters to obtain an update result;

[0047] A combination determination module, configured to determine a feature index weight combination according to the update result when the update result satisfies the iteration threshold in the algorithm parameter;

[0048] The feature screening module is used to screen the tool feature data set according to the feature index weight combination to obtain the target tool feature.

[0049] In a third aspect, an embodiment of the present application further provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the tool feature screening method as described in the first aspect is implemented.

[0050] In a fourth aspect, an embodiment of the present application further 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 tool feature screening method as described in the first aspect.

[0051] Other features and advantages of the present application will be described in the following description, and partly become apparent from the description, or understood by practicing the present application. The purpose and other advantages of the present application can be realized and obtained through the structures particularly pointed out in the description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings are used to provide further understanding of the technical solution of the present application and constitute a part of the specification. Together with the examples of the present application, they are used to explain the technical solution of the present application and do not constitute a limitation on the technical solution of the present application.

[0053] Figure 1 is a flowchart of a specific method of the tool feature screening method provided in an embodiment of the present application;

[0054] Figure 2 is a specific flow chart of preprocessing operation of original tool signal provided by the embodiment of the present application;

[0055] Figure 3 is a specific flow chart of creating a three-dimensional array provided in an embodiment of the present application;

[0056] Figure 4 It is a specific flow chart of initializing the position of a three-dimensional array based on a preset chaotic mapping algorithm and algorithm parameters provided in an embodiment of the present application;

[0057] Figure 5 It is a specific flow chart of updating the position of a target dung beetle individual through a preset flight strategy and a preset inertia weight provided in an embodiment of the present application;

[0058] Figure 6 It is a specific flow chart for updating the reproduction and foraging behavior of target dung beetle individuals provided in the embodiment of the present application;

[0059] Figure 7 It is a specific flow chart of determining the weight combination of characteristic indicators according to the update result provided in the embodiment of the present application;

[0060] Figure 8 is a system schematic diagram of a tool feature screening system provided in an embodiment of the present application;

[0061] Fig. 9 Comparison diagram before and after the improved Circle chaos mapping provided for this application example;

[0062] Fig.10 It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0064] It should be noted that, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0065] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0066] A tool feature screening method provided in the embodiment of the present application can be applied to a terminal, can also be applied to a server side, and can also be software running in a terminal or a server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer or a smart watch, etc.; the server side can be configured as an independent physical server, or a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN) and big data and artificial intelligence platforms; the software can be an application that implements the above method, etc., but is not limited to the above forms.

[0067] Embodiments of the present application can be used in numerous general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer computer devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0068] With the development of intelligent manufacturing technology, CNC machine tools account for an increasing proportion in the manufacturing industry and are increasingly widely used in the field of mechanical processing. Tools are important components of machine tools, and their condition directly affects the quality of workpieces, production efficiency, and processing costs. Therefore, it is necessary to monitor the wear of the tool and replace the tool reasonably according to the wear state of the tool. Among them, the monitoring feature set of the tool is mainly obtained by time domain, frequency domain, and time-frequency domain analysis. After feature extraction, a large amount of feature information will be obtained.

[0069] However, some features are not related to tool wear. Sampling only a single feature index as an indicator for calculating the correlation between features and labels may not fully capture the complex structure of high-order data. In addition, there is correlation between some features, which leads to information redundancy. Irrelevant and redundant features will cause a lot of unnecessary calculations, affecting the accuracy of model classification.

[0070] In order to solve the above problems, the embodiments of the present application provide a tool feature screening method, system, electronic device and storage medium. First, multiple original tool signals are obtained, and the original tool signals are preprocessed to obtain a tool feature data set, thereby extracting a tool feature data set of vibration signals with less interference, and then the preset algorithm parameters of the improved dung beetle optimization algorithm are initialized, and a three-dimensional array is created, so that the combination of different indicator weights can be stored through the three-dimensional array to provide a search space for the subsequent optimization process, and at the same time, the parameter combination of the search space in the optimization process can be defined. After that, the three-dimensional array is initialized based on the preset chaotic mapping algorithm and algorithm parameters, and multiple relatively uniformly distributed positions in the search space of the three-dimensional array are determined. The target dung beetle individual is then updated through a preset flight strategy and a preset inertia weight to update the position of the target dung beetle individual, a global exploration is performed through long-distance jumps to avoid local optimality, and the search step length is adaptively adjusted through the inertia weight to balance global exploration and local development, and the reproduction and foraging behavior of the target dung beetle individual is updated based on the algorithm parameters to obtain an updated result, and the position of the dung beetle individual is iteratively updated to gradually approach the optimal solution. When the updated result meets the iteration threshold in the algorithm parameters, the feature index weight combination is determined according to the updated result, that is, the optimal feature set is found, and finally the tool feature data set is screened according to the feature index weight combination to obtain the target tool feature, thereby achieving accurate screening of the tool feature and improving classification accuracy. The embodiment of the present application creates a three-dimensional array to comprehensively consider the pros and cons of different feature combinations, so as to define the parameter combination of the search space in the optimization process, and update the position of the target dung beetle individual through different strategies. The optimal solution of the weight of each evaluation index is selected through the cooperation of the three-dimensional array and the improved dung beetle optimization algorithm to achieve the purpose of optimal feature selection, avoid falling into the local optimal solution and insufficient exploration ability, improve the global search capability, and improve the accuracy of feature classification.

[0071] The embodiments of the present invention are further described below in conjunction with the accompanying drawings.

[0072] Reference Figure 1 , Figure 1 1 is a flowchart of a specific method of the tool feature screening method provided in an embodiment of the present application. In some embodiments, the tool feature screening method includes but is not limited to steps S101 to S106.

[0073] Step S101, acquiring a plurality of original tool signals, and performing preprocessing operations on the original tool signals to obtain a tool feature data set.

[0074] It should be noted that the original tool signal is obtained by collecting the tool in different directions.

[0075] In step S101 of some embodiments, first, the original tool signals of the tool in different directions are acquired, and the original tool signals are preprocessed to obtain vibration signals with less interference and obtain a tool feature data set, which is helpful for subsequent signal analysis and processing.

[0076] It is worth noting that the original tool signal in the embodiment of the present application is obtained by collecting vibration signals in different directions through an acceleration sensor, for example, collecting vibration signals in the X, Y and Z axis directions through an acceleration sensor.

[0077] Step S102, initializing the algorithm parameters of the preset improved dung beetle optimization algorithm and creating a three-dimensional array.

[0078] It should be noted that the three-dimensional array is used to store combinations of different indicator weights.

[0079] In step S102 of some embodiments, the algorithm parameters of a preset improved dung beetle optimization algorithm (Improved DungBeetle Optimization, IDBO) are initialized. By initializing the algorithm parameters, the feasibility of the solution can be ensured, the divergence of the improved dung beetle optimization algorithm can be prevented, the stability of the algorithm can be ensured, and a three-dimensional array is created, so that the three-dimensional array can store different indicator weight combinations, and different indicator weight combinations are used as search spaces in the optimization process, so that the sizes of different weights in the three-dimensional array can be dynamically adjusted through the iterative process of the algorithm, thereby balancing the needs of global search and local search.

[0080] It should be noted that the algorithm parameters include but are not limited to population size, parameter dimension, maximum iteration threshold, search space upper bound and search space lower bound.

[0081] Among them, the population size is used to characterize the total number of dung beetle individuals, which is also the total number of candidate solutions used by the improved dung beetle optimization algorithm. The parameter dimension is used to characterize the number of independent parameters in each candidate solution. The upper bound of the search space defines the maximum value allowed for each parameter in the optimization problem, and the lower bound of the search space defines the minimum value allowed for each parameter in the optimization problem. By initializing the algorithm parameters, the feasibility of the solution can be ensured, the improved dung beetle optimization algorithm can be prevented from diverging, and the stability of the algorithm can be ensured.

[0082] Specifically, the population size of dung beetles in the embodiment of the present application is N=50, the parameter dimension dim=3, the maximum iteration threshold tmax=200, the search space upper bound Ub=[0,0,0] and the lower bound Lb=[1,1,1].

[0083] Step S103, initializing the position of the three-dimensional array based on a preset chaotic mapping algorithm and algorithm parameters, and determining a plurality of target dung beetle individuals.

[0084] In step S103 of some embodiments, the position of the three-dimensional array is initialized based on a preset chaotic mapping algorithm and algorithm parameters to obtain multiple target dung beetle individuals with uniform position distribution. The chaotic mapping algorithm can increase the diversity of the population and prevent the algorithm from falling into a local optimal solution.

[0085] It should be noted that the chaotic mapping algorithm in the embodiment of the present application is an improved Circle chaotic mapping algorithm.

[0086] Step S104, updating the position of the target dung beetle individual through a preset flight strategy and a preset inertia weight, and updating the reproduction and foraging behavior of the target dung beetle individual based on the algorithm parameters to obtain an updated result.

[0087] In step S104 of some embodiments, the position of the target dung beetle individual is updated through a preset flight strategy and a preset inertia weight. The flight strategy enables the dung beetle individual to perform a larger random movement after the position is updated, so that it can explore a wider area, balancing the breadth and depth of the algorithm. The inertia weight makes it possible to use a larger weight in the early stage of the algorithm iteration, so that the algorithm can quickly reach the vicinity of the target value with a larger step; and in the later stage of the iteration, a smaller weight should be used, so that the algorithm moves with a smaller step, so as to accurately search near the target solution and better local optimization. The reproduction and foraging behavior of the target dung beetle individual is updated based on the algorithm parameters to obtain an updated result, which is convenient for subsequent judgment on whether it is necessary to continue to update the location of the dung beetle.

[0088] It should be noted that the flight strategy of the embodiment of the present application is the Levy Flight Strategy (Levy), and the inertia weight in the embodiment of the present application is the inertia weight based on chaos adaptation. The inertia weight based on chaos adaptation is a strategy for adjusting the search step size in the optimization algorithm. This strategy combines the concepts of chaos theory and inertia weight to improve the search capability and efficiency of the algorithm.

[0089] Step S105: when the update result satisfies the iteration threshold in the algorithm parameter, a feature index weight combination is determined according to the update result.

[0090] In step S105 of some embodiments, when the number of position iterations in the update result meets the iteration threshold in the algorithm parameters, it means that the algorithm has reached the maximum number of iterations and needs to stop calculating. The feature indicator weight combination is determined directly based on the update result, and the optimal fitness value and the optimal position are output to realize the search for the optimal solution and the optimal position to obtain the optimal weight combination.

[0091] Step S106, screening the tool feature data set according to the feature index weight combination to obtain the target tool feature.

[0092] In step S106 of some embodiments, the tool feature data set is classified according to the feature index weight combination, and the required target tool features are screened out to avoid a large amount of unnecessary calculations caused by irrelevant and redundant features, thereby achieving effective selection of tool features.

[0093] Reference Figure 2 , Figure 2 It is a specific flow chart of preprocessing the original tool signal provided by the embodiment of the present application. In some embodiments, the method includes but is not limited to step S201 and step S203.

[0094] Step S201, filtering the original tool signal.

[0095] Step S202, extracting features from the tool signal after filtering to obtain multiple columns of features and a tool wear column.

[0096] Step S203, generating a tool feature data set according to the multiple columns of features and the tool wear column.

[0097] In some embodiments, in steps S201 to S203, during the preprocessing operation of the original tool signal, it is necessary to filter the original tool signal to obtain a vibration signal with less interference, reduce the noise component in the signal, enhance the useful signal, thereby improving the signal-to-noise ratio of the signal, and the clean signal makes the feature extraction more accurate, which is helpful for subsequent signal analysis and processing. The embodiment of the present application also extracts features from the tool signal after filtering to obtain multiple columns of features and tool wear columns, which is convenient for the subsequent setting of the Spearman correlation index, and realizes the extraction of different features. Finally, a tool feature data set is generated based on the multiple columns of features and tool wear columns, so as to increase the diversity and comprehensiveness of the tool features. The multiple columns of features and tool wear columns can identify the patterns and trends of the tool under different wear conditions.

[0098] It should be noted that the multiple columns of features in the embodiment of the present application are used to characterize a comprehensive description of the tool state, including different aspects and attributes of the vibration signal, wherein different feature columns may represent different statistical characteristics of the vibration signal; the tool wear column is used to reflect the degree of tool wear.

[0099] Reference Figure 3 , Figure 3 The method includes but is not limited to step S301.

[0100] Step S301, creating a three-dimensional array according to preset index requirements.

[0101] It should be noted that the index requirements include Spearman correlation index, robustness index and mutual information index.

[0102] In step S301 of some embodiments, a three-dimensional array is created according to the preset Spearman correlation index, robustness index and mutual information index, so that the three-dimensional array can store different indicator weight combinations, and different indicator weight combinations are used as search spaces in the optimization process, so that the sizes of different weights in the three-dimensional array can be dynamically adjusted through the iterative process of the algorithm, thereby balancing the needs of global search and local search.

[0103] Specifically, the Spearman correlation index in the embodiment of the present application is used to reflect the monotonic relationship between two variables. The embodiment of the present application analyzes the rank relationship between variables. When the correlation coefficient is large, the internal correlation of the two variables is strong, and vice versa, the correlation is weak. The Spearman correlation index formula is expressed as follows:

[0104]

[0105] Where x represents multiple columns of features, y represents tool wear columns; x, y are the means of x and y respectively; n is the number of samples of data; Sp(x, y) represents the correlation coefficient of two variables. The result of Sp(x, y) is in the interval [-1, 1]. The correlation close to -1 or 1 indicates that there is a high negative correlation or positive correlation between the variables. The embodiment of the present application takes the absolute value of Sp(x, y) to represent the degree of correlation of the variables.

[0106] In some embodiments, the robustness index is used to characterize the resistance of the feature to noise and outliers. This embodiment combines the Gaussian filtering method to calculate the robustness index. The process of calculating the robustness index is specifically described below.

[0107] Specifically, for the characteristic signal sequence F = [f(t1), f(t2), ..., f(t K )] and the time series T = [t1, t2, ..., t K ], f(t K ) indicates that at time t K The corresponding eigenvalue at, K is the total length of the sample time or the number of sample feature points. For the monitoring signal value, the one-dimensional Gaussian calculation formula is expressed as follows:

[0108]

[0109] Afterwards, the Gaussian smoothing filter method is used to decompose the original feature sequence into two parts: a steady trend and a random residual. The specific expression is:

[0110] f(t k)=f T (t k )+f R (t k );

[0111] Among them, f T represents the stable part, f R Represents random margin.

[0112] Furthermore, an exponential function is introduced to characterize the formula of the robustness of the feature evaluation index. The specific formula is as follows:

[0113]

[0114] It is understandable that if the random margin f R (t k ) is smaller, the ratio of the random residual to the original feature sequence value f R (t k ) / f(t k ) is smaller, the function graph y=exp(-f R (t k ) / f(t k )) in the variable -|f R (t k ) / f(t k )| will be closer to the origin, and finally the value of Rob(F) will be closer to 1. On the contrary, if the feature sequence image curve is more jittery, the final robustness index function Rob(F) value will be closer to 0. Therefore, the value of Rob(F) is between [0,1], and the closer the value is to 1, the better the robustness of the feature sequence.

[0115] In some embodiments, the mutual information index is used as one of the indicators of the preferred feature set. Since the correlation between feature values ​​will affect the prediction accuracy and learning effect of the model, the mutual information index can be used as an indicator of the preferred feature to obtain the degree of correlation between the features.

[0116] Specifically, the mutual information I(x; y) between a feature value x and tool wear value y in any feature set is defined as:

[0117]

[0118] Among them, x represents multiple columns of features, y represents the tool wear column, and p(x i )、p(y j ) are the probability densities of random variables x and y respectively; p(x i ,y j) is the joint probability density of x and y; I(x,y) represents the mutual information between x and y. If x and y are independent, then p(x,y)=p(x)p(y), and the result value of the mutual information I(x;y) is 0, which means that x and y are unrelated. On the contrary, if the relationship between x and y is closer, the mutual information I(x;y) will be greater.

[0119] Reference Figure 4 , Figure 4 This is a specific flow chart of initializing the position of a three-dimensional array based on a preset chaotic mapping algorithm and algorithm parameters provided in an embodiment of the present application. The method includes but is not limited to step S401 and step S406.

[0120] It should be noted that the algorithm parameters in the embodiments of the present application include population size.

[0121] Step S401, initializing the position of the population size in the search space of the three-dimensional array by using a preset chaotic mapping algorithm to obtain a plurality of dung beetle individuals with uniform position distribution.

[0122] Step S402: for each dung beetle individual, determine the parameter combination of the dung beetle individual.

[0123] Step S403, performing index calculation on the parameter combination by using a preset feature evaluation index function to obtain a comprehensive index value.

[0124] Step S404, sorting the parameter combinations in descending order according to the comprehensive index values ​​to obtain a feature sequence.

[0125] Step S405, filtering out a feature set at a preset position in the feature sequence, inputting the feature set into a preset support vector machine model, and outputting a fitness value.

[0126] Step S406, screening all dung beetle individuals according to the fitness values ​​to determine a plurality of target dung beetle individuals.

[0127] In steps S401 to S406 of some embodiments, in the process of initializing the position of the three-dimensional array based on the preset chaotic mapping algorithm and algorithm parameters, first, the population size is initialized in the search space of the three-dimensional array through the preset chaotic mapping algorithm to obtain a plurality of dung beetle individuals with uniform position distribution. Individuals with relatively uniform initial position distribution can be generated through chaotic mapping, which helps to increase the diversity of the population and avoid the algorithm from falling into a local optimal solution. For each dung beetle individual, a parameter combination of each dung beetle individual is determined, wherein the parameter combination is used to characterize the characteristic index weight combination of each dung beetle individual. Afterwards, all parameter combinations are substituted into a preset characteristic evaluation index function for calculation to obtain a comprehensive index value corresponding to each dung beetle individual. Then, the comprehensive index values ​​corresponding to all dung beetle individuals are sorted in descending order to obtain a feature sequence, wherein the parameter combination corresponding to the comprehensive index value with a lower sorting has a lower correlation. Finally, a feature set at a preset position is screened out in the feature sequence, thereby improving the accuracy of subsequent feature classification, and the feature set is input into a preset support vector machine model, and a fitness value is output, thereby being able to evaluate the contribution of each dung beetle individual to the optimization target. Then, all dung beetle individuals are screened according to the fitness value to determine multiple target dung beetle individuals, which can quickly locate potential search areas, reduce the amount of calculation in non-effective areas, and improve calculation efficiency.

[0128] Specifically, the expression of the chaotic sequence generated by improving the Circle mapping in the embodiment of the present application is as follows:

[0129]

[0130] Among them, x i represents the i-th chaotic sequence number, mod(a,b) represents the modulo operation of a over b.

[0131] The feature evaluation index function in the embodiment of the present application is expressed as follows:

[0132] J(w1,w2,w3)w1×Sp(x,y)+w2×Rob(F)+w3×I(x,y);

[0133] ∑ i w i =1(i=1,2,3);

[0134] Among them, J(w1,w2,w3) is the characteristic evaluation index function, which means the weighted linear superposition value after the characteristic parameters are calculated for each index, w i is the weight of each indicator, Sp(x, y) is the Spearman correlation indicator, Rob(F) is the robustness indicator, and I(x, y) is the mutual information indicator.

[0135] It is worth noting that when the target dung beetle individuals are generated by the improved Circle chaotic mapping strategy in the embodiment of the present application, the dung beetle individuals are randomly generated and relatively evenly distributed in the search space content. The preset positions in the embodiment of the present application can be set according to the needs of the user, for example, the top ten feature sets in the feature sequence are screened out, the top eight feature sets in the feature sequence are screened out, etc., and the embodiment of the present application does not make specific restrictions.

[0136] Reference Figure 5 , Figure 5 This is a specific flow chart of updating the position of a target dung beetle individual through a preset flight strategy and a preset inertia weight provided in an embodiment of the present application. The method includes but is not limited to step S501 and step S503.

[0137] Step S501, determining the individual type of the target dung beetle individual.

[0138] Step S502: When the individual type is a ball-rolling dung beetle, the ball-rolling position of the ball-rolling dung beetle is updated according to a preset flight strategy.

[0139] In steps S501 to S502 of some embodiments, in the process of updating the position of the target dung beetle individual through a preset flight strategy and a preset inertia weight, the individual type of the target dung beetle individual is first determined. Different types of dung beetle individuals need to update their positions according to different formulas. Specifically, when the individual type is a rolling dung beetle, the rolling position of the rolling dung beetle is updated according to the preset flight strategy, so that the dung beetle individual can perform a larger random movement after updating the position, thereby exploring a wider area, balancing the breadth and depth of the algorithm.

[0140] It is worth noting that the introduction of the Levy flight strategy will produce random movements with longer jumps and multiple sudden changes in direction, that is, alternating short-distance local searches and long-distance jumping global searches. This flight method can strengthen the neighborhood search near the optimal solution while detecting solutions farther away in the algorithm space.

[0141] It should be noted that the formula for updating the position of Levy flight is:

[0142]

[0143] Among them, α is the step size control factor, is a vector operation, Levy(λ) is the step size that follows the Levy distribution. Specifically, u=N(0,σ 2 u ), σ v =1, Γ is the Gamma function.

[0144] After the dung beetle optimization algorithm introduces this strategy, the dung beetle rolling ball position update formula is as follows:

[0145] x i (t+1)=x i (t)+α×k×x i (t-1)+Levy(λ)×b×|x i (t)-X w ∣;

[0146] x i (t+1)=x i (t)+tan(θ)×|x i (t)-x i (t-1)|×Levy(λ);

[0147] Among them, x i (t+1) is the t+1th iteration position of the i-th dung beetle; x i (t) is the current iteration position of the i-th dung beetle; x i (t-1) is the previous iteration position of the i-th dung beetle; α is a natural number assigned a value of -1 or 1; k∈(0,0.2], k is a deflection coefficient, and k in the embodiment of the present application is 0.1; b is a constant in the range of (0,1), and b in the embodiment of the present application is 0.3; X w is the global worst position.

[0148] Step S503: when the individual type is a thieving dung beetle, the theft position of the thieving dung beetle is updated based on a preset inertia weight.

[0149] In step S503 of some embodiments, when the individual type is a thieving dung beetle, the embodiments of the present application need to update the stealing position of the thieving dung beetle based on a preset inertia weight. Specifically, the embodiments of the present application should use a larger weight in the early stage of the algorithm iteration to allow the algorithm to quickly reach the vicinity of the target value with a larger step; and in the later stage of the iteration, a smaller weight should be used to allow the algorithm to move with a smaller step, so as to accurately search near the target solution and better local optimization.

[0150] It is worth noting that the inertia weight in the embodiment of the present application is an inertia weight based on chaos adaptation, which is specifically expressed as follows:

[0151]

[0152] Where t is the current iteration number; T max represents the maximum number of iterations; ω s represents the initial value of inertia weight; ω e Inertia weight representing the maximum number of iterations.

[0153] In some embodiments, the embodiments of the present application will combine the position update formula of the theft phase of DBO, and the theft phase update formula of IDBO is:

[0154] x i (t+1)=X b +S×g·(|x i (t)-X * ∣+∣x i (t)-ω·X b ∣);

[0155] Wherein, g is a random vector of size 1×D following normal distribution; S is a constant, and S in the embodiment of the present application is 0.5; X * is the local optimal solution; X b is the global optimal position of the dung beetle.

[0156] Reference Figure 6 , Figure 6 This is a specific flow chart for updating the reproduction and foraging behavior of a target dung beetle individual provided in an embodiment of the present application. The method includes but is not limited to step S601 and step S606.

[0157] It should be noted that the algorithm parameters in the embodiment of the present application include an upper bound of the search space and a lower bound of the search space.

[0158] Step S601, when the individual type is a breeding dung beetle or a foraging dung beetle, a safe area for breeding dung beetles or foraging dung beetles is simulated based on a preset boundary selection strategy, an upper bound of the search space, and a lower bound of the search space.

[0159] Step S602, updating the breeding behavior of the breeding dung beetles based on the safety area, and updating the foraging behavior of the foraging dung beetles.

[0160] In steps S601 to S602 of some embodiments, when the individual type is a breeding dung beetle or a foraging dung beetle, it is necessary to determine the area where the breeding dung beetles or the foraging dung beetles can lay eggs or forage. Specifically, the embodiments of the present application will utilize a boundary selection strategy, an upper bound of the search space, and a lower bound of the search space to simulate a safe area for breeding dung beetles or foraging dung beetles, thereby improving the safety and efficiency of the search, and maintaining population diversity and algorithm stability. Thereafter, the breeding behavior of the breeding dung beetles is updated based on the safe area, and the foraging behavior of the foraging dung beetles is updated, thereby improving the stability and reliability of the algorithm and avoiding invalid or infeasible solutions.

[0161] It can be understood that the boundary selection strategy in the embodiments of the present application includes but is not limited to a fixed boundary strategy, a reflective boundary strategy, a periodic boundary strategy, a shrinking boundary strategy, and the like.

[0162] Specifically, the embodiment of the present application uses a boundary selection strategy to simulate the safe area for dung beetles to lay eggs, and the formula is as follows:

[0163]

[0164] Among them, X * is the local optimal solution; R = 1-t / t max , t max is the maximum number of iterations, t is the current number of iterations, Ub is the upper bound of the search space, Lb is the lower bound of the search space, and Ub * is the upper bound of the safe area, Lb * The lower bound of the safe zone.

[0165] Furthermore, the position update formula for dung beetle reproduction is:

[0166] B i (t+1)=X * +b1(B i (t)-Lb * )+b2×(B i (t)-Ub * );

[0167] Among them, B i (t+1) is the t+1th selection position of the i-th hatching ball; B i (t) is the current iteration position of the i-th hatching ball; b1 and b2 are two D-dimensional independent random vectors.

[0168] Furthermore, the position update formula of the dung beetle foraging is:

[0169] x i (t+1)=x i (t)+C1×(x i (t)-Lb b )+C2·(B i (t)-Ub b );

[0170] Among them, C1 is a random number that obeys the normal distribution; C2 is a random vector in the range of (0,1).

[0171] Step S603, calculate the first position of the rolling ball position of the rolling dung beetle after the ball is updated, the second position of the stealing position of the thieving dung beetle after the stealing position is updated, the third position of the breeding behavior of the breeding dung beetle after the breeding behavior is updated, the fourth position of the foraging behavior of the foraging dung beetle after the foraging behavior is updated, and the number of position iterations of the target dung beetle individual.

[0172] Step S604, calculating the error rate value of the target dung beetle individual according to the first position, the second position, the third position and the fourth position.

[0173] In steps S603 to S604 of some embodiments, after updating the positions of dung beetles of different individual types, the embodiments of the present application will also calculate the positions of dung beetles of different individual types, specifically, the first position of the rolling position of the rolling dung beetle after the rolling ball is updated, the second position of the stealing position of the thieving dung beetle after the stealing position is updated, the third position of the breeding behavior of the breeding dung beetle after the breeding behavior is updated, and the fourth position of the foraging behavior of the foraging dung beetle after the foraging behavior is updated, so as to obtain the current iteration position of different types of dung beetles, and the embodiments of the present application will also record the number of position iterations of all target dung beetle individuals to facilitate whether to continue the position update operation later. Afterwards, the error rate value of the target dung beetle individual is calculated based on the first position, the second position, the third position and the fourth position, so as to obtain the accuracy rate of the updated positions of different types of dung beetles.

[0174] Specifically, the calculation formula of the error rate value is as follows:

[0175] Err=2-Acc train -Acc test ;

[0176] Among them, Acc train Indicates the accuracy of the training set, Acc test Represents the accuracy of the test set.

[0177] Step S605: When the error rate value is less than the preset global error rate value, the target fitness value and target position corresponding to the error rate value are recorded.

[0178] Step S606, generating an update result according to the target fitness value, the target position and the number of position iterations.

[0179] In steps S605 to S606 of some embodiments, after calculating the error rate value of the target dung beetle individual, the embodiments of the present application will also compare the error rate value with the preset global error rate value. When the error rate value is less than the preset global error rate value, it means that the accuracy corresponding to the current position of the dung beetle individual is relatively high. The target fitness value and target position corresponding to the error rate value can be recorded. Finally, an update result is generated based on the target fitness value, target position and number of position iterations to facilitate subsequent judgment on whether it is necessary to repeatedly update the position of the dung beetle.

[0180] It should be noted that when the error rate value is greater than or equal to the preset global error rate value, there is no need to replace the global error rate value.

[0181] Reference Figure 7 , Figure 7 It is a specific flow chart for determining the weight combination of feature indicators according to the update result provided by the embodiment of the present application. In some embodiments, the method includes but is not limited to step S701 and step S704.

[0182] Step S701, determining a target fitness value and a target position corresponding to an update result.

[0183] Step S702, determining a plurality of candidate indicator weight combinations according to the target fitness value and the target position, and sorting all the candidate indicator weight combinations in descending order to obtain a combination sequence.

[0184] Step S703, screen out the weight combination at the preset position in the combination sequence, input the weight combination into a preset support vector machine model, and output the classification accuracy.

[0185] Step S704, comparing the classification accuracy with the preset feature indicator weights to determine a feature indicator weight combination.

[0186] In steps S701 to S704 of some embodiments, in the process of determining the feature indicator weight combination according to the update result, first, the target fitness value and target position corresponding to the update result are determined, and then a plurality of candidate indicator weight combinations are determined according to the target fitness value and the target position, and all the candidate indicator weight combinations are sorted in descending order to obtain a combination sequence, and then the weight combination of the preset position is screened out in the combination sequence, and the weight combination is input into the preset support vector machine model, and the classification accuracy is output, and finally, the classification accuracy is compared with the preset feature indicator weight to determine the feature indicator weight combination. The embodiment of the present application evaluates the combination of each indicator weight by considering the error rate of the support vector machine model to obtain the optimal weight combination.

[0187] It is worth noting that the preset positions in the embodiments of the present application can be set according to the needs of the user, for example, screening out the top ten indicator weight combinations in the combination sequence, screening out the top eight indicator weight combinations in the combination sequence, etc. The embodiments of the present application do not make specific restrictions.

[0188] See also Figure 8 The embodiment of the present application further provides a tool feature screening system, which can implement the above tool feature screening method, and the system includes:

[0189] The data acquisition module 801 is used to acquire multiple original tool signals and perform preprocessing operations on the original tool signals to obtain a tool feature data set, wherein the original tool signals are obtained by collecting the tool in different directions;

[0190] An array creation module 802 is used to initialize the algorithm parameters of the preset improved dung beetle optimization algorithm and create a three-dimensional array, wherein the three-dimensional array is used to store combinations of different indicator weights;

[0191] The individual determination module 803 is used to initialize the position of the three-dimensional array based on a preset chaotic mapping algorithm and algorithm parameters to determine multiple target dung beetle individuals;

[0192] The position updating module 804 is used to update the position of the target dung beetle individual through a preset flight strategy and a preset inertia weight, and to update the reproduction and foraging behavior of the target dung beetle individual based on the algorithm parameters to obtain an updated result;

[0193] A combination determination module 805 is used to determine a characteristic index weight combination according to the update result when the update result satisfies the iteration threshold in the algorithm parameter;

[0194] The feature screening module 806 is used to screen the tool feature data set according to the feature index weight combination to obtain the target tool feature.

[0195] The tool feature screening system of the embodiment of the present application is used to execute the tool feature screening method in the above embodiment. Its specific processing process is the same as the tool feature screening method in the above embodiment, and will not be repeated here.

[0196] In order to specifically illustrate the tool feature screening method of the embodiment of the present application, a specific example is given below.

[0197] Example 1:

[0198] The traditional DBO (Dung Beetle Optimizer) algorithm is prone to insufficient population diversity, the algorithm is prone to falling into local optimal solutions, and the algorithm's exploration ability is insufficient. This example first uses the population generated by the improved Circle mapping to replace the randomly generated initial population in the original algorithm to achieve population diversity; secondly, the Levy flight strategy is combined to make the algorithm make larger jumps during the search, which helps avoid falling into local optimal solutions; finally, an adaptive variable inertia weight strategy is proposed to enhance the algorithm's exploration ability.

[0199] The specific improvement measures include the following:

[0200] Step 1: Preprocess the original tool signal, initialize the parameters of the IDBO algorithm, and create a three-dimensional array P containing the optimization index weights.

[0201] In some embodiments, the acceleration sensor can collect vibration signals in the X, Y and Z directions, and filter the vibration signals to obtain vibration signals with less interference. A tool feature data set is extracted based on the vibration signal, including multiple columns of features and a tool wear column.

[0202] In some embodiments, a three-dimensional array P[w1,w2,w3] is created, and the three-dimensional array is used to store the combination of different indicator weights, and this combination will be used as the search space in the optimization process. In addition, the embodiment of the present application will also initialize the population size of dung beetles N=50, the parameter dimension dim=3, and the maximum number of iterations t max =200, the search space upper bound Ub = [0,0,0] and lower bound Lb = [1,1,1].

[0203] Step 2: Use the improved Circle chaotic mapping strategy to initialize N individuals with relatively uniform position distribution in the search space, then iterate the N wavelet parameter combinations into the fitness expression to calculate the fitness value of the current dung beetle individual.

[0204] refer to Fig. 9 , Fig. 9 This is a comparison diagram before and after the improved Circle chaos mapping provided in an example of this application.

[0205] Specifically, we first use the strategy of improved Circle chaotic mapping to initialize N individuals with relatively uniform position distribution in the search space, and then substitute the N groups of feature index weight combinations into the feature evaluation index function J(w1,w2,w3). According to J(w1,w2,w3), we get a comprehensive index F. According to the value of F, we sort the features in descending order, set the feature screening threshold and screen out the top ten feature sets D2. D2 is input into the support vector machine model, and the lowest error rate Err is used as the optimal solution of the objective function.

[0206] It can be understood that the feature evaluation index function formula in this example is as follows:

[0207] J(w1,w2,w3)=w1×Sp(x,y)+w2×Rob(F)+w3×I(x,y);

[0208] ∑ i w i = l(i = 1, 2, 3);

[0209] Among them, J(w1,w2,w3) is the characteristic evaluation index function, which means the weighted linear superposition value after the characteristic parameters are calculated for each index, w i is the weight of each indicator, Sp(x, y) is the Spearman correlation indicator, Rob(F) is the robustness indicator, and I(x, y) is the mutual information indicator.

[0210] Specifically, the weight range of each weight indicator is shown in Table 1.

[0211] Table 1 Weight parameter range table

[0212]

[0213]

[0214] Furthermore, the formula for the minimum error rate Err is:

[0215] Err=2-Acc train -Acc test ;

[0216] Among them, Acc train Indicates the accuracy of the training set, Acc test Represents the accuracy of the test set.

[0217] Furthermore, this example also sets a global minimum error rate value E best , and compare the Err obtained each time with E best For comparison, if Err <Ebest t Then replace it.

[0218] Step 3: Use the Levy flight strategy to update the dung beetle rolling ball position according to the formulas in step S501 to step S502.

[0219] Specifically, each time a dung beetle individual is determined to be a ball-rolling dung beetle, first, according to the Levy flight strategy, the individual is made to perform a larger random movement after the position is updated, and then the ball-rolling position of the dung beetle is updated according to the formulas in step S501 to step S502.

[0220] Step 4: Update the dung beetle stealing position according to the formula in step S503 in combination with the inertia weight based on chaos adaptation.

[0221] Specifically, each time a dung beetle is determined to be a stealing dung beetle, the formula in step S503 is executed to update the weight formula, and the weight is adaptively adjusted in the iterative process, so that the algorithm has a larger weight in the early stage and a smaller weight in the later stage. Then, the dung beetle stealing position is updated according to the execution.

[0222] Step 5: Update the dung beetle reproduction and foraging behavior update formula according to steps S601 to S606.

[0223] Step 6: Determine whether the algorithm has reached the maximum number of iterations. If so, stop the calculation and output the optimal fitness value and optimal position; otherwise, repeat steps 1 to 5.

[0224] Specifically, the improved Circle chaos mapping, Levy flight strategy and chaotic adaptive inertia weight strategy are combined to update the position of the dung beetle according to the formulas of steps S501 to S502, S503, and S601 to S606, and calculate Err after the updated position, and compare it with E best Compare and record the optimal solution and optimal position.

[0225] Step 7: Find E based on the IDBO result best The corresponding combination of feature index weights is compared with the DBO algorithm and the traditional weight manual selection method.

[0226] In some embodiments, this example lists five combinations of feature index weight selections based on experience, and selects the top ten features based on the combination. The feature set of each combination is input into the SVM (Support Vector Machine) model to calculate the classification accuracy. Finally, it is compared with the feature index weights optimized by IDBO.

[0227] In some embodiments, the embodiments of the present application utilize the powerful search strategy of the IDBO algorithm to efficiently explore the solution space of the weights of each feature screening evaluation index function, and evaluate the combination of each index weight by considering the error rate of the SVM classification model to obtain the optimal weight combination.

[0228] Table 2 shows the weight values ​​obtained by the improved optimization algorithm IDBO in this embodiment, and Table 3 shows the accuracy of the feature set after manual screening.

[0229] Table 2 Optimal solution based on IDBO algorithm

[0230] <![CDATA[w1]]> <![CDATA[w2]]> <![CDATA[w3]]> Classification accuracy 0.14 0.19 0.67 92.04%

[0231] Table 3 Based on traditional manual weight selection method

[0232]

[0233]

[0234] It can be seen from Table 2 that the optimal solution of the feature evaluation index function parameters obtained by the improved optimization algorithm IDBO in the embodiment of the present application is J(0.14, 0.19, 0.67), and its classification accuracy is higher than that of the traditional manual weight selection method.

[0235] An embodiment of the present application further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor is used to execute the tool feature screening method in the above embodiment of the present application.

[0236] See also Fig.10 , Fig.10 The hardware structure of the electronic device provided in the embodiment of the present application is illustrated, and the electronic device includes:

[0237] The processor 1001 may be implemented by a general-purpose CPU (Central Processing Unit), 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 application;

[0238] The memory 1002 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 1002 can store an operating system and other application programs. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 1002, and the processor 1001 calls and executes the tool feature screening method of the embodiment of this application;

[0239] Input / output interface 1003, used to implement information input and output;

[0240] The communication interface 1004 is used to realize the communication interaction between the device and other devices. The communication can be realized through a wired manner (such as USB, network cable, etc.) or a wireless manner (such as mobile network, WIFI, Bluetooth, etc.);

[0241] A bus 1005 , which transmits information between various components of the device (e.g., the processor 1001 , the memory 1002 , the input / output interface 1003 , and the communication interface 1004 );

[0242] The processor 1001 , the memory 1002 , the input / output interface 1003 and the communication interface 1004 are connected to each other in communication within the device via the bus 1005 .

[0243] An embodiment of the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above-mentioned tool feature screening method is implemented.

[0244] 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.

[0245] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0246] It can be understood by those skilled in the art that Figure 1-10 The technical solutions shown in the figure do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than those shown in the figure, or a combination of certain steps, or different steps.

[0247] The system 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 present embodiment.

[0248] 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.

[0249] 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.

[0250] 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.

[0251] In the several embodiments provided in the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the division of the above 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 systems or units, which can be electrical, mechanical or other forms.

[0252] The units described above 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.

[0253] 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.

[0254] 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, 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 a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store programs.

[0255] The preferred embodiments of the present application are described above with reference to the accompanying drawings, but the scope of the rights of the present application is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present application should be within the scope of the rights of the present application.

Claims

1. A tool feature screening method, It is characterized in that The method comprises: Acquire multiple original tool signals, and perform preprocessing operations on the original tool signals to obtain a tool feature data set, wherein the original tool signals are obtained by collecting the tool in different directions; Initialize the algorithm parameters of the preset improved dung beetle optimization algorithm and create a three-dimensional array, wherein the three-dimensional array is used to store combinations of different indicator weights; Initializing the position of the three-dimensional array based on a preset chaotic mapping algorithm and the algorithm parameters to determine a plurality of target dung beetle individuals; The position of the target dung beetle individual is updated through a preset flight strategy and a preset inertia weight, and the reproduction and foraging behavior of the target dung beetle individual is updated based on the algorithm parameters to obtain an updated result; When the update result satisfies the iteration threshold in the algorithm parameter, determining a feature index weight combination according to the update result; The tool feature data set is screened according to the feature index weight combination to obtain target tool features.

2. The tool feature screening method according to claim 1, It is characterized in that The preprocessing operation is performed on the original tool signal to obtain a tool feature data set, including: Performing filtering processing on the original tool signal; Perform feature extraction on the tool signal after filtering to obtain multiple columns of features and tool wear columns; A tool feature data set is generated according to the plurality of feature columns and the tool wear column.

3. The tool feature screening method according to claim 1, It is characterized in that The step of creating a three-dimensional array comprises: A three-dimensional array is created according to preset index requirements, wherein the index requirements include a Spearman correlation index, a robustness index, and a mutual information index.

4. The tool feature screening method according to claim 1, It is characterized in that The algorithm parameters include the population size; the three-dimensional array is initialized based on the preset chaotic mapping algorithm and the algorithm parameters to determine a plurality of target dung beetle individuals, including: Initializing the position of the population size in the search space of the three-dimensional array by using a preset chaotic mapping algorithm to obtain a plurality of dung beetle individuals with uniform position distribution; For each of the dung beetle individuals, determining a parameter combination of the dung beetle individual; Performing index calculation on the parameter combination through a preset characteristic evaluation index function to obtain a comprehensive index value; Sort the parameter combinations in descending order according to the comprehensive index values ​​to obtain a feature sequence; Filtering a feature set at a preset position in the feature sequence, inputting the feature set into a preset support vector machine model, and outputting a fitness value; All the dung beetle individuals are screened according to the fitness values ​​to determine a plurality of target dung beetle individuals.

5. The tool feature screening method according to claim 1, It is characterized in that The updating of the position of the target dung beetle individual by using a preset flight strategy and a preset inertia weight includes: Determining the individual type of the target dung beetle individual; When the individual type is a ball-rolling dung beetle, updating the ball-rolling position of the ball-rolling dung beetle according to a preset flight strategy; When the individual type is a thieving dung beetle, the stealing position of the thieving dung beetle is updated based on a preset inertia weight.

6. The tool feature screening method according to claim 5, It is characterized in that The algorithm parameters include an upper bound of the search space and a lower bound of the search space; the reproduction and foraging behavior of the target dung beetle individual is updated based on the algorithm parameters to obtain an updated result, including: When the individual type is a breeding dung beetle or a foraging dung beetle, a safe area of ​​the breeding dung beetle or the foraging dung beetle is simulated based on a preset boundary selection strategy, the search space upper bound, and the search space lower bound; updating the breeding behavior of the breeding dung beetle based on the safety area, and updating the foraging behavior of the foraging dung beetle; Calculate the first position of the rolling ball position of the rolling dung beetle after being updated, the second position of the stealing position of the thieving dung beetle after being updated, the third position of the breeding behavior of the breeding dung beetle after being updated, the fourth position of the foraging behavior of the foraging dung beetle after being updated, and the number of iterations of the position of the target dung beetle individual; Calculating an error rate value of the target dung beetle individual according to the first position, the second position, the third position, and the fourth position; When the error rate value is less than a preset global error rate value, recording the target fitness value and target position corresponding to the error rate value; An update result is generated according to the target fitness value, the target position and the number of position iterations.

7. The tool feature screening method according to claim 6, It is characterized in that Determining the feature index weight combination according to the update result includes: Determining a target fitness value and a target position corresponding to the update result; Determine a plurality of candidate indicator weight combinations according to the target fitness value and the target position, and sort all the candidate indicator weight combinations in descending order to obtain a combination sequence; Selecting a weight combination at a preset position in the combination sequence, inputting the weight combination into a preset support vector machine model, and outputting a classification accuracy rate; The classification accuracy is compared with the preset feature indicator weights to determine the feature indicator weight combination.

8. A tool feature screening system, It is characterized in that The system comprises: A data acquisition module, used for acquiring a plurality of original tool signals and performing preprocessing operations on the original tool signals to obtain a tool feature data set, wherein the original tool signals are obtained by collecting the tool in different directions; An array creation module, used to initialize the algorithm parameters of the preset improved dung beetle optimization algorithm and create a three-dimensional array, wherein the three-dimensional array is used to store a combination of different indicator weights; An individual determination module is used to initialize the position of the three-dimensional array based on a preset chaotic mapping algorithm and the algorithm parameters to determine a plurality of target dung beetle individuals; A position updating module, used to update the position of the target dung beetle individual through a preset flight strategy and a preset inertia weight, and to update the reproduction and foraging behavior of the target dung beetle individual based on the algorithm parameters to obtain an update result; A combination determination module, configured to determine a feature index weight combination according to the update result when the update result satisfies the iteration threshold in the algorithm parameter; The feature screening module is used to screen the tool feature data set according to the feature index weight combination to obtain the target tool feature.

9. An electronic device, It is characterized in that The electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the tool feature screening method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, It is 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 tool feature screening method according to any one of claims 1 to 7.