Intelligent tool selection method based on k neighbor entropy estimation and Kriging method

Abstract

The invention discloses an intelligent tool selection method based on k neighbor entropy estimation and a Kriging method, and the method comprises the following steps: S1, formulating a cutting tool pre-screening method based on rules, and obtaining an alternative tool set for machining; s2, performing tool performance reasoning and tool sorting on the basis of the alternative tool set obtained by pre-screening in S1; and S3, in combination with analysis of cutter and user behaviors, obtaining a cutting tool comprehensive recommendation method with process scene adaptability. According to the intelligent cutter selection method based on the k neighbor entropy estimation and the Kriging method, cutter selection experience in cutting machining records is mined, so that a cutting cutter comprehensive recommendation method with certain process scene adaptability is obtained; according to the method, when a user formulates a process manual, a predicted machining result, possible behaviors of the user and possible combinations of cutters can be given according to input of the user, and intelligent selection of the cutters is achieved.

Description

Technical Field

[0001] The invention relates to the technical field of tool selection, and in particular to a tool intelligent selection method based on k-nearest neighbor entropy estimation and Kriging method. Background Art

[0002] In the process of process formulation, the selected machining tools affect the surface quality of the machined parts and the cost of machining. In most cases, the current selection of cutting tools is based on the experience of workers or experts and the relevant tool manuals. This is subjective and will also lead to some experience loss as personnel turnover. In recent years, the intelligent selection of cutting tools has received more and more attention, and there is a tendency to use algorithms to achieve intelligent tool recommendations. The tool recommendation algorithm refers to the process of obtaining a more suitable tool from a larger number of alternative tools through certain judgment criteria during the process of process formulation.

[0003] The existing tool recommendation algorithms are usually designed from two perspectives: one is based on case reasoning, and the other is based on rule reasoning. The former lacks certain constraints and is prone to inappropriate tool recommendations. As the case library continues to expand, it is impossible to establish a meaningful index to speed up the search of cases. The latter is based on rule reasoning and is too dependent on the experience of process personnel, requiring continuous maintenance and improvement by personnel. The data fusion algorithm utilizes the current rapidly developing intelligent technology, combines multiple aspects such as tools and users, and complements each other to achieve more intelligent tool recommendations. Summary of the invention

[0004] The purpose of the present invention is to provide a method for intelligent tool selection based on k-nearest neighbor entropy estimation and Kriging method, which can give predicted processing results, possible user behaviors and possible tool combinations according to user input when the user formulates a process manual, thereby realizing intelligent selection of cutting tools.

[0005] To achieve the above object, the present invention provides a method for intelligent tool selection based on k-nearest neighbor entropy estimation and Kriging method, comprising the following steps:

[0006] S1. Develop a rule-based cutting tool pre-screening method to obtain a set of candidate tools for machining;

[0007] S2, performing tool performance reasoning and tool sorting based on the candidate tool set pre-screened in S1;

[0008] S3. Combined with the analysis of tools and user behaviors, a comprehensive recommendation method for cutting tools with the ability to adapt to process scenarios is obtained.

[0009] Preferably, the specific process of S1 is:

[0010] The user inputs the processing conditions into the inference engine, which combines expert knowledge to search for suitable tools in the tool database. The processing conditions include machine tool materials and process information.

[0011] Pre-screening is performed based on the feature requirements input by the user, which include tool parameters, tool materials and machining accuracy;

[0012] Most of the unsuitable tools are screened out based on their functions, topological relationships and geometric shapes, and after several layers of filtering, a pre-screened set of candidate tools is obtained.

[0013] Preferably, in S2, tool performance reasoning includes analyzing the information gain ratio using k-nearest neighbor entropy estimation and inferring the tool using Kriging interpolation, and obtaining various evaluation indicators by analyzing the tool parameters; then the tools are sorted, and finally the sorting of the alternative tool set is obtained according to the comprehensive multidimensional indicators as a benchmark sorting.

[0014] Preferably, in S2, the information gain ratio is used to calculate the weight of the influence of the tool parameters on the processing results by using the k-nearest neighbor entropy estimation analysis, including the following two parts:

[0015] First, the k-nearest neighbor entropy estimates the entropy of the multidimensional evaluation index in a tool set. The input of this method is the tool parameters of N processing records and the corresponding processing result evaluation index. In the tool application scenario, the evaluation of whether the tool selection is reasonable is usually analyzed from three perspectives: the surface roughness of the workpiece being processed, the wear of the processing tool, and the processing cost. Therefore, the k-nearest neighbor entropy estimation is calculated by the following formula:

[0016]

[0017] Among them, H(x) represents the entropy of the three-dimensional vector of processing quality, wear and cost; Γ is the gamma function, ψ is the derivative of the logarithm of the gamma function, N is the number of samples, the number of k nearest neighbors selected, D is the dimension of x, ε i For x i The Euclidean distance to the kth neighbor;

[0018] Secondly, the information gain ratio analyzes the correlation between tool parameters and indicators through the change of entropy and analyzes the importance of tool parameters;

[0019] For the tool vector with m parameters and its corresponding D-dimensional evaluation index, a parameter is selected to divide the input set into z sets, and the data with the same value are divided into the same set. The information entropy H of the D-dimensional evaluation index in each set is calculated separately. j , then the information gain ratio of the tool parameters is: After obtaining the information gain ratio G of all parameters riAfter that, all ratios are normalized to obtain the tool parameter importance vector G = [G r1 ,G r2 ,…,G rm ];

[0020] For the continuous attributes in tool parameters, the dichotomy method is used to calculate the information gain of the continuous attributes. For the attributes with incomplete information in tool parameters, the mathematical expectation is used to replace the default value.

[0021] Preferably, in S2, a multidimensional space is formed between the tool having m parameters and its corresponding processing result evaluation index, and Kriging interpolation uses known points in the space to estimate unknown points, specifically:

[0022] Define the basic interpolation formula as where w i is the weight, y i is the value of a known point. Assuming that the spatial attributes are uniform, the space has the same variance σ everywhere. 2 , if unbiased estimation is satisfied, then: The weights are solved by:

[0023]

[0024] Among them, r ij The semivariogram function between point i and point j is fitted by a Gaussian function: Where r(0) = 0, c0 and c are fitting parameters, d represents the distance, and γ is the parameter of the Gaussian model. After solving all weights, the predicted value of the unobserved target in the index space can be obtained through the basic interpolation formula;

[0025] The performance prediction values ​​of all tools in the candidate tool set are obtained through Kriging interpolation, including roughness Ra, tool life and processing cost. According to the performance prediction values ​​of all tools in the candidate tool set, the tools are ranked and recommended.

[0026] Preferably, in S3, based on the benchmark ranking of the tools obtained in S2, a fusion algorithm based on process scenarios is used to generate a comprehensive ranking based on the benchmark ranking of the tools, including the following steps:

[0027] S31, using collaborative filtering method to call the user's usage records, calculate the similarity of tool reputation, and predict the user's preference prediction values ​​for different tools by calculating the similarity matrix to obtain the user preference ranking of the tool set;

[0028] S32, using the Pagerank method to analyze the use of tool combinations from the processing records, constructing the adjacency matrix of the tool association graph, and using the main eigenvector approximate solution obtained to obtain the tool combination ranking;

[0029] S33, summarizing the benchmark ranking, user preference ranking and tool combination ranking, and obtaining a comprehensive ranking result through weighted sum ranking.

[0030] Therefore, the present invention adopts the above-mentioned tool intelligent selection method based on k-nearest neighbor entropy estimation and Kriging method, and the beneficial effects are as follows:

[0031] (1) The rule-based cutting tool pre-screening method of the present invention performs screening according to multiple conditions such as machine tool materials, process information, tool parameters, material and processing accuracy, which can quickly eliminate a large number of unsuitable tools, reduce the amount of subsequent calculations, and improve the efficiency of the entire tool selection process.

[0032] (2) The present invention uses k-nearest neighbor entropy estimation to analyze the information gain ratio to determine the tool parameter weights, and then combines Kriging interpolation to infer tool performance. It can accurately predict the performance of the tool during processing, such as roughness Ra, tool life and processing cost, and perform benchmark sorting based on this, providing a scientific basis for the accurate selection of tools.

[0033] (3) The present invention considers user preferences and tool combination usage respectively through collaborative filtering method and Pagerank method to generate a comprehensive ranking result, so that tool recommendations are not only based on processing performance, but also in line with user usage habits and optimize tool combinations, thereby improving the overall coordination and adaptability of the processing process.

[0034] (4) The data fusion algorithm in the present invention utilizes the current rapidly developing intelligent technology, combines multiple aspects such as tools and users, and complements each other to achieve more intelligent tool recommendations.

[0035] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is an overall technical block diagram of an embodiment of a method for intelligently selecting a tool based on k-nearest neighbor entropy estimation and Kriging method of the present invention;

[0037] Figure 2 It is a logic diagram of a rule-based pre-screening method of an embodiment of a tool intelligent selection method based on k-nearest neighbor entropy estimation and Kriging method of the present invention;

[0038] Figure 3 It is a tool performance reasoning logic diagram of an embodiment of a tool intelligent selection method based on k-nearest neighbor entropy estimation and Kriging method of the present invention;

[0039] Figure 4 It is a tool benchmark sorting flow chart of an embodiment of a tool intelligent selection method based on k-nearest neighbor entropy estimation and Kriging method of the present invention;

[0040] Figure 5 It is a comprehensive sorting schematic diagram of an embodiment of a method for intelligent tool selection based on k-nearest neighbor entropy estimation and Kriging method of the present invention. DETAILED DESCRIPTION

[0041] The technical solution of the present invention is further described below through the accompanying drawings and embodiments.

[0042] Unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0043] like Figure 1 As shown, the technical framework of the tool intelligent selection method based on k-nearest neighbor entropy estimation and Kriging method of the present invention has a three-layer structure, including the following steps:

[0044] S1. Develop a rule-based cutting tool pre-screening method to obtain a set of candidate tools for machining;

[0045] S2, performing tool performance reasoning and tool sorting based on the candidate tool set pre-screened in S1;

[0046] S3. Combined with the analysis of tools and user behaviors, a comprehensive recommendation method for cutting tools with the ability to adapt to process scenarios is obtained.

[0047] In S1, the rule-based cutting tool pre-screening method is to pre-screen out the tools that are not suitable for the current machining features, and reduce the amount of calculation for tool optimization. The specific process is as follows:

[0048] The user inputs the processing conditions into the inference engine, which will then combine expert knowledge to search for suitable tools in the tool database. The processing conditions include machine tool materials and process information.

[0049] Pre-screening is performed based on the feature requirements input by the user, and the feature requirements include the tool parameters, tool materials, and machining accuracy mentioned above. This embodiment takes square shoulder milling as an example to illustrate the pre-screening method. Square shoulder milling belongs to non-rotating square shoulder processing, and generally has several main parameters such as width, length, cutting depth, angle, and fillet radius. In addition, material, machining accuracy, etc. will also affect the selection of tools. The logic of pre-screening based on the feature requirements input by the user is as follows: Figure 2 As shown, select the integral milling cutter and the replaceable cutter head milling cutter, query the integral milling cutter and replaceable cutter head data table, the logic is as follows:

[0050] Step 1: Determine the cutting diameter:

[0051] 1) If the cutting diameter is within the range of plane width*(0.4-0.6), if it is not within the range of (0.4-0.6), proceed to the next step of judgment;

[0052] 2) If the cutting diameter is less than the plane width*0.8, proceed to the next step;

[0053] Step 2: Determine the fillet radius:

[0054] 1) If the maximum fillet radius is equal to the minimum fillet radius, the fillet radius shall not be less than the maximum fillet radius;

[0055] 2) If the maximum fillet radius is not equal to the minimum fillet radius, the fillet radius is greater than the maximum fillet radius;

[0056] 3) If the maximum fillet radius and the minimum fillet radius are 0, determine whether the fillet radius range is (0.4-1.2). If so, proceed to the next step, otherwise there is no result.

[0057] Step 3: Relationship between processed materials and applicable processing materials

[0058] 1) If the material being processed belongs to the set of applicable processing materials, proceed to the next step of judgment; otherwise, there is no result.

[0059] 2) If the machining process is roughing, finishing or semi-finishing, then the tool can enter the pre-screening tool set; if it does not meet this requirement, there will be no result.

[0060] Most of the unsuitable tools are screened out based on their functions, topological relationships and geometric shapes, and after several layers of filtering, a pre-screened set of candidate tools is obtained.

[0061] like Figure 3As shown in Figure 2, in S2, tool performance reasoning includes two steps: using k-nearest neighbor entropy estimation to analyze the information gain ratio and using Kriging interpolation to infer the tool. By analyzing the tool parameters, various evaluation indicators are obtained; then the tools are sorted, and finally the ranking of the candidate tool set is obtained according to the comprehensive multidimensional indicators as the benchmark ranking, as shown in Figure 2. Figure 4 shown.

[0062] In S2, the following two key calculation methods are applied:

[0063] (1) The information gain ratio is used to calculate the weight of the influence of tool parameters on machining results by using k-nearest neighbor entropy estimation, which lays the foundation for the Kriging method. It mainly includes the following two parts:

[0064] First, the k-nearest neighbor entropy estimates the entropy of the multidimensional evaluation index in a tool set. The input of this method is the tool parameters of N processing records and the corresponding processing result evaluation index. In the tool application scenario, the evaluation of whether the tool selection is reasonable is usually analyzed from three perspectives: the surface roughness of the workpiece being processed, the wear of the processing tool, and the processing cost. Therefore, the k-nearest neighbor entropy estimation is calculated by the following formula:

[0065]

[0066] Among them, H(x) represents the entropy of the three-dimensional vector of processing quality, wear, and cost. Γ is the gamma function, ψ is the derivative of the logarithm of the gamma function, N is the number of samples, k is the number of nearest neighbors selected, D is the dimension of x, ε i For x i Euclidean distance to the kth neighbor.

[0067] Secondly, the information gain ratio analyzes the correlation between tool parameters and indicators through the change of entropy, and analyzes the importance of tool parameters. For the input tool vector with m parameters and its corresponding D-dimensional evaluation index, a parameter is selected to divide the input set into z, and the data with the same value are divided into the same set. The information entropy H of the D-dimensional evaluation index in each set is calculated separately. j , then the information gain ratio of the tool parameters is:

[0068]

[0069] After obtaining the information gain ratio G of all parameters ri After that, all ratios are normalized to obtain the tool parameter importance vector G = [G r1 ,G r2 ,…,G rm ]. For the continuous attributes in tool parameters, the dichotomy method is used to calculate the information gain of continuous attributes. For the attributes with incomplete information in tool parameters, mathematical expectations are used to replace the default values.

[0070] 2) Kriging interpolation inference tool

[0071] In S2, Kriging is an interpolation method for multidimensional space, which can model multidimensional space. A tool with m parameters and its corresponding processing result evaluation index constitute a multidimensional space. Kriging interpolation is to use known points in space to estimate unknown points. Therefore, it is possible to infer the approximate performance of the pre-screened candidate tools through existing examples.

[0072] Define the basic interpolation formula as where w i is the weight, y i is the value of the known point. Assuming that the spatial properties are uniform, the space has the same variance σ everywhere 2 , if unbiased estimation is satisfied, then: The weights are solved by:

[0073]

[0074] Among them, r ij The semivariogram function between point i and point j is fitted by a Gaussian function: Where r(0) = 0, c0 and c are fitting parameters, d represents the distance, and γ is the parameter of the Gaussian model. After solving all weights, the predicted value of the unobserved target in the indicator space can be obtained through the basic interpolation formula.

[0075] Kriging interpolation can be used to obtain the performance prediction values ​​of all tools in the candidate tool set, including roughness Ra, tool life, and processing cost, and then the tools can be ranked and recommended based on the performance prediction values ​​of all tools in the candidate tool set.

[0076] In S3, based on the baseline ranking of the tools obtained in S2, in order to better reflect the user's usage preferences and the combination of tools, a fusion algorithm based on process scenarios is used to generate a comprehensive ranking based on the baseline ranking of the tools, including the following steps:

[0077] S31. Use collaborative filtering method to call the user's usage records, calculate the similarity of tool reputation, predict the user's preference for different tools by calculating the similarity matrix, and obtain the user preference ranking of the tool set.

[0078] The formula for calculating the similarity of word-of-mouth between knives is:

[0079]

[0080] Among them, N a is the number of process personnel who have selected tool a, N b is the number of process personnel who have selected tool b.

[0081] After calculating the similarity of all existing tools in the data set, the similarity matrix W can be obtained: ij , based on the similarity matrix, the interest of a user user1 in tool q can be calculated as follows:

[0082] P user1c =∑ i∈N(user1)∩S(c,k) W ic R user1i ;

[0083] P user1q =∑ i∈N(user1)∩S(q,l) W iq R user1i ;

[0084] Among them, P user1q represents the predicted value of user1's preference for tool q, N(user1) represents the set of tools selected by user1, S(q,l) represents the l tools with the most similar reputation to tool q, and W iq is the similarity between tool i and tool q, obtained from the similarity matrix, R user1i represents user1’s preference for tool i. Based on the predicted values ​​of user preference for different tools obtained by this method, the user preference ranking of the tool set can be obtained.

[0085] S32. The Pagerank method is used to analyze the use of tool combinations from the processing records, and the adjacency matrix of the tool association graph is constructed. The main eigenvector approximate solution is used to obtain the tool combination ranking. The specific process is as follows:

[0086] First, construct the adjacency matrix of the tool relationship graph: For the tool set T to be selected after rule screening, assume there is a tool set of 5 tools: T = {t1, t2, t3, t4, t5}, these tools have all been used to process "material 1 and feature 1", and t5 has never been involved in processing. Among them, {t1, t3} have been processed together 3 times, {t1, t2} have been processed together 5 times, {t3, t4} have been processed together 4 times, {t2, t4} have been processed together 2 times, then the tool hyperlink matrix is:

[0087]

[0088] For the dangling node t5 (not connected to other tools), the corresponding rows are all filled to 1 / 5 to obtain HM′, and a random noise is further added to obtain the improved adjacency matrix: Among them, E represents the unit matrix, α is a variable scalar, generally 0.85-0.99, and n is the number of tools.

[0089] Since the actual tool selection set may contain hundreds of tools, directly solving high-order equations will consume a lot of computing resources. Therefore, the power method is used to solve the eigenvector of the main eigenvalue. The calculation method is as follows: (u) =π (u-1) HM", through continuous iteration, an approximate solution can be finally obtained, and the tool combination ranking can be obtained by using the approximate solution of the main eigenvector obtained.

[0090] S33, such as Figure 5 As shown in the figure, the benchmark ranking, user preference ranking and tool combination ranking are summarized and the comprehensive ranking result is obtained by weighted sum ranking:

[0091]

[0092] Among them, β i is the weight value, Score i is the ranking of the three sorting methods, β i =β0+pos, pos is the custom tendency, β0 is the adjustable initial value.

[0093] The ranking result comprehensively considers the predicted processing results, possible user behaviors, and possible tool combinations. It is a general and complete tool recommendation method.

[0094] Therefore, the present invention adopts the above-mentioned intelligent tool selection method based on k-nearest neighbor entropy estimation and Kriging method, and constructs a three-layer technical framework. Through a series of steps such as rule-based pre-screening, tool performance reasoning and sorting, and comprehensive sorting combined with process scenarios, it is possible to accurately screen and recommend the most suitable tool for a specific processing task from a large number of tools, effectively improving the intelligence level, accuracy and efficiency of tool selection, while taking into account user preferences and tool combination usage, and can provide a comprehensive, reasonable and efficient tool selection plan for cutting processing, significantly improving processing quality and economic benefits.

[0095] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for intelligent tool selection based on k-nearest neighbor entropy estimation and Kriging method, characterized in that: The following steps are involved: S1. Develop a rule-based cutting tool pre-screening method to obtain a set of candidate tools for machining; S2, performing tool performance reasoning and tool sorting based on the candidate tool set pre-screened in S1; S3. Combined with the analysis of tools and user behaviors, a comprehensive recommendation method for cutting tools with the ability to adapt to process scenarios is obtained.

2. The method for intelligent tool selection based on k-nearest neighbor entropy estimation and Kriging method according to claim 1 is characterized in that: The specific process of S1 is as follows: The user inputs the processing conditions into the inference engine, which combines expert knowledge to search for suitable tools in the tool database. The processing conditions include machine tool materials and process information. Pre-screening is performed based on the feature requirements input by the user, which include tool parameters, tool materials and machining accuracy; Most of the unsuitable tools are screened out based on their functions, topological relationships and geometric shapes, and after several layers of filtering, a pre-screened set of candidate tools is obtained.

3. The method for intelligent tool selection based on k-nearest neighbor entropy estimation and Kriging method according to claim 2 is characterized in that: In S2, tool performance reasoning includes analyzing the information gain ratio using k-nearest neighbor entropy estimation and inferring tools using Kriging interpolation. By analyzing the tool parameters, various evaluation indicators are obtained; then the tools are sorted, and finally the ranking of the alternative tool set is obtained according to the comprehensive multidimensional indicators as the benchmark ranking.

4. The method for intelligent tool selection based on k-nearest neighbor entropy estimation and Kriging method according to claim 3 is characterized in that: In S2, the information gain ratio is estimated and analyzed using the k-nearest neighbor entropy to calculate the weight of the tool parameters on the machining results, which includes the following two parts: First, the k-nearest neighbor entropy estimates the entropy of the multidimensional evaluation index in a tool set. The input of this method is the tool parameters of N processing records and the corresponding processing result evaluation index. In the tool application scenario, the evaluation of whether the tool selection is reasonable is usually analyzed from three perspectives: the surface roughness of the workpiece being processed, the wear of the processing tool, and the processing cost. Therefore, the k-nearest neighbor entropy estimation is calculated by the following formula: Among them, H(x) represents the entropy of the three-dimensional vector of processing quality, wear and cost; Γ is the gamma function, ψ is the derivative of the logarithm of the gamma function, N is the number of samples, the number of k nearest neighbors selected, D is the dimension of x, ε i For x i The Euclidean distance to the kth neighbor; Secondly, the information gain ratio analyzes the correlation between tool parameters and indicators through the change of entropy and analyzes the importance of tool parameters; For the tool vector with m parameters and its corresponding D-dimensional evaluation index, a parameter is selected to divide the input set into z sets, and the data with the same value are divided into the same set. The information entropy H of the D-dimensional evaluation index in each set is calculated separately. j , then the information gain ratio of the tool parameters is: After obtaining the information gain ratio G of all parameters ri After that, all ratios are normalized to obtain the tool parameter importance vector G = [G r1 ,G r2 ,…,G rm ]; For the continuous attributes in tool parameters, the dichotomy method is used to calculate the information gain of the continuous attributes. For the attributes with incomplete information in tool parameters, the mathematical expectation is used to replace the default value.

5. The method for intelligent tool selection based on k-nearest neighbor entropy estimation and Kriging method according to claim 4 is characterized in that: In S2, a multidimensional space is formed between the tool with m parameters and its corresponding processing result evaluation index. Kriging interpolation uses the known points in the space to estimate the unknown points, specifically: Define the basic interpolation formula as where w i is the weight, y i is the value of a known point. Assuming that the spatial attributes are uniform, the space has the same variance σ everywhere. 2 , if unbiased estimation is satisfied, then: The weights are solved by: Among them, r ij The semivariogram function between point i and point j is fitted by a Gaussian function: Where r(0) = 0, c0 and c are fitting parameters, d represents the distance, and γ is the parameter of the Gaussian model. After solving all weights, the predicted value of the unobserved target in the index space can be obtained through the basic interpolation formula; The performance prediction values ​​of all tools in the candidate tool set are obtained through Kriging interpolation, including roughness Ra, tool life and processing cost. According to the performance prediction values ​​of all tools in the candidate tool set, the tools are ranked and recommended.

6. The method for intelligent tool selection based on k-nearest neighbor entropy estimation and Kriging method according to claim 5 is characterized in that: In S3, based on the baseline ranking of the tools obtained in S2, a fusion algorithm based on process scenarios is used to generate a comprehensive ranking based on the baseline ranking of the tools, including the following steps: S31, using collaborative filtering method to call the user's usage records, calculate the similarity of tool reputation, and predict the user's preference prediction values ​​for different tools by calculating the similarity matrix to obtain the user preference ranking of the tool set; S32, using the Pagerank method to analyze the use of tool combinations from the processing records, constructing the adjacency matrix of the tool association graph, and using the main eigenvector approximate solution obtained to obtain the tool combination ranking; S33, summarizing the benchmark ranking, user preference ranking and tool combination ranking, and obtaining a comprehensive ranking result through weighted sum ranking.

Images