A part repairing method, device, equipment and readable storage medium

Abstract

The application relates to the technical field of manufacturing, and discloses a part repairing method, device, equipment and readable storage medium, the method comprising the following steps: obtaining original feature data of a target part from a feature dimension affecting the service life of the part; performing selection and dimension reduction processing on the original feature data by using a regression model to obtain target feature data; calculating the variation coefficient of the features in the target feature data, and determining the weight of the corresponding features by using the variation coefficient; performing trade-off sequencing on different laser cladding repair schemes by using a multi-attribute decision model and in combination with the differences of the different laser cladding repair schemes and the weights of the features in the target feature data; and selecting a target laser cladding repair scheme for repairing the target part from the different laser cladding repair schemes by using the sequencing result. The application has the technical effect of being capable of effectively determining the best target laser cladding repair scheme for repairing the target part.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of manufacturing, in particular to a part repair method, device, equipment and readable storage medium. BACKGROUND

[0002] In the fields of modern industrial manufacturing such as rail transit, wind power equipment, marine vessels and engineering machinery, shaft parts as key transmission and bearing components in mechanical equipment, their performance directly affects the running stability and reliability of the entire mechanical system. Especially in high-load and high-wear conditions such as rail transit, marine vessels and engineering machinery, the fatigue life of shaft parts becomes a key factor restricting the service life and maintenance cost of equipment.

[0003] Laser cladding technology as an advanced surface repair and strengthening technology can significantly improve the fatigue life and wear resistance of shaft parts by cladding a layer of alloy material with excellent mechanical properties and corrosion resistance on the surface of the shaft parts. However, laser cladding repair processes / programs have diversity, and the fatigue life of shaft parts after different laser cladding repair programs varies. How to scientifically and accurately evaluate and select the optimal process is of great significance to maximize cost-effectiveness and ensure product quality in large-scale production and manufacturing.

[0004] Currently, the specific implementation of laser cladding repair shaft part fatigue life decision evaluation has significant deficiencies in high-dimensional data processing, weight distribution objectivity and multi-attribute decision calculation model optimization. It is impossible to determine the best laser cladding repair program.

[0005] In summary, how to effectively determine the best laser cladding repair program for repairing parts and other issues is a technical problem that needs to be solved by technical personnel in the field. SUMMARY

[0006] The purpose of the present application is to provide a part repair method, device, equipment and readable storage medium to determine the best laser cladding repair program for repairing parts.

[0007] To solve the above technical problems, the present application provides the following technical solutions.

[0008] A part repair method, comprising:

[0009] Obtaining original feature data of a target part from feature dimensions affecting the life of the part;

[0010] Using a regression model to select and reduce the dimension of the original feature data to obtain target feature data;

[0011] Calculating the coefficient of variation of the features in the target feature data, and using the coefficient of variation to determine the weight of the corresponding features;

[0012] The multi-attribute decision model is used to compromise and sort the different laser cladding repair schemes in combination with the differences of the different laser cladding repair schemes and the weights of the features in the target feature data corresponding to the features.

[0013] The target laser cladding repair scheme for repairing the target part is selected from the different laser cladding repair schemes by using the sorting result.

[0014] Preferably, the original feature data is selected and reduced in dimension by using a regression model, including:

[0015] The regression model is constructed.

[0016] The optimal value of the regularization parameter λ is selected by using cross-validation to optimize the regression model.

[0017] The regression model performs coefficient compression processing on the original feature data to realize selection and dimension reduction of the original feature data.

[0018] Preferably, the coefficient of variation of the features in the target feature data is calculated, including:

[0019] The mean value of the current feature in the target feature data and the mean square deviation of the evaluation feature are calculated.

[0020] The ratio of the mean value to the mean square deviation is determined as the coefficient of variation.

[0021] Preferably, the weight of the corresponding feature is determined by using the coefficient of variation, including:

[0022] The coefficients of variation of the features in the target feature data are summed to obtain a total coefficient of variation.

[0023] The ratio of the coefficient of variation of the current feature in the target feature data to the total coefficient of variation is determined as the weight of the current feature.

[0024] Preferably, the multi-attribute decision model is used to compromise and sort the multiple laser cladding repair schemes in combination with the differences of the different laser cladding repair schemes and the weights of the features in the target feature data corresponding to the features, including:

[0025] The different laser cladding repair schemes and the target feature data, and the weights of the features corresponding to the features are input into the multi-attribute decision model.

[0026] The group utility value and the individual regret value of each laser cladding repair scheme, and the decision index value are calculated by using the multi-attribute decision model.

[0027] The group utility value and the individual regret value are used to perform trade-off sequencing on a plurality of laser cladding repair schemes.

[0028] Preferably, the original feature data of the target part is acquired from a feature dimension affecting the service life of the part, including:

[0029] The original feature data is acquired from at least one feature dimension of material properties, geometric parameters, load conditions and process parameters;

[0030] Correspondingly, before the original feature data is selected and processed by the regression model, the method further includes:

[0031] The original feature data is subjected to data cleaning and missing value processing;

[0032] The classification variables in the original feature data are subjected to one-hot encoding.

[0033] Preferably, the method further includes:

[0034] The target laser cladding repair scheme is executed to repair the target part; the target part is a shaft part.

[0035] A part repair device, comprising:

[0036] A feature acquisition module is configured to acquire original feature data of a target part from a feature dimension affecting the service life of the part;

[0037] A feature selection module is configured to select and process the original feature data by a regression model to obtain target feature data;

[0038] A weight determination module is configured to calculate a coefficient of variation of features in the target feature data and determine weights of the corresponding features by using the coefficient of variation;

[0039] A scheme sequencing module is configured to perform trade-off sequencing on a plurality of laser cladding repair schemes by using a multi-attribute decision model and combining differences of different laser cladding repair schemes and the weights of the features in the target feature data;

[0040] A scheme selection module is configured to execute the first laser cladding repair scheme to repair the target part.

[0041] An electronic device, comprising:

[0042] A memory is configured to store a computer program;

[0043] A processor is configured to execute the computer program to implement the steps of the part repair method.

[0044] A readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described part repair method.

[0045] The method provided in this application includes: obtaining original feature data of a target part from the feature dimensions affecting the lifespan of the part; selecting and reducing the dimensionality of the original feature data using a regression model to obtain target feature data; calculating the coefficient of variation of the features in the target feature data and determining the weight of the corresponding features using the coefficient of variation; using a multi-attribute decision model and combining the differences between different laser cladding repair schemes and the weights corresponding to the features in the target feature data to perform a trade-off ranking of different laser cladding repair schemes; and using the ranking results to select a target laser cladding repair scheme for repairing the target part from the different laser cladding repair schemes.

[0046] In this application, the original feature data corresponding to the feature dimensions affecting the lifespan of the target part are first obtained. Then, feature extraction and dimensionality reduction are performed on the original feature data based on a regression model. Specifically, a regression algorithm is used to select and reduce the dimensionality of the original attributes affecting the fatigue life of the target part, accurately screening key features, i.e., the target feature data, to reduce the dimensionality of the decision model. Next, the coefficient of variation method is used to objectively calculate the weight of each attribute based on the dispersion of the attribute data, thereby avoiding interference from subjective human factors and ensuring the fairness of weight allocation. Finally, using a multi-attribute decision model, the importance of each attribute and the differences in solutions are comprehensively considered to rank different laser cladding repair solutions. Based on the ranking results, the optimal target laser cladding repair solution for repairing the target part can be selected.

[0047] In other words, this application can solve the problem of redundant and complex evaluation indicators for high-dimensional data, improve the objectivity of weight allocation and decision evaluation of shaft parts, provide an effective technical means for reliability assessment of laser repair of shaft parts, and effectively determine the optimal target laser cladding repair scheme for the target parts.

[0048] Accordingly, embodiments of this application also provide a part repair apparatus, device, and readable storage medium corresponding to the above-described part repair method, which have the above-described technical effects, and will not be repeated here. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1This is a flowchart illustrating the implementation of a part repair method in an embodiment of this application.

[0051] Figure 2 This is a flowchart illustrating a specific implementation of a part repair method in this application.

[0052] Figure 3 This is a schematic diagram of the structure of a parts repair device according to an embodiment of this application;

[0053] Figure 4 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application;

[0054] Figure 5 This is a schematic diagram of the specific structure of an electronic device in an embodiment of this application. Detailed Implementation

[0055] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0056] Please refer to Figure 1 , Figure 1 This is a flowchart of a part repair method according to an embodiment of this application, which includes the following steps.

[0057] S101. Obtain the original feature data of the target part from the feature dimensions that affect the life of the part.

[0058] To select the optimal laser cladding repair scheme that can improve the lifespan of the target part, this embodiment first obtains the original feature data corresponding to the feature dimensions that can affect the lifespan of the part. Since material properties, geometric parameters, load conditions, process parameters, and the usage scenario of the part can affect the lifespan of the part, in practical applications, original feature data corresponding to at least one of the feature dimensions among material properties, geometric parameters, load conditions, process parameters, and the usage scenario of the part can be obtained.

[0059] In one specific embodiment of this application, obtaining the original feature data of the target part from the feature dimensions affecting the part's lifespan includes:

[0060] Obtain raw feature data from at least one of the following feature dimensions: material properties, geometric parameters, load conditions, and process parameters;

[0061] Accordingly, before using regression models to select and reduce the dimensionality of the original feature data, the following steps are also included:

[0062] Perform data cleaning on the original feature data and handle missing values;

[0063] One-hot encoding is performed on the categorical variables in the original feature data.

[0064] Among them, material properties include powder composition, powder particle size, uniformity, tensile strength, hardness, yield strength, elongation, etc.

[0065] Geometric parameters include repair area, repair thickness, and repair shape;

[0066] Load conditions include load type, load frequency, cyclic stress amplitude, and mean stress;

[0067] Process parameters include laser power, spot radius, defocusing amount, and pre- and post-heat treatment methods.

[0068] The process parameters correspond to different laser cladding repair schemes.

[0069] In other words, raw feature data related to the target part can be collected, including material properties, geometric parameters, load conditions, process parameters, and application environment. The acquired raw feature data is cleaned, missing values ​​are removed, and categorical variables undergo one-hot encoding and other preprocessing to ensure data quality and consistency. Through data preprocessing, a complete and standardized dataset is obtained, providing a foundation for subsequent feature extraction and model building.

[0070] S102. Use a regression model to select and reduce the dimensionality of the original feature data to obtain the target feature data.

[0071] Considering that the repair of target parts using laser cladding involves multiple factors affecting the part's lifespan, including but not limited to the composition of the repair powder, laser power, defocusing amount, spot diameter, and cladding speed, this embodiment employs a regression model to select and reduce the dimensionality of the original feature data in order to obtain key target feature data with lower dimensionality.

[0072] In one specific embodiment of this application, a regression model is used to select and reduce the dimensionality of the original feature data, including:

[0073] Construct a regression model;

[0074] Cross-validation is used to select the optimal value of the regularization parameter λ in order to optimize the regression model;

[0075] The regression model performs coefficient compression on the original feature data to select and reduce the dimensionality of the original feature data.

[0076] In other words, LASSO regression can be used for feature selection and dimensionality reduction. Specifically, a LASSO regression model can be constructed, and then the optimal value of the regularization parameter λ can be selected through cross-validation to optimize model performance. LASSO regression reduces the coefficients of unimportant features to zero by compressing the coefficients, thereby achieving feature selection. This yields a set of target feature data that has a key impact on the fatigue life of the target part, reducing the dimensionality of the decision evaluation model.

[0077] For example, the LASSO regression model can be used to screen out key features that affect the fatigue life of the target part from the original feature data, including but not limited to target feature data such as laser power, defocusing amount, spot diameter, cladding speed, powder feeding speed, and preheating temperature.

[0078] S103. Calculate the coefficient of variation of the features in the target feature data, and use the coefficient of variation to determine the weight of the corresponding feature.

[0079] To eliminate the subjectivity of subjective judgments, this embodiment employs the coefficient of variation (COP) method to objectively calculate the weights of the screened attributes. The COP method is an objective weighting method based on the degree of data dispersion. It reflects the relative importance of each attribute by calculating its COP. A larger COP indicates greater dispersion of the attribute and a greater impact on the evaluation results; therefore, it should be assigned a larger weight. This process avoids the influence of subjective factors on weight allocation, improving the objectivity and fairness of the weight distribution.

[0080] Specifically, the coefficient of variation of each feature in the defect feature data can be calculated separately, and then the weight of the feature can be determined by combining the coefficient of variation.

[0081] In one specific embodiment of this application, calculating the coefficient of variation of features in the target feature data includes:

[0082] Calculate the mean of the current feature in the target feature data and the root mean square error of the evaluation feature;

[0083] The ratio of the mean to the standard deviation is defined as the coefficient of variation.

[0084] For example, the following steps and formulas can be used to calculate the coefficient of variation:

[0085] (1) Using the formula below, calculate the first... Item features Mean;

[0086] Where n is the number of feature items in the target feature data, m is the numerical value of that feature item, i is the variable, and j is the current item. Let i be the feature value of variable i and j be the current term.

[0087] (2) Using the formula below, calculate the first... Evaluation features Mean squared error;

[0088] .

[0089] (3) Using the formula below, calculate the first... Evaluation features Coefficient of variation;

[0090] .

[0091] In one specific embodiment of this application, the weight of a corresponding feature is determined using the coefficient of variation, including:

[0092] The total coefficient of variation is obtained by summing the coefficients of variation of the features in the target feature data;

[0093] The weight of the current feature is determined by the ratio of the coefficient of variation of the current feature to the total coefficient of variation in the target feature data.

[0094] For example, when calculating weights, you can refer to the following steps and formulas:

[0095] The coefficients of variation of each feature are normalized to obtain the coefficients of variation for each feature. Weight, .

[0096] In other words, the objective weights of the selected features are calculated based on the coefficient of variation method. The specific steps include calculating the coefficient of variation for each feature to reflect the range and dispersion of its value. A larger coefficient of variation indicates a greater impact of the feature on the evaluation results, and therefore it should be assigned a greater weight. The coefficients of variation for all features are then normalized to obtain the objective weights of each feature, providing a basis for subsequent decision-making and evaluation.

[0097] S104. Using a multi-attribute decision model, and combining the differences between different laser cladding repair schemes with the weights of features in the target feature data, a compromise ranking of different laser cladding repair schemes is performed.

[0098] After obtaining the weights corresponding to the features in the target feature data, a multi-attribute decision model can be used, combined with the differences between different laser cladding repair schemes and the weights corresponding to the features in the target feature data, to make a compromise ranking of different laser cladding repair schemes.

[0099] For example, the VIKOR (Multi-Criterion Decision Analysis (MCDA) method) multi-attribute decision model can be used to comprehensively consider the importance of each attribute and the differences between schemes, and to rank laser cladding repair schemes with trade-offs, thus achieving the scientific selection of the optimal scheme. The VIKOR model is a trade-off ranking method based on ideal and negative ideal solutions. It calculates the distance of each scheme to the ideal and negative ideal solutions, comprehensively considers the importance of each attribute and the differences between schemes, and obtains a trade-off ranking result for each scheme. In the ranking process, the VIKOR model can balance the importance of each attribute, avoiding over-focus on a single attribute or local optima, thereby achieving the selection of the globally optimal scheme. This method can more scientifically and accurately evaluate the merits of different laser cladding repair schemes, providing effective decision support for industrial production.

[0100] In one specific embodiment of this application, a multi-attribute decision model is used, and by combining the differences between different laser cladding repair schemes and the weights corresponding to features in the target feature data, a trade-off ranking of multiple laser cladding repair schemes is performed, including:

[0101] Different laser cladding repair schemes and target feature data, as well as the weights corresponding to the features, are input into the multi-attribute decision model;

[0102] The group utility value, individual regret value, and decision index value of each laser cladding repair scheme are calculated using a multi-attribute decision model.

[0103] Multiple laser cladding repair schemes were ranked and compromised using group utility value, individual regret value, and decision index value.

[0104] Among them, the laser cladding repair scheme uses 316L repair powder, preheats to 150℃, has a laser power of 1300W, a spot size of 2.5mm, a cladding speed of 10mm / s, and a defocusing amount of 0.

[0105] For example: Suppose that there are n key features affecting the fatigue life of a target part, C1, C2, …, Cj, …, Cn, with weights W1, W2, …, Wn respectively, and m repair schemes X1, X2, …, Xi, …, Xm. Where f ij This represents the evaluation value of the i-th scheme under the j-th feature.

[0106] Key features may include two types: benefit-oriented features and cost-oriented features. For benefit-oriented features, higher values ​​are better, while for cost-oriented features, lower values ​​are better. The methods for calculating positive and negative ideal solutions differ between the two types of key features.

[0107] The specific calculation method is as follows:

[0108] The ideal solution;

[0109] Negative ideal solution;

[0110] (1) The characteristics of the benefit-oriented model are as follows:

[0111] , where min represents the minimum and max represents the maximum.

[0112] (2) The characteristics of cost-type are as follows:

[0113] .

[0114] Group utility value:

[0115] .

[0116] Individual Regret Value:

[0117] .

[0118] Q Index:

[0119] .

[0120] Then, based on the obtained S, R, and Q values, and taking into account the importance of each attribute and the differences between the schemes, a compromise ranking is performed.

[0121] S105. Using the sorting results, select the target laser cladding repair scheme from the different laser cladding repair schemes to repair the target part.

[0122] After completing the trade-off ranking, a target laser cladding repair scheme can be selected from different laser cladding repair schemes based on the ranking results. This target laser cladding repair scheme is the optimal scheme / process for repairing the target part.

[0123] In one specific embodiment of this application, the method further includes: performing a target laser cladding repair scheme to repair a target part; the target part is a shaft-type part. That is, performing the target laser cladding repair scheme thereby repairing the shaft-type part.

[0124] The method provided in this application includes: obtaining original feature data of a target part from the feature dimensions affecting the lifespan of the part; selecting and reducing the dimensionality of the original feature data using a regression model to obtain target feature data; calculating the coefficient of variation of the features in the target feature data and determining the weight of the corresponding features using the coefficient of variation; using a multi-attribute decision model and combining the differences between different laser cladding repair schemes and the weights corresponding to the features in the target feature data to perform a trade-off ranking of different laser cladding repair schemes; and using the ranking results to select a target laser cladding repair scheme for repairing the target part from the different laser cladding repair schemes.

[0125] In this application, the original feature data corresponding to the feature dimensions affecting the lifespan of the target part are first obtained. Then, feature extraction and dimensionality reduction are performed on the original feature data based on a regression model. Specifically, a regression algorithm is used to select and reduce the dimensionality of the original attributes affecting the fatigue life of the target part, accurately screening key features, i.e., the target feature data, to reduce the dimensionality of the decision model. Next, the coefficient of variation method is used to objectively calculate the weight of each attribute based on the dispersion of the attribute data, thereby avoiding interference from subjective human factors and ensuring the fairness of weight allocation. Finally, using a multi-attribute decision model, the importance of each attribute and the differences in solutions are comprehensively considered to rank different laser cladding repair solutions. Based on the ranking results, the optimal target laser cladding repair solution for repairing the target part can be selected.

[0126] In other words, this application can solve the problem of redundant and complex evaluation indicators for high-dimensional data, improve the objectivity of weight allocation and decision evaluation of shaft parts, provide an effective technical means for reliability assessment of laser repair of shaft parts, and effectively determine the optimal target laser cladding repair scheme for the target parts.

[0127] To facilitate a better understanding and implementation of the part repair method provided in the embodiments of this application by those skilled in the art, the following is combined with... Figure 2 Please provide an explanation.

[0128] like Figure 2 As shown, the original feature data is obtained first.

[0129] The original feature data includes, but is not limited to, material properties, geometric parameters, load conditions, and process parameters. Then, the original feature data undergoes preprocessing operations such as data cleaning (e.g., handling missing values, one-hot encoding of categorical variables to standardize the data).

[0130] Then, LASSO regression is used for feature extraction. Specifically, a regression model is first constructed, then optimized using a regularization parameter λ, and finally validated using the regression model to achieve variable selection and structural output. The output structure is the target feature data.

[0131] Next, objective weights are determined based on the coefficient of variation. Specifically, the mean of the j-th indicator is calculated, the root mean square error of the j-th indicator is calculated, and the coefficient of variation of the j-th indicator is calculated. The coefficients of variation of each indicator are then normalized to obtain the weights of each indicator.

[0132] Finally, based on VIKOR's decision evaluation method, the alternatives are ranked. Specifically, the positive ideal (PIS) and negative ideal (NIS) of each indicator are determined, the group utility value S and individual regret value R of each decision alternative are calculated, the decision indicator value Q of each alternative is calculated, and finally, the alternatives are ranked according to the obtained S, R, and Q.

[0133] Based on the ranking results, the most suitable laser cladding repair scheme for the target part can be selected.

[0134] Corresponding to the above method embodiments, this application also provides a part repair device, and the part repair device described below can be referred to in correspondence with the part repair method described above.

[0135] See Figure 3 As shown, the device includes the following modules:

[0136] The feature acquisition module 101 is used to acquire the original feature data of the target part from the feature dimensions that affect the life of the part;

[0137] The feature selection module 102 is used to select and reduce the dimensionality of the original feature data using a regression model to obtain the target feature data;

[0138] The weight determination module 103 is used to calculate the coefficient of variation of features in the target feature data and to determine the weight of the corresponding feature using the coefficient of variation.

[0139] The scheme ranking module 104 is used to use a multi-attribute decision model and combine the differences between different laser cladding repair schemes and the weights corresponding to the features in the target feature data to perform a compromise ranking of multiple laser cladding repair schemes.

[0140] The scheme selection module 105 is used to execute the top-ranked laser cladding repair scheme to repair the target part.

[0141] The apparatus provided in this application includes: acquiring original feature data of a target part from the feature dimensions affecting the lifespan of the part; selecting and reducing the dimensionality of the original feature data using a regression model to obtain target feature data; calculating the coefficient of variation of the features in the target feature data and determining the weight of the corresponding features using the coefficient of variation; using a multi-attribute decision model and combining the differences between different laser cladding repair schemes and the weights corresponding to the features in the target feature data to perform a trade-off ranking of different laser cladding repair schemes; and using the ranking result to select a target laser cladding repair scheme for repairing the target part from the different laser cladding repair schemes.

[0142] In this application, the original feature data corresponding to the feature dimensions affecting the lifespan of the target part are first obtained. Then, feature extraction and dimensionality reduction are performed on the original feature data based on a regression model. Specifically, a regression algorithm is used to select and reduce the dimensionality of the original attributes affecting the fatigue life of the target part, accurately screening key features, i.e., the target feature data, to reduce the dimensionality of the decision model. Next, the coefficient of variation method is used to objectively calculate the weight of each attribute based on the dispersion of the attribute data, thereby avoiding interference from subjective human factors and ensuring the fairness of weight allocation. Finally, using a multi-attribute decision model, the importance of each attribute and the differences in solutions are comprehensively considered to rank different laser cladding repair solutions. Based on the ranking results, the optimal target laser cladding repair solution for repairing the target part can be selected.

[0143] In other words, this application can solve the problem of redundant and complex evaluation indicators for high-dimensional data, improve the objectivity of weight allocation and decision evaluation of shaft parts, provide an effective technical means for reliability assessment of laser repair of shaft parts, and effectively determine the optimal target laser cladding repair scheme for the target parts.

[0144] In one specific embodiment of this application, the feature selection module is specifically used to construct a regression model;

[0145] Cross-validation is used to select the optimal value of the regularization parameter λ in order to optimize the regression model;

[0146] The regression model performs coefficient compression on the original feature data to select and reduce the dimensionality of the original feature data.

[0147] In one specific embodiment of this application, the weight determination module is specifically used to calculate the mean of the current feature in the target feature data and the root mean square error of the evaluation feature;

[0148] The ratio of the mean to the standard deviation is defined as the coefficient of variation.

[0149] In one specific embodiment of this application, the weight determination module is specifically used to sum the coefficients of variation of the features in the target feature data to obtain the total coefficient of variation;

[0150] The weight of the current feature is determined by the ratio of the coefficient of variation of the current feature to the total coefficient of variation in the target feature data.

[0151] In one specific embodiment of this application, the scheme ranking module is specifically used to input different laser cladding repair schemes and target feature data, as well as the weights corresponding to the features, into a multi-attribute decision model;

[0152] The group utility value, individual regret value, and decision index value of each laser cladding repair scheme are calculated using a multi-attribute decision model.

[0153] Multiple laser cladding repair schemes were ranked and compromised using group utility value, individual regret value, and decision index value.

[0154] In one specific embodiment of this application, the feature acquisition module is specifically used to acquire original feature data from at least one feature dimension among material properties, geometric parameters, load conditions, and process parameters; correspondingly, before using a regression model to select and reduce the dimensionality of the original feature data, a feature prediction module is also included, which is used to clean the original feature data and handle missing values; and to perform one-hot encoding on the categorical variables in the original feature data.

[0155] In one specific embodiment of this application, it further includes: a repair execution module, used to execute a target laser cladding repair scheme to repair the target part; the target part is a shaft-type part.

[0156] Corresponding to the above method embodiments, this application also provides an electronic device. The electronic device described below and the component repair method described above can be referred to each other.

[0157] See Figure 4 As shown, the electronic device includes:

[0158] Memory 332 is used to store computer programs;

[0159] The processor 322 is used to implement the steps of the part repair method in the above method embodiment when executing a computer program.

[0160] For details, please refer to Figure 5 , Figure 5This is a schematic diagram of the specific structure of an electronic device provided in this embodiment. The electronic device can vary significantly due to differences in configuration or performance. It may include one or more central processing units (CPUs) (e.g., one or more processors) and a memory 332. The memory 332 stores one or more computer programs 342 or data 344. The memory 332 can be temporary or permanent storage. The program stored in the memory 332 may include one or more modules (not shown in the diagram), each module may include a series of instruction operations on the data processing device. Furthermore, the processor 322 may be configured to communicate with the memory 332 and execute the series of instruction operations stored in the memory 332 on the electronic device 301.

[0161] Electronic device 301 may also include one or more power supplies 326, one or more wired or wireless network interfaces 350, one or more input / output interfaces 358, and / or one or more operating systems 341.

[0162] The steps in the component repair method described above can be implemented by the structure of the electronic device.

[0163] Corresponding to the above method embodiments, this application also provides a readable storage medium. The readable storage medium described below can be referred to in conjunction with the part repair method described above.

[0164] A readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the part repair method described in the above method embodiments.

[0165] The readable storage medium can specifically be a USB flash drive, external hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, or any other readable storage medium capable of storing program code.

[0166] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0167] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0168] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0169] Finally, it should be noted that in this document, relationships such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "include," "contain," or any other variations are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0170] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for repairing parts, characterized in that, include: Obtain the original feature data of the target part from the feature dimensions that affect the lifespan of the part; The original feature data is selected and dimensionality reduced using a regression model to obtain the target feature data; Calculate the coefficient of variation of the features in the target feature data, and use the coefficient of variation to determine the weight of the corresponding feature; Using a multi-attribute decision model, and combining the differences between different laser cladding repair schemes with the weights corresponding to the features in the target feature data, different laser cladding repair schemes are ranked by trade-offs; wherein, the laser cladding repair scheme corresponds to different process parameters; Using the ranking results, a target laser cladding repair scheme for repairing the target part is selected from different laser cladding repair schemes; The target laser cladding repair scheme is executed to repair the target part; the target part is a shaft-type part.

2. The part repair method according to claim 1, characterized in that, The original feature data is selected and dimensionality reduced using a regression model, including: Construct the regression model; Cross-validation is used to select the optimal value of the regularization parameter λ to optimize the regression model; The regression model performs coefficient compression on the original feature data to select and reduce the dimensionality of the original feature data.

3. The part repair method according to claim 1, characterized in that, Calculating the coefficient of variation of features in the target feature data includes: Calculate the mean of the current feature and the root mean square error of the evaluation feature in the target feature data; The ratio of the mean to the root mean deviation is determined as the coefficient of variation.

4. The part repair method according to claim 3, characterized in that, Determining the weights of corresponding features using the coefficient of variation includes: The total coefficient of variation is obtained by summing the coefficients of variation of the features in the target feature data; The ratio of the coefficient of variation of the current feature to the total coefficient of variation in the target feature data is determined as the weight of the current feature.

5. The part repair method according to claim 1, characterized in that, Using a multi-attribute decision model, and combining the differences between different laser cladding repair schemes with the weights corresponding to features in the target feature data, a trade-off ranking of multiple laser cladding repair schemes is performed, including: Different laser cladding repair schemes and the target feature data, as well as the weights corresponding to the features, are input into the multi-attribute decision model; The group utility value, individual regret value, and decision index value of each laser cladding repair scheme are calculated using the multi-attribute decision model. The group utility value, the individual regret value, and the decision index value are used to make a trade-off ranking of multiple laser cladding repair schemes.

6. The part repair method according to claim 1, characterized in that, From the feature dimensions that affect the lifespan of a component, obtain the original feature data of the target component, including: The original feature data is obtained from at least one of the following feature dimensions: material properties, geometric parameters, load conditions, and process parameters. Accordingly, before using a regression model to select and reduce the dimensionality of the original feature data, the following steps are also included: The original feature data is cleaned, and missing values ​​are removed. One-hot encoding is performed on the categorical variables in the original feature data.

7. A parts repair device, characterized in that, include: The feature acquisition module is used to obtain the original feature data of the target part from the feature dimensions that affect the life of the part; The feature selection module is used to select and reduce the dimensionality of the original feature data using a regression model to obtain target feature data; The weight determination module is used to calculate the coefficient of variation of the features in the target feature data and use the coefficient of variation to determine the weight of the corresponding feature. The scheme ranking module is used to perform a trade-off ranking of multiple laser cladding repair schemes by utilizing a multi-attribute decision model and combining the differences between different laser cladding repair schemes with the weights corresponding to the features in the target feature data; wherein, the laser cladding repair schemes correspond to different process parameters; The scheme selection module is used to execute the top-ranked laser cladding repair scheme to repair the target part; The repair execution module is used to execute the target laser cladding repair scheme to repair the target part; the target part is a shaft part.

8. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to implement the steps of the part repair method as described in any one of claims 1 to 6 when executing the computer program.

9. A readable storage medium, characterized in that, The readable storage medium stores a computer program that, when executed by a processor, implements the steps of the part repair method as described in any one of claims 1 to 6.

Images