A method for evaluating the performance of CNC machine tool castings

Abstract

The present invention relates to the technical field of machine tool casting evaluation, and specifically to a method for evaluating the performance of CNC machine tool castings, the method comprising: performing preprocessing operations on machine tool casting production data, analyzing the distribution fluctuation of each data indicator, judging the amplitude of its change in the overall data, and obtaining the importance analysis results of the indicator. In the present invention, by performing in-depth preprocessing and comprehensive analysis on the machine tool casting production data, the method can more accurately capture and analyze the data changes of different material types and production batches, ensuring that the real impact of each factor in the evaluation process can be effectively reflected. Secondly, by re-adjusting the indicator weights and constructing an evaluation model based on similarity analysis, the ability to cope with data missing and quality fluctuations in the evaluation process is enhanced. Even in the case of unstable raw material supply or changes in the production process, the actual impact of these changes on casting performance can be effectively mapped.

Description

Technical Field

[0001] The invention relates to the technical field of machine tool casting evaluation, and in particular to a method for evaluating the performance of a numerically controlled machine tool casting. Background Art

[0002] The field of machine tool casting evaluation technology is a field that focuses on evaluating and improving the quality and performance of CNC machine tool castings. Machine tool castings are an important part of the structure of CNC machine tools, and their quality directly affects the accuracy, stability and service life of the machine tools. The field includes the entire process from material selection, design optimization, production process to final performance evaluation. Through a systematic evaluation of castings, potential problems in the manufacturing process can be identified, and then the production process can be optimized to improve product quality and performance.

[0003] Among them, the CNC machine tool casting performance evaluation method refers to a set of methods and processes designed to systematically evaluate the performance of CNC machine tool castings. The purpose is to ensure that the castings can meet the requirements of precision, strength and durability in design and application, and also seek to optimize the production process through evaluation results to improve the overall performance and quality of castings. Therefore, manufacturers can better understand the performance of castings in actual applications, predict their lifespan, and determine whether the target working conditions and performance standards are met.

[0004] Although existing technologies have achieved certain results in the performance evaluation of CNC machine tool castings, they have failed to fully capture the dynamic changes of all variables when dealing with complex and diverse production data. For example, when dealing with different types of materials or diverse production batches, traditional methods are not sufficient to accurately reflect the true impact of each factor on the final casting performance, resulting in a failure to fully consider the actual importance of certain key factors when optimizing the production process. In addition, they perform poorly in the face of missing data or large quality fluctuations. For example, in the casting production process, when there is an unstable supply of raw materials leading to fluctuations in casting properties, or there are unrecorded operational changes in the production process, it is difficult to accurately reflect the impact of the changes on casting performance, resulting in deviations between the evaluation results and the actual situation, which in turn limits the effective decision-making on production process adjustments and quality improvement measures to a certain extent. Summary of the invention

[0005] The present application provides a method for evaluating the performance of CNC machine tool castings, thereby solving the problem that although the existing technology has achieved certain results in the evaluation of the performance of CNC machine tool castings, it has failed to fully capture the dynamic changes of all variables when processing complex and diverse production data. For example, when processing different types of materials or various production batches, the traditional method is not sufficient to accurately reflect the true impact of each factor on the final casting performance, resulting in a failure to fully consider the actual importance of certain key factors when optimizing the production process. In addition, it performs poorly in the face of missing data or large quality fluctuations. For example, in the casting production process, when there is an unstable supply of raw materials that causes fluctuations in casting properties, or there are unrecorded operational changes in the production process, it is difficult to accurately reflect the impact of the changes on the casting performance, resulting in a deviation between the evaluation results and the actual situation, which in turn limits the effective decision-making on production process adjustments and quality improvement measures to a certain extent.

[0006] In view of the above problems, the present application provides a method for evaluating the performance of CNC machine tool castings.

[0007] The present application provides a method for evaluating the performance of a CNC machine tool casting, wherein the method comprises:

[0008] S1: Preprocess the machine tool casting production data, analyze the distribution fluctuation of each data indicator, judge its variation range in the overall data, and obtain the importance analysis results of the indicator;

[0009] S2: Based on the importance analysis results of the indicators, readjust the indicator weights to reflect the impact of the indicators on the quality of machine tool castings, and generate an indicator configuration with optimized weights;

[0010] S3: using the index configuration after weight optimization, calculating the similarity or difference between the parameters of the tool castings, and constructing a similarity analysis model;

[0011] S4: Analyze the correlation between machine tool casting parameters and performance through the similarity analysis model, identify key parameters, and obtain key parameters and their influence evaluation results;

[0012] S5: combining the weight-optimized index configuration and key parameters and their influence evaluation results, adjusting the structure and parameters of the similarity analysis model, mapping the current performance of the machine tool casting, and establishing an efficiency evaluation model;

[0013] S6: Apply the performance evaluation model to analyze the current machine tool casting production data, identify the key factors affecting casting performance, and formulate a casting production optimization evaluation plan based on the analysis results.

[0014] Preferably, the importance analysis results of the indicators include the distribution fluctuation score of the precision indicators of machine tool castings, the variation amplitude score of the surface roughness indicators and the contribution score of the mechanical property indicators. The indicator configuration after weight optimization includes the optimized weight values ​​of the precision indicators of machine tool castings, the optimized weight values ​​of the surface roughness indicators and the optimized weight values ​​of the mechanical property indicators. The similarity analysis model includes parameter difference evaluation calculated according to the optimized weights of machine tool castings, performance comparison indicators between castings and similarity scores. The key parameters and their influence evaluation results include the scores of precision parameters, key surface treatment parameters and mechanical property parameters that affect the performance of machine tool castings. The performance evaluation model includes model evaluation indicators, performance mapping and parameter influence measures adjusted for the performance of machine tool castings. The casting production optimization evaluation plan includes recommended process adjustment measures, raw material quality improvement plans and manufacturing parameter optimization for machine tool casting production.

[0015] Preferably, the step of preprocessing the machine tool casting production data, analyzing the distribution fluctuation of each data indicator, judging its variation range in the overall data, and obtaining the importance analysis result of the indicator also includes:

[0016] S101: Check the production data of machine tool castings including precision measurement values, surface roughness indexes and material strength parameters, identify and delete abnormal values, fill in missing data items, and obtain processed casting production data;

[0017] S102: using the processed casting production data, performing volatility and central trend analysis, calibrating the fluctuation range and center position of each machine tool casting production data indicator, and obtaining data distribution characteristic analysis results;

[0018] S103: According to the data distribution characteristic analysis result, determine the change range of each indicator, compare and analyze the fluctuation size of the indicator data, and obtain the importance analysis result of the indicator.

[0019] Preferably, based on the results of the importance analysis of the indicators, the indicator weights are readjusted to reflect the impact of the indicators on the quality of machine tool castings, and the step of generating the indicator configuration after weight optimization also includes:

[0020] S201: Arranging the importance analysis results of the indicators, referring to their contributions to the quality of machine tool castings, setting weights for each indicator, and obtaining an initialization weight table;

[0021] S202: According to the initialized weight table, by calculating the entropy weight method, increasing the weight ratio of the key performance indicators, reducing the weight distribution of the secondary performance indicators, and matching them with the quality contribution of the machine tool casting to obtain an adjusted weight distribution table;

[0022] S203: According to the adjusted weight allocation table, iteratively evaluate and adjust the weight ratio of each performance indicator in the machine tool casting production to obtain an indicator configuration after weight optimization.

[0023] Preferably, the entropy weight method is based on the improved formula I:

[0024] ;

[0025] Calculate the weight of each performance indicator, and adjust the weight of each performance indicator according to the calculation result to generate an adjusted weight distribution table;

[0026] in, Represents the improved The weight of the indicator, It is the improved The entropy value of an indicator, is the quality coefficient of the indicator, is the stability coefficient of the indicator, is the complexity coefficient of the indicator, is the weighted The sample in The probability distribution of the indicators, is the normalization constant, represents the sample size, is the number of indicators.

[0027] Preferably, the step of using the index configuration after weight optimization to calculate the similarity or difference between the parameters of the tool castings and constructing a similarity analysis model also includes:

[0028] S301: Select the index configuration after weight optimization, mark the performance index and weight of each machine tool casting, establish an analysis benchmark, and obtain an analysis benchmark index and weight list;

[0029] S302: using the analysis benchmark index and the weight list, selecting key performance indicators in the production of machine tool castings, sorting and comparing the performance data of each casting according to the weight list, and analyzing the difference in performance data between castings to generate a quantitative result of performance difference;

[0030] S303: Based on the performance difference quantification result, a similarity analysis model is obtained by mapping the similarity or difference of the performance of the machine tool casting.

[0031] Preferably, the steps of analyzing the correlation between machine tool casting parameters and performance through the similarity analysis model, identifying key parameters, and obtaining key parameters and their influence evaluation results also include:

[0032] S401: Based on the similarity analysis model, the Euclidean distance calculation method and the hierarchical analysis method are used to screen and classify the machine tool casting parameter data, and a parameter correlation matrix is ​​constructed based on the similarity or difference between the parameters;

[0033] S402: Based on the parameter correlation matrix, use matrix operations to quantify the correlation strength between each parameter and the performance of the machine tool casting, obtain the impact value of each parameter on the performance, and obtain a parameter impact value list;

[0034] S403: Based on the parameter influence value list, key parameters affecting the performance of machine tool castings are screened by setting thresholds, and the parameters and their corresponding influence values ​​are summarized and sorted as key indicators to obtain key parameter and influence evaluation results.

[0035] Preferably, the Euclidean distance calculation method is based on the improved formula II:

[0036] ;

[0037] Screen and classify the machine tool casting parameter data and construct the parameter correlation matrix;

[0038] in, is the improved Euclidean distance, Represents the precision parameter vector of machine tool castings, Represents the surface roughness parameter vector of machine tool castings, Represents the material strength parameter vector of machine tool castings, represents the ambient temperature parameter vector, represents the weight value of the i-th parameter, is the total number of parameters.

[0039] Preferably, the step of adjusting the structure and parameters of the similarity analysis model in combination with the weight-optimized index configuration and key parameters and their influence evaluation results, mapping the current situation of the machine tool casting performance, and establishing the performance evaluation model also includes:

[0040] S501: Based on the weight-optimized indicator configuration and key parameters and their influence evaluation results, adjust the input layer configuration of the similarity analysis model, adjust the data interface, directly input indicators and parameters, and generate a model input layer adjustment plan;

[0041] S502: Based on the model input layer adjustment plan, adjust the data processing flow and internal data path of the model, optimize the data flow mode, and generate a structure optimization adjustment result;

[0042] S503: Apply the structural optimization adjustment results to conduct a comprehensive test on the model to evaluate its effect in mapping the performance of machine tool castings, and make adjustments based on the test feedback to establish a performance evaluation model.

[0043] Preferably, the step of applying the performance evaluation model to analyze the current machine tool casting production data, identifying the key factors affecting the casting performance, and formulating a casting production optimization evaluation plan based on the analysis results also includes:

[0044] S601: using the performance evaluation model to analyze current machine tool casting production data, processing the data through the model, identifying key factors affecting casting performance, including differences in material quality or machining accuracy, and generating an initial key factor list;

[0045] S602: According to the initial key factor list, evaluate the influence of each factor on the performance of the machine tool casting, determine the factor with the greatest influence through comparative analysis, and generate an influence evaluation result;

[0046] S603: Based on the impact assessment results, consider the feasibility and benefits of the measures, design multiple optimization measures including adjusting key parameters, improving process flow or optimizing material use, and establish a casting production optimization assessment plan.

[0047] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0048] Through in-depth preprocessing and comprehensive analysis of machine tool casting production data, the method can more accurately capture and analyze data changes of different material types and production batches, ensuring that the real impact of various factors in the evaluation process can be effectively reflected. Especially in diverse and complex production environments, the adaptability and accuracy of the evaluation model are improved. Secondly, by re-adjusting the indicator weights and constructing an evaluation model based on similarity analysis, the ability to cope with data missing and quality fluctuations in the evaluation process is enhanced. Even in the case of unstable raw material supply or changes in the production process, the actual impact of these changes on casting performance can be effectively mapped. This makes the evaluation results more accurate and reliable, providing more effective decision support for production process adjustments and quality improvements.

[0049] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 The present invention provides a schematic diagram of the overall process of a method for evaluating the performance of a CNC machine tool casting;

[0051] Figure 2 The present invention proposes a specific flow chart of S1 of a method for evaluating the performance of CNC machine tool castings;

[0052] Figure 3 The present invention proposes a specific flow chart of S2 of a method for evaluating the performance of CNC machine tool castings;

[0053] Figure 4 The present invention proposes a specific flow chart of S3 of a method for evaluating the performance of CNC machine tool castings;

[0054] Figure 5 The present invention proposes a specific flow chart of S4 of a method for evaluating the performance of CNC machine tool castings;

[0055] Figure 6 The present invention proposes a specific flow chart of S5 of a method for evaluating the performance of CNC machine tool castings;

[0056] Figure 7 The present invention proposes a specific flow chart of S6 of a method for evaluating the performance of CNC machine tool castings. DETAILED DESCRIPTION

[0057] The present application provides a method for evaluating the performance of CNC machine tool castings.

[0058] Application Overview

[0059] There are problems in the prior art. Although the prior art has achieved certain results in the performance evaluation of CNC machine tool castings, it fails to fully capture the dynamic changes of all variables when processing complex and diverse production data. For example, when processing different types of materials or various production batches, traditional methods are not sufficient to accurately reflect the true impact of each factor on the final casting performance, resulting in failure to fully consider the actual importance of certain key factors when optimizing the production process. In addition, it performs poorly in the face of missing data or large quality fluctuations. For example, in the casting production process, when there is an unstable supply of raw materials leading to fluctuations in casting properties, or there are unrecorded operational changes in the production process, it is difficult to accurately reflect the impact of the changes on the casting performance, resulting in deviations between the evaluation results and the actual situation, which in turn limits the effective decision-making on production process adjustments and quality improvement measures to a certain extent.

[0060] In response to the above technical problems, the overall idea of ​​the technical solution provided by this application is as follows:

[0061] like Figure 1As shown, the present application provides a method for evaluating the performance of CNC machine tool castings, wherein the method comprises: S1: performing preprocessing operations on the production data of machine tool castings, analyzing the distribution fluctuation of each data indicator, judging its variation range in the overall data, and obtaining the importance analysis result of the indicator;

[0062] S2: Based on the results of the importance analysis of the indicators, readjust the indicator weights to reflect the impact of the indicators on the quality of machine tool castings and generate the indicator configuration after weight optimization;

[0063] S3: Use the index configuration after weight optimization to calculate the similarity or difference between the parameters of the tool castings and build a similarity analysis model;

[0064] S4: Analyze the correlation between machine tool casting parameters and performance through similarity analysis model, identify key parameters, and obtain key parameters and their influence evaluation results;

[0065] S5: Combine the index configuration after weight optimization and the key parameters and their influence evaluation results, adjust the structure and parameters of the similarity analysis model, map the current situation of machine tool casting performance, and establish an efficiency evaluation model;

[0066] S6: Use the performance evaluation model to analyze the current machine tool casting production data, identify the key factors affecting casting performance, and develop a casting production optimization evaluation plan based on the analysis results.

[0067] The results of the importance analysis of indicators include the distribution fluctuation score of the precision indicators of machine tool castings, the variation range score of the surface roughness indicators and the contribution score of the mechanical performance indicators. The indicator configuration after weight optimization includes the optimized weight values ​​of the precision indicators of machine tool castings, the optimized weight values ​​of the surface roughness indicators and the optimized weight values ​​of the mechanical performance indicators. The similarity analysis model includes parameter difference evaluation calculated according to the optimized weights of machine tool castings, performance comparison indicators and similarity scores between castings. The evaluation results of key parameters and their influence include the scores of precision parameters, key surface treatment parameters and mechanical performance parameters that affect the effectiveness of machine tool castings. The effectiveness evaluation model includes model evaluation indicators, performance mapping and parameter influence measures adjusted for the performance of machine tool castings. The casting production optimization evaluation plan includes recommended process adjustment measures, raw material quality improvement plans and manufacturing parameter optimization for machine tool casting production.

[0068] like Figure 2 As shown, the steps of preprocessing the machine tool casting production data, analyzing the distribution fluctuation of each data indicator, judging its variation range in the overall data, and obtaining the importance analysis results of the indicators also include:

[0069] S101: Check the production data of machine tool castings including precision measurement values, surface roughness indexes and material strength parameters, identify and delete abnormal values, fill in missing data items, and obtain the processed casting production data. The specific process is as follows;

[0070] Sub-step S101 uses data preprocessing technology to perform data cleaning operations based on the production data of machine tool castings. It uses the DataFrame.dropna() method through Python's Pandas library to delete missing values, and uses the DataFrame.replace() method to identify and replace outliers. For example, precision measurement values, surface roughness indicators, and material strength parameters that exceed the normal production range are marked as abnormal and replaced or deleted. The DataFrame.fillna() method is used to fill missing data items. The identification criteria for outliers and missing values ​​are preset according to production standards. For example, precision measurement values ​​that deviate from the mean by more than 3 standard deviations are considered abnormal, and the processed casting production data is generated.

[0071] S102: using the processed casting production data, performing volatility and central trend analysis, calibrating the fluctuation range and center position of each machine tool casting production data indicator, and obtaining the specific process of the data distribution characteristic analysis result is as follows;

[0072] Sub-step S102 uses statistical analysis methods based on the processed casting production data to perform volatility and central tendency analysis. The standard deviation is calculated using the np.std() function of Python's NumPy library to evaluate volatility, and the mean is calculated using the np.mean() function to determine the central tendency. The fluctuation range and center position of each machine tool casting production data indicator are calibrated. The calculation logic of the fluctuation range and center position is preset according to statistical principles, such as using the mean ± standard deviation to define the fluctuation range, and generate data distribution feature analysis results.

[0073] S103: According to the data distribution characteristic analysis results, the change range of each indicator is determined, and the fluctuation size of the indicator data is compared and analyzed to obtain the specific process of the indicator importance analysis results;

[0074] Based on the analysis results of data distribution characteristics, sub-step S103 uses the variation range analysis method to perform importance analysis. The correlation between each indicator and the target performance is calculated using the stats.pearsonr() function of Python's SciPy library to determine the variation range of each indicator. The fluctuation size of the indicator data is compared and analyzed. The comparison standards of the variation range and fluctuation size are preset according to the correlation analysis. For example, the closer the absolute value of the correlation coefficient is to 1, the closer the correlation between the indicator and the target performance is. The importance analysis results of the indicators are generated.

[0075] like Figure 3As shown, based on the results of the importance analysis of the indicators, the indicator weights are readjusted to reflect the impact of the indicators on the quality of machine tool castings, and the steps of generating the indicator configuration after weight optimization also include:

[0076] S201: Arrange the importance analysis results of the indicators, refer to their contributions to the quality of machine tool castings, set weights for each indicator, and obtain the specific process of initializing the weight table as follows;

[0077] In sub-step S201, weights are initialized based on the results of the importance analysis of the indicators. A DataFrame is constructed using Python's Pandas library to list each indicator and its importance score. The DataFrame.assign() method is used to assign initial weights to each indicator based on its contribution. The basis for weight assignment is to convert the indicator's importance score into a weight percentage to ensure that the initial weight of each indicator reflects its relative contribution to the quality of machine tool castings and to generate an initialization weight table.

[0078] S202: According to the initialization weight table, by calculating the entropy weight method, the weight ratio of the key performance indicators is increased, the weight distribution of the secondary performance indicators is reduced, and it is matched with the quality contribution of the machine tool casting to obtain the adjusted weight distribution table. The specific process is as follows;

[0079] Sub-step S202 performs entropy weight calculation based on the initialized weight table, and uses the entropy() function in Python's SciPy library to calculate the entropy value of each indicator according to its distribution in the casting production data set. The weight ratio of key performance indicators is adjusted according to the calculation result of the entropy value, and the weight distribution of secondary performance indicators is reduced. The adjustment process uses the amount of information reflected by the entropy value to determine the increase or decrease of the weight, ensuring that the adjusted weight more accurately matches the relationship between the indicator and the contribution to the quality of machine tool castings, and generates an adjusted weight distribution table.

[0080] The entropy weight method is based on the improved formula I:

[0081] ;

[0082] Calculate the weight of each performance indicator, and adjust the weight of each performance indicator according to the calculation result to generate an adjusted weight distribution table;

[0083] in, Represents the improved The weight of an indicator reflects the relative importance of the indicator in the overall evaluation. It is the improved The entropy value of an indicator is used to quantify the amount of information and the uniformity of the distribution of the indicator. The smaller the amount of information (the larger the entropy value), the smaller the variability of the indicator and the lower its importance. It is the quality coefficient of the indicator, reflecting the directness and importance of the indicator on the quality of machine tool castings. By analyzing historical production data and quality control records and combining expert experience evaluation, the degree of direct impact of each indicator on the quality of machine tool castings is determined. It is the stability coefficient of the indicator, which indicates the variation range of the indicator under different production batches or conditions. The higher the stability, the larger the coefficient. By analyzing the variability of the same indicator in historical production batches, the fluctuation range of the indicator under different production conditions is calculated to measure its stability. It is the complexity coefficient of the indicator, taking into account the process complexity or cost required to achieve or maintain the indicator. The higher the complexity, the larger the coefficient value. It is determined by evaluating the process complexity, technical difficulty and cost required to achieve or maintain each indicator, using expert scoring or cost-benefit analysis methods. is the weighted The sample in The probability distribution of an indicator and the probability value after considering the actual distribution characteristics of the indicator. is the normalization constant, , used to ensure the normalization of entropy calculation. Represents the sample size, which is the total number of samples considered in the analysis. is the number of indicators, which refers to the total number of indicators considered in the evaluation process.

[0084] The improved formula I is implemented as follows:

[0085] First, the production data of machine tool castings is collected, including but not limited to accuracy measurement values, surface roughness, material strength and other indicators. Preprocessing includes cleaning the data, identifying and processing outliers, and filling in missing data.

[0086] The data of each indicator is normalized, and the probability distribution of each indicator in the casting production data set is calculated by converting the value of each indicator into its relative position or proportion in the total data set.

[0087] Based on the probability distribution of each indicator using Formula I Calculate its entropy value. The entropy value reflects the amount of information and the uniformity of distribution of each indicator in the casting production data set.

[0088] According to formula I, the entropy value of each indicator is comprehensively considered , quality coefficient , stability coefficient and complexity coefficient Recalculate the weight of each indicator .

[0089] The improved weight of each indicator is calculated It is organized into a table to reflect the contribution of each indicator to the performance evaluation of machine tool castings after considering quality, stability and complexity.

[0090] Finally, the adjusted weight distribution table is used to analyze the machine tool casting production data in the performance evaluation model to identify the key factors affecting casting performance, and accordingly, a casting production optimization evaluation plan is formulated.

[0091] S203: According to the adjusted weight allocation table, iteratively evaluate and adjust the weight ratio of each performance indicator in the machine tool casting production, and obtain the specific process of the indicator configuration after weight optimization;

[0092] In sub-step S203, based on the adjusted weight distribution table, iterative evaluation is performed to adjust the weights. The adjusted weight distribution table is iteratively updated using Python's Pandas library. In each iteration, the DataFrame.apply() method is used to adjust the weight ratio of each performance indicator based on the new analysis results of the production data. Dynamic adjustment logic is used in the iterative process to update the weights based on the latest evaluation of the impact of the indicators on casting quality to reflect the impact of changes in production data on weight distribution. The weight distribution is iterated until the weight distribution is stable, generating an indicator configuration with optimized weights.

[0093] like Figure 4 As shown, the steps of using the index configuration after weight optimization to calculate the similarity or difference between the parameters of the tool castings and constructing the similarity analysis model also include:

[0094] S301: Select the index configuration after weight optimization, mark the performance index and weight of each machine tool casting, establish the analysis benchmark, and obtain the specific process of the analysis benchmark index and weight list as follows;

[0095] In sub-step S301, based on the indicator configuration after weight optimization, an analysis benchmark is established. A DataFrame is constructed through Python's Pandas library to list each machine tool casting performance indicator and its corresponding optimization weight. The DataFrame.to_dict() method is used to convert the indicators and weights into a dictionary format for subsequent processing, ensuring that each performance indicator and its weight are clearly marked and easy to process by the algorithm to generate a list of analysis benchmark indicators and weights.

[0096] S302: using the analysis benchmark index and weight list, selecting key performance indicators in the production of machine tool castings, sorting and comparing the performance data of each casting according to the weight list, and analyzing the difference in performance data between castings, the specific process of generating quantitative results of performance differences is as follows;

[0097] Sub-step S302 performs sorting and comparison of performance data based on the analysis of benchmark indicators and weight lists, uses Python's NumPy library for weighted calculations, and uses the numpy.dot() function to multiply the performance data of each casting by the weight list to calculate the weighted performance value. This process evaluates the differences between castings based on the weighted performance data, sorts them by the calculated weighted performance values, compares the performance differences between different castings, and generates quantitative results of performance differences.

[0098] S303: Based on the performance difference quantification result, the specific process of obtaining the similarity analysis model by mapping the similarity or difference of the performance of the machine tool casting is as follows;

[0099] Sub-step S303 builds a similarity analysis model based on the quantified results of performance differences. The pairwise_distances function in Python's Scikit-learn library is used to input weighted performance data and calculate the Euclidean distance between castings. This function calculates the distance between each pair of castings based on the input performance data. The smaller the distance, the higher the similarity between the castings. The similarity or difference of casting performance is mapped to a specific value to generate a similarity analysis model.

[0100] like Figure 5 As shown, the steps of analyzing the correlation between machine tool casting parameters and performance through a similarity analysis model, identifying key parameters, and obtaining key parameters and their influence evaluation results also include:

[0101] S401: Based on the similarity analysis model, the Euclidean distance calculation method and the hierarchical analysis method are used to screen and classify the machine tool casting parameter data, and the specific process of constructing the parameter correlation matrix based on the similarity or difference between the parameters is as follows;

[0102] In sub-step S401, based on the similarity analysis model, the machine tool casting parameter data is screened and classified. The similarity or difference between parameters is calculated through Python's NumPy library combined with the Euclidean distance calculation method. The improved Euclidean distance calculation takes into account the indicator configuration after weight optimization. The distance between weighted parameters is calculated through the numpy.linalg.norm() function. This process establishes a similarity analysis benchmark based on weight adjustment, and uses the hierarchical analysis method to redistribute the weights of key performance indicators through Python's ahp library to construct a parameter correlation matrix. The matrix quantifies the parameter correlation strength based on the similarity or difference between parameters and generates a parameter correlation matrix.

[0103] The Euclidean distance calculation method is based on the improved formula II:

[0104] ;

[0105] Screen and classify the machine tool casting parameter data and construct the parameter correlation matrix;

[0106] in, It is the improved Euclidean distance, which is used to measure the difference between two groups of machine tool casting parameters. Represents the precision parameter vector of machine tool castings, where each element The accuracy values ​​corresponding to different castings or different measuring points. Represents the surface roughness parameter vector of machine tool castings, where each element Corresponding to the surface roughness values ​​of different castings or different measuring points. Represents the material strength parameter vector of the machine tool casting, where each element Corresponding to material strength values ​​for different castings or different tests. Represents the ambient temperature parameter vector, where each element Corresponding to the ambient temperature values ​​under different castings or different conditions. It represents the weight value of the i-th parameter, which is calculated by the AHP method and reflects the relative importance of this parameter in the quality assessment of machine tool castings. is the total number of parameters, indicating the number of parameter dimensions considered when calculating the improved Euclidean distance.

[0107] The improved formula II is implemented as follows:

[0108] Determine and collect parameter data related to machine tool casting performance, including accuracy, surface roughness, material strength, ambient temperature, etc.

[0109] Through the analytic hierarchy process, a judgment matrix was constructed and pairwise comparisons were performed to determine the importance weight of each parameter relative to the casting performance.

[0110] Calculate the weight of each parameter, solve the maximum eigenvalue of the judgment matrix and its corresponding eigenvector, and obtain the weight after normalization. At the same time, perform a consistency check to ensure that the CR value is less than 0.1 to ensure the consistency of the judgment. If the consistency is insufficient, the judgment matrix needs to be adjusted.

[0111] Applying the determined weights, Formula II is calculated for each pair of machine tool castings, and the improved Euclidean distance value is used to compare and analyze the casting performance and identify castings with similar or significantly different performance.

[0112] S402: Based on the parameter correlation matrix, matrix operations are used to quantify the correlation strength between each parameter and the performance of the machine tool casting, and the influence value of each parameter on the performance is obtained. The specific process of obtaining the parameter influence value list is as follows;

[0113] Sub-step S402 performs matrix operations based on the parameter correlation matrix, using Python's NumPy library for operations. The numpy.matmul() function performs matrix multiplication on the parameter correlation matrix and the index configuration after weight optimization to quantify the correlation strength between each parameter and the performance of the machine tool casting. During the calculation process, the influence value of each parameter is obtained by multiplying the corresponding element in the correlation matrix by its weight, which reflects the degree of influence of the parameter on the performance and generates a list of parameter influence values.

[0114] S403: Based on the parameter influence value list, the key parameters affecting the performance of the machine tool casting are screened by setting thresholds, and the parameters and their corresponding influence values ​​are summarized and sorted as key indicators. The specific process of obtaining the key parameters and their influence evaluation results is as follows;

[0115] In sub-step S403, based on the parameter influence value list, the threshold screening method is used to identify the key parameters that affect the performance of machine tool castings. The screening threshold is set through the Python script, and the numpy.where() function is used to filter out the parameters with influence values ​​higher than the threshold. The parameters and their corresponding influence values ​​are summarized and sorted as key indicators. The setting of the screening threshold is based on the prior knowledge and historical data analysis of the performance impact in the casting production process to ensure that only key parameters are selected and the evaluation results of key parameters and their influence are obtained.

[0116] like Figure 6 As shown, combining the index configuration after weight optimization and the key parameters and their influence evaluation results, adjusting the structure and parameters of the similarity analysis model, mapping the current situation of the machine tool casting performance, and establishing the performance evaluation model also include:

[0117] S501: Based on the weight-optimized indicator configuration and key parameters and their influence evaluation results, the input layer configuration of the similarity analysis model is adjusted, the data interface is adjusted, and the indicators and parameters are directly inputted. The specific process of generating the model input layer adjustment plan is as follows;

[0118] In sub-step S501, based on the indicator configuration and key parameters after weight optimization and their influence evaluation results, the input layer configuration of the similarity analysis model is adjusted. The indicator configuration and key parameter data are read using the Pandas library through a Python script and used as the direct input of the model. The data frame is merged using the DataFrame.merge() method to establish a new structure of the data interface. The basis for adjusting the data interface is to ensure that all important performance indicators and key parameters can be effectively identified and processed by the model, and to generate a model input layer adjustment plan.

[0119] S502: Based on the model input layer adjustment plan, the data processing flow and internal data path of the model are adjusted to optimize the data flow mode, and the specific process of generating the structural optimization adjustment result is as follows;

[0120] The S502 sub-step adjusts the model's data processing flow and internal data paths based on the model input layer adjustment plan, optimizes the model's data preprocessing steps using Python's Scikit-learn library, and connects multiple data processing steps in series through the Pipeline function, including data standardization and dimensionality reduction. It uses the StandardScaler() and PCA() methods to optimize the data flow method, simplifies the internal data transmission path through the serialized processing flow, improves processing efficiency, and generates structural optimization and adjustment results.

[0121] S503: Apply the structural optimization adjustment results, conduct comprehensive tests on the model, evaluate its effect in mapping the performance of machine tool castings, and make adjustments based on the test feedback. The specific process of establishing the performance evaluation model is as follows;

[0122] The S503 sub-step applies the structural optimization adjustment results to conduct a comprehensive test on the model. The cross_val_score function in Python's Scikit-learn library is used for cross-validation to evaluate the effectiveness of the model in mapping the performance of machine tool castings. During the test, the model is comprehensively evaluated by setting different test parameters and conditions, such as fold selection and scoring index setting. According to the test feedback, GridSearchCV is used to tune the parameters to ensure that the performance of the model is optimal and establish an efficiency evaluation model.

[0123] like Figure 7 As shown in the figure, the performance evaluation model is used to analyze the current machine tool casting production data, identify the key factors affecting casting performance, and formulate the casting production optimization evaluation plan based on the analysis results, which also includes:

[0124] S601: Use the performance evaluation model to analyze the current machine tool casting production data, process the data through the model, identify the key factors affecting the casting performance, including differences in material quality or processing accuracy, and generate the initial key factor list.

[0125] In sub-step S601, based on the performance evaluation model, the current machine tool casting production data is analyzed. The performance evaluation model is loaded using the Scikit-learn library through a Python script, and the model.predict() method is used to process the production data set to identify the key factors affecting casting performance. This includes taking production data as input, calculating the weights of factors affecting performance based on the model training results, such as material quality or machining accuracy differences, and screening key factors based on the weights to generate an initial list of key factors.

[0126] S602: Based on the initial key factor list, evaluate the impact of each factor on the performance of machine tool castings, determine the factor with the greatest impact through comparative analysis, and generate the specific process of the impact assessment result as follows;

[0127] Sub-step S602 performs an impact assessment based on the initial key factor list, calculates the impact of each factor using the NumPy library through a Python script, calculates the absolute size of the factor impact value and sorts these values ​​through the numpy.abs() and numpy.argsort() functions, compares the factor impact values, determines the factor with the greatest impact on casting performance, and generates an impact assessment result.

[0128] S603: Based on the impact assessment results, consider the feasibility and benefits of the measures, design multiple optimization measures including adjusting key parameters, improving process flow or optimizing material use, and establish the specific process of casting production optimization assessment plan;

[0129] In sub-step S603, optimization measures are designed based on the impact assessment results. A series of optimization measures are designed by evaluating the feasibility and benefits of the measures, including adjusting key parameters, improving process flow or optimizing material usage. The decision tree algorithm is used to select the optimization plan. The decision tree model is implemented through Python scripts and Scikit-learn library. The impact assessment results are used as input, and the model outputs the optimization measure plan. Factors including cost, implementation difficulty and expected benefits are considered to establish a casting production optimization evaluation plan.

[0130] Although the present application has been described in conjunction with specific features and embodiments thereof, it is obvious that various modifications and combinations may be made thereto without departing from the spirit and scope of the present application. Accordingly, this specification and the drawings are merely exemplary illustrations of the present application as defined by the appended claims, and are deemed to have covered any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include these modifications and variations.

Claims

1. A method for evaluating the performance of CNC machine tool castings, characterized in that: The method comprises: S1: Preprocess the machine tool casting production data, analyze the distribution fluctuation of each data indicator, and judge its variation range in the overall data to obtain the importance analysis results of the indicators, including: Check the production data of machine tool castings including precision measurement values, surface roughness index and material strength parameters, identify and delete abnormal values, fill in missing data items, and obtain processed casting production data; Using the processed casting production data, performing volatility and central trend analysis, calibrating the fluctuation range and center position of each machine tool casting production data indicator, and obtaining data distribution characteristic analysis results; According to the data distribution characteristics analysis results, determine the change range of each indicator, compare and analyze the fluctuation size of the indicator data, and obtain the importance analysis results of the indicator; S2: Based on the importance analysis results of the indicators, readjust the indicator weights to reflect the impact of the indicators on the quality of machine tool castings, and generate the indicator configuration after weight optimization, including: Arrange the importance analysis results of the indicators, set weights for each indicator with reference to their contributions to the quality of machine tool castings, and obtain an initialization weight table; According to the initialization weight table, by calculating the entropy weight method, the weight ratio of the key performance indicators is increased, the weight distribution of the secondary performance indicators is reduced, and it is matched with the quality contribution of the machine tool casting to obtain an adjusted weight distribution table; According to the adjusted weight allocation table, iteratively evaluate and adjust the weight ratio of each performance indicator in the production of machine tool castings to obtain an indicator configuration after weight optimization; The entropy weight method is based on the improved formula I: ； Calculate the weight of each performance indicator, and adjust the weight of each performance indicator according to the calculation result to generate an adjusted weight distribution table; in, Represents the improved The weight of the indicator, It is the improved The entropy value of an indicator, is the quality coefficient of the indicator, is the stability coefficient of the indicator, is the complexity coefficient of the indicator, is the weighted The sample in The probability distribution of the indicators, is the normalization constant, represents the sample size, is the number of indicators; S3: Using the index configuration after weight optimization, calculating the similarity or difference between the parameters of the tool casting, and constructing a similarity analysis model, including: Select the index configuration after weight optimization, mark the performance index and weight of each machine tool casting, establish an analysis benchmark, and obtain an analysis benchmark index and weight list; Using the analytical benchmark indicators and weight list, select key performance indicators in the production of machine tool castings, sort and compare the performance data of each casting according to the weight list, and analyze the differences in performance data between castings to generate quantitative results of performance differences; Based on the performance difference quantification results, a similarity analysis model is obtained by mapping the similarity or difference of the performance of the machine tool casting; S4: By using the similarity analysis model, the correlation between machine tool casting parameters and performance is analyzed, key parameters are identified, and key parameters and their influence evaluation results are obtained, including: Based on the similarity analysis model, the Euclidean distance calculation method and the hierarchical analysis method are used to screen and classify the machine tool casting parameter data, and the parameter correlation matrix is ​​constructed based on the similarity or difference between the parameters; Based on the parameter correlation matrix, matrix operations are used to quantify the correlation strength between each parameter and the performance of the machine tool casting, obtain the impact value of each parameter on the performance, and obtain a list of parameter impact values; Based on the parameter influence value list, the key parameters affecting the performance of the machine tool casting are screened by setting thresholds, and the parameters and their corresponding influence values ​​are summarized and sorted as key indicators to obtain the key parameters and their influence evaluation results; The Euclidean distance calculation method is based on the improved formula II: ； Screen and classify the machine tool casting parameter data and construct the parameter correlation matrix; in, is the improved Euclidean distance, Represents the precision parameter vector of machine tool castings, where each element The accuracy values ​​corresponding to different castings or different measuring points, Represents the surface roughness parameter vector of machine tool castings, where each element Corresponding to the surface roughness values ​​of different castings or different measuring points, Represents the material strength parameter vector of the machine tool casting, where each element Corresponding to different castings or different test material strength values, Represents the ambient temperature parameter vector, where each element Corresponding to the ambient temperature values ​​under different castings or different conditions, represents the weight value of the i-th parameter, is the total number of parameters; S5: combining the weight-optimized index configuration and key parameters and their influence evaluation results, adjusting the structure and parameters of the similarity analysis model, mapping the current performance of the machine tool casting, and establishing an efficiency evaluation model; S6: Apply the performance evaluation model to analyze the current machine tool casting production data, identify the key factors affecting casting performance, and formulate a casting production optimization evaluation plan based on the analysis results; The importance analysis results of the indicators include the distribution fluctuation score of the precision indicators of machine tool castings, the variation range score of the surface roughness indicators and the contribution score of the mechanical performance indicators. The indicator configuration after weight optimization includes the optimized weight value of the precision indicators of machine tool castings, the optimized weight value of the surface roughness indicators and the optimized weight value of the mechanical performance indicators. The similarity analysis model includes parameter difference evaluation calculated according to the optimized weights of machine tool castings, performance comparison indicators between castings and similarity scores. The key parameters and their influence evaluation results include the scores of precision parameters, key surface treatment parameters and mechanical performance parameters that affect the performance of machine tool castings. The performance evaluation model includes model evaluation indicators, performance mapping and parameter influence measures adjusted for the performance of machine tool castings. The casting production optimization evaluation plan includes recommended process adjustment measures, raw material quality improvement plans and manufacturing parameter optimization for machine tool casting production.

Images