Method for optimizing technological parameters of entropy alloy coating in high-speed laser cladding

Abstract

The invention provides a process parameter optimization method for an entropy alloy coating in high-speed laser cladding, and belongs to the field of high-speed laser cladding. Comprising the following steps that firstly, laser cladding process parameters serve as input, the aspect ratio and microhardness of a cladding coating serve as output, and a high-speed laser single-channel cladding experiment is conducted through a center composite design (CCD) experiment scheme; then, according to a CCD experiment result, establishing a back propagation (GABP) neural network prediction model optimized by a genetic algorithm based on laser cladding process parameters and cladding layer aspect ratio and microhardness; and finally, on the basis of a second-generation non-dominated sorting genetic algorithm (NSGA-II), by taking the aspect ratio of the cladding layer and the microhardness as optimization targets and taking a GABP neural network prediction model as a fitness function of the NSGA-II, obtaining a Pareto leading-edge solution set through global optimization to complete process parameter optimization. The method has the technical effects that the defects of a trial and error method, a single variable method and other traditional optimization methods can be avoided, the experiment time is saved, the cost is saved, the cladding layer which is metallurgically bonded with the matrix and has no obvious cracks and holes can be obtained through the determined optimal process parameter combination, and the method is simple, high in universality and suitable for industrial production. And the method can be popularized and used for determining process parameters of other coatings in high-speed laser cladding.

Description

Technical Field

[0001] The invention belongs to the technical field of high-speed laser cladding, and in particular relates to a method for optimizing process parameters of a high-speed laser cladding medium-entropy alloy coating. Background Art

[0002] High-speed laser cladding is an advanced surface modification technology that rapidly melts and deposits one or more layers of metal powder on the surface of the substrate to form a cladding layer metallurgically bonded to the substrate to improve the material's wear resistance, corrosion resistance, high temperature resistance and other properties. It is widely used in the fields of aviation, energy, and automobiles. However, this process technology involves multiple parameters, such as laser power, scanning speed, powder feeding rate, etc. The selection and optimization of these parameters have an important influence on the quality and performance of the cladding layer.

[0003] Traditional optimization methods, such as single factor method and orthogonal method, are mainly based on experiments and experience accumulation. By adjusting the design process parameters, observing their influence on the quality of the cladding coating, and then gradually optimizing to achieve the best cladding effect. Although such methods are simple and easy to implement, they are inefficient, strong, and costly, and it is difficult to consider the interaction between multiple process parameters at the same time.

[0004] Therefore, there is an urgent need for an efficient and comprehensive process parameter optimization method that integrates the interactions of multiple process parameters to improve the overall quality of high-speed laser cladding alloy coatings. Intelligent optimization methods such as neural networks, genetic algorithms, and particle swarm algorithms have been widely used in process parameter optimization. These methods can automatically find the optimal process parameter combination, reduce manual intervention, and improve optimization efficiency. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art and to provide a new technical solution for a method for optimizing process parameters of a high-speed laser cladding medium-entropy alloy coating.

[0006] According to one aspect of the present invention, a method for optimizing process parameters of a high-speed laser cladding medium-entropy alloy coating is provided, comprising the following steps:

[0007] Step S1: taking the process parameters of the laser cladding medium entropy alloy coating as input and the aspect ratio (λ) and microhardness (HV) of the cladding coating as output, a high-speed laser single-pass cladding experiment is performed based on the CCD experimental scheme and relevant data collection is completed; wherein the CCD experimental scheme adopts a three-factor three-level experimental design;

[0008] Step S2: According to the CCD experimental results, a back propagation (GABP) neural network prediction model based on the relationship between the process parameters of the laser cladding medium entropy alloy coating and the quality of the cladding coating is established, and the accuracy of the prediction model is tested to see whether it meets the requirements for subsequent process parameter optimization;

[0009] Step S3: Based on the prediction model established in step S2, the second generation fast non-dominated sorting genetic algorithm (NSGA-II) is used to optimize the process parameters of the laser cladding medium entropy alloy coating, and the optimal combination of process parameters of the laser cladding medium entropy alloy coating is determined and experimentally verified.

[0010] Optionally, in step S1, the laser cladding medium entropy alloy coating process parameters include laser power A, scanning speed B, and powder feeding amount C.

[0011] Optionally, in step S1, the aspect ratio of the cladding coating is calculated as follows:

[0012]

[0013] In the above formula: λ represents the aspect ratio, W represents the melt width, the unit is mm; H represents the melt height, the unit is mm.

[0014] Optionally, in step S1, the preset particle size range of the powder of the medium entropy alloy coating is 10 μm to 60 μm.

[0015] Optionally, in step S1, before the cladding experiment, the surface of the substrate is first ground and polished with sandpaper, and then the contaminants on the surface of the substrate are ultrasonically cleaned with anhydrous ethanol and acetone, and then placed in a ventilated and dry place for use; at the same time, the medium-entropy alloy powder is placed in a vacuum drying oven and dried for 2 hours for use, with a preset temperature of 100°C.

[0016] Optionally, in step S1, after the cladding experiment is completed, a cross-sectional sample of the cladding coating is obtained by using a wire cutting machine, and the cross-sectional sample is ground, polished and corroded.

[0017] Optionally, in step S2, a prediction model of a back propagation neural network optimized by a genetic algorithm is established with laser power A, scanning speed B, and powder feeding amount C as inputs and aspect ratio (λ) and microhardness (HV) of the cladding coating as outputs;

[0018] Among them, the prediction model needs to normalize the parameters and performance data of each sample using the following formula before training and testing:

[0019]

[0020] In the above formula: X' is the value of the sample after normalization; X is the value of the sample before normalization; X max , X min are the maximum and minimum values ​​of the samples respectively;

[0021] The prediction results of the prediction model need to be denormalized using the following formula to obtain the actual value:

[0022]

[0023] In the above formula: y i is the predicted output after sample denormalization; is the predicted output before sample denormalization; max(y) and min(y) are the maximum and minimum values ​​of the sample before denormalization.

[0024] Optionally, in step S2, the goodness of fit R 2 Check whether the accuracy of the prediction model meets the requirements of subsequent multi-objective optimization of process parameters; 2 The calculation method is as follows:

[0025]

[0026] In the above formula: n is the number of samples in a single-pass cladding test, y i is the measured value of the single-pass cladding experiment, is the fitting value of the single-pass cladding layer performance index model, y i It is the average value of the measured values ​​obtained from all single-pass cladding experiments.

[0027] Optionally, R 2 Greater than 0.9.

[0028] Optionally, in step S3, the multi-objective optimization mathematical model is as follows:

[0029]

[0030] In the above formula, f1 is the aspect ratio prediction model based on GABP neural network, and f2 is the microhardness prediction model based on GABP neural network. F represents the nonlinear mapping relationship between the input and output of the NSGA-II optimization model; T is the laser cladding process parameter set, including laser power A, scanning speed B and powder feeding amount C; CP is the optimization object. i Represents individual C i The process parameters, C i-min , C i-max It is individual C i The minimum and maximum values ​​of various process parameters.

[0031] Optionally, in step S3, an optimal laser cladding medium entropy alloy coating process parameter verification experiment is carried out to calculate the relative error between the predicted value and the actual measured value and analyze the cladding layer quality optimization effect to verify the effectiveness of the optimization method.

[0032] A technical effect of the present invention is:

[0033] In the embodiment of the present application, the process parameter optimization method of the high-speed laser cladding medium-entropy alloy coating can simultaneously consider multiple conflicting objectives by combining the global search capabilities of the genetic algorithm and the BP neural network and the advantages of the NSGA-II algorithm in dealing with multi-objective optimization problems, so as to find a set of optimal laser cladding medium-entropy alloy coating process parameter combinations. The optimal process parameter combination determined by the present invention can obtain a cladding coating that is metallurgically bonded to the substrate and has no obvious cracks and holes, and the method is simple and highly versatile, and can be used to determine the process parameters of other materials for high-speed laser cladding. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 A schematic flow chart of a method for optimizing process parameters of a high-speed laser cladding medium-entropy alloy coating according to an embodiment of the present invention;

[0035] Figure 2 A schematic diagram of the geometric features of the cross section of the cladding coating in one embodiment of the present invention;

[0036] Figure 3 It is a flow chart of the NSGA-Ⅱ multi-objective optimization algorithm in one embodiment of the present invention;

[0037] Figure 4 The output value of the GABP neural network test sample aspect ratio in Example 1 is compared with the actual value;

[0038] Figure 5 The output value of the GABP neural network test sample microhardness in Example 1 is compared with the actual value;

[0039] Figure 6 is the Pareto frontier solution set of the target response value of the cladding coating in Example 1;

[0040] Figure 7 The cross-sectional morphology of the cladding coating of the comparative test group and the optimal process parameter group in Example 1;

[0041] Figure 8 is the Pareto frontier solution set of the target response value of the cladding coating in Example 2;

[0042] Fig. 9 The cross-sectional morphology of the cladding coating of the comparative test group and the optimal process parameter group in Example 2. DETAILED DESCRIPTION

[0043] Exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangements of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present application.

[0044] The embodiments of the present application will be described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and should not be construed as limitations on the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0045] According to one aspect of the present invention, see Figure 1 , provides a method for optimizing process parameters of high-speed laser cladding medium-entropy alloy coatings, which relates to the field of high-speed laser cladding technology, and is used to improve the forming quality of cladding coatings and solve the problem of low efficiency of the method for optimizing process parameters of high-speed laser cladding medium-entropy alloy coatings. x Medium entropy alloy coatings are taken as an example.

[0046] Specifically, the process parameter optimization method of the high-speed laser cladding medium-entropy alloy coating

[0047] The steps include:

[0048] Step S1: With the process parameters of the laser cladding medium entropy alloy coating as input and the aspect ratio (λ) and microhardness (HV) of the cladding coating as output, a CCD test scheme is used to conduct a high-speed laser single-pass cladding experiment and complete relevant data measurement (i.e., collect relevant data); wherein the CCD experimental scheme adopts a three-factor three-level experimental design. For example, the CCD experimental scheme adopts a three-factor three-level experimental scheme of the Central Composite Design (CCD) of the Response Surface module in the Design-Expert13 software.

[0049] Step S2: According to the CCD experimental results, a prediction model based on the process parameters of the laser cladding medium entropy alloy coating and the aspect ratio (λ) and microhardness (HV) of the cladding layer is established, and it is tested whether the accuracy of the prediction model meets the requirements of subsequent process parameter optimization.

[0050] Step S3: Based on the prediction model established in step S2, the second generation fast non-dominated genetic algorithm (NSGA-Ⅱ) is used to optimize the process parameters of the laser cladding medium entropy alloy coating, and the optimal combination of process parameters of the laser cladding medium entropy alloy coating is determined and experimentally verified.

[0051] Specifically, a multi-objective optimization model of process parameters is first established as follows;

[0052]

[0053] Among them, f1 is the aspect ratio prediction model based on GABP neural network, and f2 is the microhardness prediction model based on GABP neural network. F represents the nonlinear mapping relationship between the input and output of the NSGA-II optimization model, T is the laser cladding process parameter set, which includes three process parameters: laser power P, scanning speed V, and powder feeding speed S; the prediction model constructed by GABP neural network is the fitness function of NSGA-II algorithm, CP is the optimization object; C i Show individual C i The process parameters, c i-min 、c i-max It is individual C i The minimum and maximum values ​​of various process parameters.

[0054] Then, the parameters of the NSGA-Ⅱ algorithm are set, where the number of individuals in the population is N=50, the maximum number of iterations is G=120, the crossover probability is Pc=0.88, and the mutation probability is Pm=0.3.

[0055] The NSGA-Ⅱ multi-objective optimization algorithm process is shown in Figure 3 .

[0056] Specifically, the average aspect ratio and microhardness value of the 20 groups of samples obtained in step S1 are compared and analyzed with the aspect ratio and microhardness value of the cladding coating prepared under the optimal process parameters.

[0057] Specifically, a sample with good cladding coating bonding, no obvious defects and better target response value in step S1 is selected for comparative analysis with the cross-sectional morphology of the cladding layer obtained under the determined optimal laser cladding medium entropy alloy coating process parameter combination.

[0058] In an embodiment of the present application, the process parameter optimization method for high-speed laser cladding medium-entropy alloy coating combines the global search capabilities of genetic algorithm and BP neural network and the advantages of NSGA-II algorithm in dealing with multi-objective optimization problems. It can simultaneously consider multiple conflicting objectives to find a set of optimal laser cladding medium-entropy alloy coating process parameter combinations. The optimal process parameter combination determined by the present invention can obtain a cladding layer that is metallurgically bonded to the substrate and has no obvious cracks and holes. The method is simple and has high versatility, and can be used to determine the process parameters of other materials for high-speed laser cladding.

[0059] Optionally, in step S1, the process parameters of the laser cladding medium entropy alloy coating include laser power A, scanning speed B, and powder feeding amount C. This helps to reasonably and accurately optimize the process parameters of the high-speed laser cladding medium entropy alloy coating.

[0060] Optionally, in step S1, the aspect ratio of the cladding coating is calculated as follows:

[0061]

[0062] In the above formula: λ represents the aspect ratio, W represents the weld width, the unit is mm; H represents the weld depth, the unit is mm.

[0063] In the above embodiment, it is helpful to accurately calculate the aspect ratio of the cladding coating.

[0064] Optionally, in step S1, the medium entropy alloy powder is obtained by mechanically mixing FeCrNi powder and Mo powder in a molar ratio of 1:1:1:x, and the preset particle size range is 15μm to 55μm. That is, the medium entropy alloy powder is FeCrNiMo x .

[0065] Optionally, in step S1, before the cladding experiment, the cladding surface of the substrate is first ground and polished with sandpaper, and then the surface contaminants of the substrate are cleaned with anhydrous ethanol and acetone, and then placed in a ventilated and dry place for use; at the same time, the powder to be clad is placed in a vacuum drying oven and dried for 2 hours for use, and the preset temperature is 100°C.

[0066] The above-mentioned embodiment helps to carry out the cladding experiment stably and accurately.

[0067] Optionally, in step S1, after the cladding experiment is completed, a cross-section sample of the cladding coating is obtained by using a wire cutting machine, and the cross-section of the sample is ground, polished and etched, which is helpful for accurate analysis of the cross-section sample.

[0068] Optionally, in step S2, a cladding coating performance prediction model is established by optimizing a back propagation neural network (BPNN) with a genetic algorithm (GA) using laser power A, scanning speed B, and powder feeding amount C as inputs and an aspect ratio (λ) and microhardness (HV) of the cladding coating as outputs;

[0069] Among them, the prediction model needs to normalize the parameters and performance data of each sample using the following formula before training and testing:

[0070]

[0071] In the above formula: X' is the value of the sample after normalization; X is the value of the sample before normalization; X max , X min are the maximum and minimum values ​​of the samples respectively;

[0072] The prediction results of the prediction model need to be denormalized using the following formula to obtain the actual value:

[0073]

[0074] In the above formula: y iis the predicted output after sample denormalization; is the predicted output before sample denormalization; max(y) and min(y) are the maximum and minimum values ​​of the sample before denormalization.

[0075] In the above implementation, since the units and dimensions of the process parameters are inconsistent, in order to improve the prediction accuracy and convergence speed of the model, the neural network needs to normalize the parameters and performance data of each sample before training and testing. And the prediction results of the prediction model need to be denormalized to obtain the actual value, which helps to ensure the accuracy of the prediction results.

[0076] Optionally, in step S2, the goodness of fit R 2 Check whether the accuracy of the prediction model meets the requirements of subsequent multi-objective optimization of process parameters; 2 The calculation method is as follows:

[0077]

[0078] In the above formula: n is the number of samples in a single-pass cladding test, y i is the measured value of the single-pass cladding experiment, is the fitting value of the performance index model of a single-pass cladding layer, It is the average value of the measured values ​​obtained from all single-pass cladding experiments.

[0079] In the above implementation, it is helpful to quickly and accurately check whether the accuracy of the prediction model meets the requirements of subsequent multi-objective optimization of process parameters.

[0080] Optionally, R 2 Greater than 0.9. 2 The closer it is to 1, the better the fitting accuracy. It is generally believed that R 2 It is greater than 0.9, which means that the accuracy requirement of subsequent multi-objective optimization of process parameters is met.

[0081] Optionally, in step S3, an optimal laser cladding medium entropy alloy coating process parameter verification experiment is carried out to calculate the relative error between the predicted value and the actual measured value and the optimization effect to verify the effectiveness of the optimization method.

[0082] In the above implementation, the verification method of the prediction model is relatively simple and helps to verify the effect of the optimal process parameter combination on improving the aspect ratio and microhardness of the cladding coating.

[0083] Optionally, in step S3, the average aspect ratios and microhardness values ​​of the 20 groups of samples obtained in step S1 are compared and analyzed with the aspect ratios and microhardness values ​​of the cladding coatings prepared under optimal process parameters.

[0084] Optionally, in step S3, a sample with good cladding coating bonding, no obvious defects and better target response value in step S1 is selected for comparative analysis with the cross-sectional morphology of the cladding coating obtained under the determined optimal laser cladding medium entropy alloy coating process parameter combination.

[0085] The process parameter optimization method for high-speed laser cladding medium-entropy alloy coating is a multi-objective intelligent optimization algorithm based on the combination of genetic algorithm and BP neural network (GABP) and the second-generation non-dominated sorting genetic algorithm (NSGA-II), which can better improve the optimization efficiency of high-speed laser cladding process parameters.

[0086] In a specific embodiment, see Figure 1 Firstly, according to the central composite experimental design in the response surface, a high-speed laser single-pass cladding test was completed; secondly, with laser power, scanning speed, and powder feeding amount as optimization variables, and the aspect ratio and microhardness of the cladding coating as optimization targets, a back propagation neural network prediction model optimized by genetic algorithm was established; thirdly, the goodness of fit R 2 The reliability and prediction accuracy of the model were tested. Then, a multi-objective optimization model for the process parameters of high-speed laser cladding medium-entropy alloy coating was constructed with the goal of maximizing the aspect ratio and microhardness of the cladding coating. Then, the second-generation fast non-dominated genetic algorithm was used to obtain the Pareto frontier solution set of the optimization variables and target response values, and the optimal process parameters were determined. Finally, a verification experiment for the optimal process parameter combination was carried out to quantify the improvement effect of the cladding layer performance.

[0087] Example 1

[0088] For example, the base material is 45 steel, and the medium entropy alloy powder is FeCrNiMo 0.25 .

[0089] Exemplarily, in step S1, the corresponding relationship between the process parameters and the coding values ​​of the CCD experimental scheme is shown in Table 1 below.

[0090] Table 1 shows the corresponding relationship between the process parameters and coding values ​​of the CCD experimental scheme

[0091]

[0092] In a specific embodiment, the high-speed laser single-pass cladding experimental process in step S1 specifically includes the following sub-steps:

[0093] S11: Substrate pretreatment. Before the cladding test, the cladding surface of the brake drum substrate material needs to be ground and polished to remove the surface oxide scale, and the surface oil and rust need to be cleaned with anhydrous ethanol and acetone, and then placed in a ventilated and dry place for use.

[0094] S12: Drying of cladding powder: Place the cladding powder in a drying oven and vacuum dry it at 100°C for 2 hours for use, and the particle size is 15 μm to 53 μm;

[0095] S13: Check the instruments and equipment and set the experimental conditions and parameters, then move the CNC test bench to achieve single-pass cladding.

[0096] S14: Test results: Cut the cladding substrate sample obtained in step S13 perpendicular to the laser cladding direction to obtain a cross section of the cladding layer, then grind and polish the cross section using 400 mesh, 600 mesh, 1000 mesh, and 2000 mesh sandpaper and polishing paste, and then use a corrosive liquid to corrode the cross section of the cladding layer, so as to facilitate the subsequent observation and measurement of the molten width and height of the single-channel cladding coating. Figure 3 , W represents the melt width, and H represents the melt height. The experimental scheme and results are shown in Table 2 below.

[0097] Table 2 shows the experimental scheme and results

[0098]

[0099]

[0100] Exemplarily, in step S2, the parameter settings of the prediction model of the back propagation neural network optimized by the genetic algorithm (ie, the GABP neural network model) are shown in Table 3 below.

[0101] Table 3 shows the parameter settings of the GABP neural network model

[0102]

[0103] In a specific implementation, the aspect ratio test sample output value of the GABP neural network prediction model is compared with the actual value. Figure 4 ; Comparison between the output values ​​of the microhardness test samples of the GABP neural network prediction model and the actual values ​​can be found in Figure 5 .

[0104] See also Figure 6 and Figure 7 ,From the prediction results, the trained prediction model has high ,precision and good generalization ability, and can effectively predict and ,analyze the relationship between laser cladding process parameters and ,the cladding layer characteristics.

[0105] Exemplarily, in step S3, the average aspect ratio and microhardness value of the 20 groups of samples obtained in step S1 are compared with the aspect ratio and microhardness value of the cladding layer prepared under the optimal process parameters. See Table 4 for the comparative analysis.

[0106] Table 4 is a comparative analysis of the performance evaluation indicators of the cladding layer

[0107]

[0108]

[0109] It can be seen from the above that the performance evaluation indexes of the cladding layer obtained by the optimal process parameter group optimized by the present invention are better than the average values ​​of 20 groups prepared by CCD test. The improvement rates of aspect ratio and microhardness of the cladding layer are 15.14% and 8.50% respectively.

[0110] For example, in step S3, the cross-sectional morphology of the cladding layer under the optimal process parameter combination is compared and analyzed with the sample with good cladding layer bonding, no obvious defects and better target response value in step S1. According to the comparative analysis results, the macroscopic morphology quality of the cladding layer with the best process parameters is significantly better than that of the comparative test group.

[0111] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, but the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A method for optimizing process parameters of high-speed laser cladding medium-entropy alloy coatings, characterized in that: The steps include: Step S1: taking the process parameters of the high-speed laser cladding medium-entropy alloy coating as input and the aspect ratio and microhardness of the cladding coating as output, a high-speed laser single-pass cladding experiment is performed based on the CCD experimental scheme and relevant data collection is completed; wherein the CCD experimental scheme adopts a three-factor three-level experimental design; Step S2: According to the CCD experimental results, a back propagation neural network prediction model based on the relationship between the process parameters of the high-speed laser cladding medium-entropy alloy coating and the coating quality is established by genetic algorithm optimization, and the accuracy of the prediction model is tested to see whether it meets the requirements for subsequent process parameter optimization; Step S3: Based on the prediction model established in step S2, the laser cladding process parameters are optimized using the second-generation non-dominated sorting genetic algorithm to determine the optimal laser cladding medium entropy alloy coating process parameter combination and conduct experimental verification.

2. The process parameter optimization method for high-speed laser cladding medium-entropy alloy coating according to claim 1 is characterized in that: In step S1, the process parameters of the high-speed laser cladding medium-entropy alloy coating include laser power A, scanning speed B, and powder feeding amount C.

3. The process parameter optimization method for high-speed laser cladding medium-entropy alloy coating according to claim 1 is characterized in that: In step S1, the aspect ratio of the cladding layer is calculated as follows: In the above formula: λ represents the aspect ratio, W represents the melt width, the unit is mm; H represents the melt height, the unit is mm.

4. The process parameter optimization method for high-speed laser cladding medium-entropy alloy coating according to claim 1 is characterized in that: In step S1, the medium entropy alloy powder needs to be mechanically mixed for 20 minutes to 30 minutes before use, and the preset particle size range is 10 μm to 60 μm.

5. The process parameter optimization method for high-speed laser cladding medium-entropy alloy coating according to claim 4 is characterized in that: In the step S1, before the high-speed laser cladding experiment, the surface of the substrate is first ground and polished with sandpaper, and then the pollutants on the surface of the substrate are ultrasonically cleaned with anhydrous ethanol and acetone, and then placed in a ventilated and dry place for use; at the same time, the medium-entropy alloy powder is placed in a vacuum drying oven and dried for 2 hours for use, and the preset temperature is 100°C.

6. The process parameter optimization method for high-speed laser cladding medium-entropy alloy coating according to claim 1 is characterized in that: In step S1, after the cladding experiment is completed, a cladding layer sample is obtained by using a wire cutting machine, and the cross section of the sample is ground, polished and corroded.

7. The process parameter optimization method for high-speed laser cladding medium-entropy alloy coating according to claim 2 is characterized in that: In step S2, a prediction model of a back propagation neural network optimized by a genetic algorithm is established with laser power A, scanning speed B, and powder feeding amount C as inputs and aspect ratio and microhardness of the cladding layer as outputs; Among them, the prediction model needs to normalize the parameters and performance data of each sample using the following formula before training and testing: In the above formula: X' is the value of the sample after normalization; X is the value of the sample before normalization; X max , X min are the maximum and minimum values ​​of the samples respectively; The prediction results of the prediction model need to be denormalized using the following formula to obtain the actual value: In the above formula: y i is the predicted output after sample denormalization; is the predicted output before sample denormalization; max(y) and min(y) are the maximum and minimum values ​​of the sample before denormalization.

8. The process parameter optimization method for high-speed laser cladding medium-entropy alloy coating according to claim 7 is characterized in that: In step S2, the goodness of fit R is used 2 Check whether the accuracy of the prediction model meets the requirements of subsequent multi-objective optimization of process parameters; 2 The calculation method is as follows: In the above formula: n is the number of samples in a single-pass cladding test, y i is the measured value of the single-pass cladding experiment, is the fitting value of the performance index model of a single-pass cladding layer, It is the average value of the measured values ​​obtained from all single-pass cladding experiments.

9. The process parameter optimization method for high-speed laser cladding medium-entropy alloy coating according to claim 8, characterized in that: In step S2, R 2 Greater than 0.

9.

10. The process parameter optimization method for high-speed laser cladding medium-entropy alloy coating according to claim 1, characterized in that: In step S3, the multi-objective optimization mathematical model is as follows: Where, f1 is the aspect ratio prediction model based on GABP neural network, f2 is the microhardness prediction model based on GABP neural network; F represents the nonlinear mapping relationship between the input and output of the second-generation fast non-dominated sorting genetic algorithm optimization model; T is the laser cladding process parameter set, including laser power A, scanning speed B and powder feeding amount C; CP is the optimization object; C i Represents individual C i The process parameters, C i-min , C i-max It is individual C i The minimum and maximum values ​​of various process parameters.

11. The process parameter optimization method for high-speed laser cladding medium-entropy alloy coating according to claim 1, characterized in that: In step S3, a verification test of the optimal laser cladding medium entropy alloy coating process parameters is carried out, the relative error between the predicted value and the actual measured value is calculated, and the quality optimization effect of the high-speed laser cladding layer is analyzed to verify the effectiveness of the optimization method.

Images