Quality optimization method and system for micro-focus digital radiographic inspection of small-diameter pipe welds

Abstract

This invention discloses a quality optimization method and system for microfocus digital radiographic inspection of small-diameter pipe welds in aircraft engines. Using an actual microfocus X-ray source DR inspection system as the target, the process parameter optimization problem is studied. Using CNR and SR as quality indicators, a multivariate nonlinear response regression model is established to model the interactions between multi-objective process parameters and quality indicators. Finally, a non-dominated sorting genetic algorithm is used to determine the optimal process parameter sequence. When performing microfocus X-ray DR inspection of small-diameter pipe welds in aircraft engines, the aforementioned steps can be used to study the effects of different combinations of multi-factor parameters on the spatial resolution and contrast-to-noise ratio of the inspection image. This allows for the direct derivation of optimized DR inspection parameters, improving inspection image quality.

Description

Technical Field

[0001] The present invention belongs to the technical field of micro-focus X-ray DR detection of welds of small-diameter tubes of aircraft engines, and relates to a quality optimization method and system for micro-focus digital ray detection of welds of small-diameter tubes of aircraft engines. Background Art

[0002] With the rapid development of digital X-ray radiography (DR) technology, it is gradually replacing film photography in practical inspections. DR imaging evaluation metrics primarily include spatial resolution (SR), contrast-to-noise ratio (CNR), and signal-to-noise ratio (SNR). Process parameters influencing these metrics include X-ray source focal size, tube voltage, tube current, magnification, and penetration thickness. These process parameters interact with each other to influence inspection metrics. When using microfocus X-ray DR imaging to inspect welds on small-diameter aeroengine tubes, the precise definition of DR inspection parameters is difficult because, compared to conventional X-ray sources, the focal size of microfocus X-ray sources is non-fixed and is related to the source target power. Furthermore, small-diameter aeroengine tubes are characterized by a large number of part numbers, diverse manufacturing materials, diverse welding processes, and widely varying quality requirements due to their varying functions. Other X-ray DR inspection process parameter development methods determine optimal DR imaging process parameters by varying one DR inspection process parameter while maintaining the remaining parameters constant. These methods fail to optimize multiple process parameter combinations and do not describe the relationship between the inspection process and the quality indicators CNR and SR. These methods, which target specific workpieces for optimal imaging process parameters, are not applicable to microfocus X-ray DR inspection of small-diameter pipe welds in aircraft engines. In summary, existing optimization methods are unable to optimize multiple process parameter combinations, describe the relationship between the inspection process and the quality indicators CNR and SR, and therefore suffer from low inspection image quality. Summary of the Invention

[0003] In response to the problems existing in the prior art, the present invention provides a method and system for optimizing the quality of micro-focus digital radiographic inspection of small-diameter tube welds in aircraft engines, thereby optimizing a combination of multiple process parameters, and realizing the relationship between the inspection process and the quality indicators CNR and SR, thereby improving the inspection image quality.

[0004] The present invention is achieved through the following technical solutions:

[0005] A quality optimization method for micro-focus digital radiographic inspection of small diameter pipe welds, comprising:

[0006] S1, determine the process parameters and quality indicators based on the characteristics of micro-focus X-ray DR inspection of small diameter pipe welds;

[0007] S2, based on multivariate nonlinear response regression, a multi-objective response surface regression model between process parameters and quality indicators was established;

[0008] S3, through the multi-objective response surface regression model between quality indicators and process parameters, the optimal solution in the Pareto solution set of the NSGA-II algorithm is obtained;

[0009] S4, select the optimal solution in the Pareto solution set for actual testing and output the optimal DR detection process parameters.

[0010] Preferably, the quality indicators include CNR and SR; SR is spatial resolution, and CNR is contrast-to-noise ratio.

[0011] Preferably, the process parameters include tube voltage, tube current, magnification and transillumination thickness; the tube voltage is used to evaluate the contrast sensitivity of the detection image; the exposure is equal to the product of the tube current and the exposure time, and is used to evaluate the SNR of the detection image; the magnification is used to evaluate the unsharpness of the detection image; the transillumination thickness is used to select the optimal process parameters for small-diameter tubes with different wall thicknesses.

[0012] Preferably, the specific process of establishing a multi-objective response surface regression model between process parameters and quality indicators is as follows:

[0013] S201, determining the experimental factors and constraint intervals of the response surface: The experimental factors of the response surface include tube voltage x1, tube current x2, magnification x3, and transillumination thickness x4. Based on the actual system detection capabilities and engineering experience, the value range of the process parameters to be optimized is determined, thereby determining the constraint intervals.

[0014] S202, response surface experimental design: After level coding the experimental factors of the response surface, construct an experimental plan table to conduct actual experiments and record experimental data;

[0015] S203, response surface regression fitting: performing regression fitting on the test data to obtain a regression equation between the quality index and the process parameters, thereby establishing a relationship model of the interaction between the process parameters and the quality index.

[0016] Preferably, the experimental design method of the response surface experimental design adopts Box-Behnken Design.

[0017] Preferably, the specific process of obtaining the optimal solution of the NSGA-II algorithm is:

[0018] S301, determining the constraint conditions and range of the solution of the regression equation between the quality index and the process parameter according to the actual test conditions;

[0019] S302, determining the settings of NSGA-II algorithm parameters according to actual test conditions;

[0020] S303, using the NSGA-II algorithm to solve the multi-objective response surface regression model between the process parameters and the quality indicators, and obtaining the Pareto optimal solution set of the multi-objective model of the NSGA-II algorithm.

[0021] Preferably, the NSGA-II algorithm parameters include the maximum number of iterations, the population size, the optimal front-end individual sparseness and the fitness function deviation.

[0022] Preferably, the specific process of S303 is: first, the population is initialized and set, and then the initialized population is subjected to non-dominated sorting, selection, crossover, and mutation to generate a parent population. After that, the parent population and the child population are merged into a new population and it is determined whether a new parent population has been generated. If not, a fast non-dominated sort, calculation of congestion, and elite strategy operations are performed to generate a new parent population; otherwise, the generated parent population is subjected to selection, crossover, and mutation operations to generate a child population, and it is determined whether the evolutionary generation number Gen is equal to the maximum evolutionary generation number. If so, the algorithm stops running; otherwise, it continues running until the Pareto optimal solution set of the multi-objective model of the NSGA-II algorithm is obtained.

[0023] Preferably, two sets of optimal solutions in the Pareto optimal solution set of the multi-objective model of the NSGA-II algorithm are selected to verify and analyze the multi-objective response surface regression model between the process parameters and the quality indicators, and the error between the predicted value and the actual value of the model is calculated and compared and analyzed. If the predicted value is consistent with the actual value, it indicates that the test method of the component model is effective; if not, continue to return to reconstruct the multi-objective response surface regression model between the process parameters and the quality indicators until the results of the verification prediction value and the actual value are consistent.

[0024] A small diameter pipe weld micro-focus digital radiographic inspection quality optimization system, comprising:

[0025] Parameter index determination module, used to determine process parameters and quality indicators based on the characteristics of micro-focus X-ray DR inspection of small-diameter pipe welds;

[0026] Model building module, used to establish a multi-objective response surface regression model between process parameters and quality indicators based on multivariate nonlinear response regression;

[0027] The calculation module is used to obtain the optimal solution in the Pareto solution set of the NSGA-II algorithm through a multi-objective response surface regression model between quality indicators and process parameters;

[0028] The test module is used to select the optimal solution in the Pareto solution set for actual testing and output the optimal DR detection process parameters.

[0029] Compared with the prior art, the present invention has the following beneficial technical effects:

[0030] The present invention provides a quality optimization method and system for microfocus digital radiographic inspection of small-diameter pipe welds in aircraft engines. This method addresses the characteristics of microfocus X-ray DR (DDR) inspection of small-diameter pipe welds in aircraft engines, using an actual microfocus X-ray source DR inspection system as the target, to study process parameter optimization. Using CNR and SR as quality indicators, a multi-objective nonlinear response regression model is established to model the interactions between multi-objective process parameters and quality indicators. Finally, a nondominated sorting genetic algorithm (NSGA-II) is used to determine the optimal process parameter sequence. This multi-objective optimization model structure for small-diameter pipe weld inspection process parameters is constructed, thereby optimizing multiple process parameter combinations and clarifying the relationship between the inspection process and the quality indicators CNR and SR, thereby improving inspection image quality. When performing micro-focus X-ray DR inspection on weld seams of small-diameter tubes of aircraft engines, the above steps can be used for processing, and the effects of different combinations of multi-factor parameters on the spatial resolution and contrast-to-noise ratio of the detection image can be studied, and optimized DR detection process parameters can be formulated. This can save the detection process test cost for radiographic inspection of weld seams of small-diameter tubes of aircraft engines of various specifications. The method of the present invention is used for testing, which is suitable for micro-focus X-ray DR inspection of weld seams of small-diameter tubes of aircraft engines, and can directly derive optimized DR detection parameters, thereby improving the quality of the detection image. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Flowchart of the multi-objective optimization model for process parameters of small diameter weld inspection;

[0032] Figure 2 Schematic diagram of the Pareto optimal solution set in the embodiment;

[0033] Figure 3 Graph comparing CNR and predicted value with actual value in the embodiment. DETAILED DESCRIPTION

[0034] The present invention will be further described in detail below with reference to specific embodiments, which are intended to explain the present invention rather than to limit it.

[0035] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0036] A quality optimization method for micro-focus digital radiographic inspection of small diameter pipe welds includes:

[0037] S1. Based on the characteristics of microfocus X-ray DR inspection of small-diameter pipe welds, process parameters and quality indicators were determined. The CNR (Contrast Ratio) indicator, consisting of contrast and signal-to-noise ratio (SNR), characterizes the detail recognition capability of the inspection image, while the SR indicator indicates the minimum spacing between details required to resolve the inspection image. CNR and SR are fundamental evaluation metrics for DR inspection images. Therefore, a multi-objective optimization approach was conducted with the goal of improving the CNR and SR of the inspection image.

[0038] S2, based on multivariate nonlinear response regression, a multi-objective response surface regression model between process parameters and quality indicators was established;

[0039] S3, through the multi-objective response surface regression model between quality indicators and process parameters, the optimal solution in the Pareto solution set of the NSGA-II algorithm is obtained;

[0040] S4, select the optimal solution in the Pareto solution set for actual testing and output the optimal DR detection process parameters.

[0041] The quality indicators include CNR and SR; SR is spatial resolution, and CNR is contrast-to-noise ratio.

[0042] The process parameters include tube voltage, tube current, magnification, and transillumination thickness. Tube voltage is used to evaluate the contrast sensitivity of the detected image. Exposure, which is equal to the product of tube current and exposure time, is used to evaluate the SNR of the detected image. Magnification is used to evaluate the blur of the detected image. Transillumination thickness is used to select the optimal process parameters for small-diameter tubes of different wall thicknesses. Tube voltage, tube current, magnification, and transillumination thickness were selected as the four process parameters with the greatest impact on image quality for study, while other influencing parameters remained the same. Tube voltage is a basic transillumination parameter for DR detection, and its magnitude directly affects contrast sensitivity. Exposure, another important transillumination parameter in DR detection, is equal to the product of tube current and exposure time. It refers to the radiation dose reaching the detector and directly affects the SNR of the detected image. Magnification is a key parameter that affects the blur of the detected image. Choosing a reasonable magnification can help improve the SR of the detected image. When conducting digital radiographic inspection of welds on small-diameter aero-engine tubes, the wall thicknesses of different small-diameter tubes vary, and therefore the required penetration thicknesses are also inconsistent. The penetration thickness determines the actual tube voltage and image quality meter rating, and also affects the penetration thickness ratio and effective penetration area. Therefore, penetration thickness needs to be used as a process parameter to select the optimal process parameters for small-diameter tubes of varying wall thicknesses.

[0043] The specific process of establishing the multi-objective response surface regression model between process parameters and quality indicators is as follows:

[0044] S201, determining the experimental factors and constraint intervals of the response surface: The experimental factors of the response surface include tube voltage x1, tube current x2, magnification x3, and transillumination thickness x4. Based on the actual system detection capabilities and engineering experience, the value range of the process parameters to be optimized is determined, thereby determining the constraint intervals.

[0045] S202, response surface experimental design: After level coding the experimental factors of the response surface, construct an experimental plan table to conduct actual experiments and record experimental data;

[0046] S203, response surface regression fitting: performing regression fitting on the test data to obtain a regression equation between the quality index and the process parameters, thereby establishing a relationship model of the interaction between the process parameters and the quality index.

[0047] The experimental design method of the response surface experimental design adopts Box-Behnken Design.

[0048] The specific process of obtaining the optimal solution of the NSGA-II algorithm is as follows:

[0049] S301, determining the constraint conditions and range of the solution of the regression equation between the quality index and the process parameter according to the actual test conditions;

[0050] S302, determining the settings of NSGA-II algorithm parameters according to actual test conditions;

[0051] S303, using the NSGA-II algorithm to solve the multi-objective response surface regression model between the process parameters and the quality indicators, and obtaining the Pareto optimal solution set of the multi-objective model of the NSGA-II algorithm.

[0052] The NSGA-II algorithm parameters include the maximum number of iterations, the population size, the optimal front-end individual sparseness and the fitness function deviation.

[0053] The specific process of S303 is as follows: first, the population is initialized and set, and then the initialized population is subjected to non-dominated sorting, selection, crossover, and mutation to generate a parent population. After that, the parent population and the child population are merged into a new population and it is determined whether a new parent population has been generated. If not, a fast non-dominated sort, calculation of congestion, and elite strategy operations are performed to generate a new parent population; otherwise, the generated parent population is subjected to selection, crossover, and mutation operations to generate a child population. It is determined whether the evolutionary generation number Gen is equal to the maximum evolutionary generation number. If so, the algorithm stops running; otherwise, it continues running until the Pareto optimal solution set of the multi-objective model of the NSGA-II algorithm is obtained.

[0054] The multi-objective response surface regression model between process parameters and quality indicators was verified and analyzed. Two sets of optimization solutions from the Pareto optimal solution set of the multi-objective model of the NSGA-II algorithm were selected. The errors between the predicted values and the actual values of the model were calculated and compared and analyzed. If the predicted values were consistent with the actual values, it indicated that the experimental method of constructing the multi-objective response surface regression model between process parameters and quality indicators was effective. If they were not consistent, the multi-objective response surface regression model between process parameters and quality indicators was reconstructed until the results of verification that the predicted values were consistent with the actual values were consistent.

[0055] A small diameter pipe weld micro-focus digital radiographic inspection quality optimization system, comprising:

[0056] Parameter index determination module, used to determine process parameters and quality indicators based on the characteristics of micro-focus X-ray DR inspection of small-diameter pipe welds;

[0057] Model building module, used to establish a multi-objective response surface regression model between process parameters and quality indicators based on multivariate nonlinear response regression;

[0058] The calculation module is used to obtain the optimal solution in the Pareto solution set of the NSGA-II algorithm through a multi-objective response surface regression model between quality indicators and process parameters;

[0059] Experimental module: used to select the optimal solution in the Pareto solution set for actual testing and output the optimal DR detection process parameters.

[0060] like Figure 2 As shown in the figure, the micro-focus X-ray DR inspection method for small diameter pipe welds of aircraft engines: taking the penetration thickness of 1mm, 1.5mm, and 2mm as an example, the steps are:

[0061] (1) Determination of detection process parameters: ① Determine the target to be optimized, i.e., the quality index: The quality of digital radiographic detection images is mainly characterized by three indicators: contrast, SR, and SNR. The CNR index, which is composed of contrast and SNR, represents the detail recognition ability of the detection image, and the SR index represents the minimum spacing of the detection image to distinguish details. CNR and SR are the basic evaluation indicators of DR detection images. Therefore, multi-objective optimization is performed with the improvement of the CNR and SR of the detection image as the optimization target. ② Determine process parameters: The four process parameters that have the greatest impact on the imaging quality, namely, tube voltage, tube current, magnification, and penetration thickness, are selected for research, while the other influencing parameters remain the same. Tube voltage is the basic penetration parameter of DR detection, and the size of the tube voltage directly affects the contrast sensitivity. Exposure is another important penetration parameter in DR detection. Exposure is equal to the product of tube current and exposure time, which refers to the radiation dose reaching the detector, which directly affects the SNR of the detection image. Magnification is an important parameter that affects the unsharpness of the detection image. Reasonable selection of magnification is conducive to improving the SR of the detection image. When conducting digital radiographic inspection of welds on small-diameter aero-engine tubes, the wall thicknesses of different small-diameter tubes vary, and therefore the required penetration thicknesses are also inconsistent. The penetration thickness determines the actual tube voltage and image quality meter rating, and also affects the penetration thickness ratio and effective penetration area. Therefore, penetration thickness needs to be used as a process parameter to select the optimal process parameters for small-diameter tubes of varying wall thicknesses.

[0062] (2) Establish the regression equation between quality indicators and process parameters: ① Determine the experimental factors and constraint intervals. The experimental factors of the response surface are: tube voltage x1, tube current x2, magnification x3 and transillumination thickness x4. According to the detection capability of the actual system and engineering experience, determine the value range of the process parameters that need to be optimized. ② Response surface experimental design: After level coding the experimental factors, determine the experimental design method as Box-Behnken Design (BBD), the total number of experiments is 29, the number of factors is 4 and the number of factor levels is 3, and construct the experimental plan table for actual experiments. ③ Response surface regression fitting: Perform regression fitting on the experimental data to obtain the regression equation between quality indicators and process parameters. ;

[0063] (3) Obtain the optimal solution of the NSGA-II algorithm: ① Determine the constraints of the optimization equation: Determine the constraints and range of the solution based on the actual experimental conditions. ② NSGA-II algorithm parameter setting: The parameters of the NSGA-II algorithm include the maximum number of iterations, the number of populations, the sparseness of the optimal front-end individuals, and the fitness function deviation. Determine the algorithm parameters based on actual conditions. ③ NSGA-II algorithm solution: The NSGA-II algorithm first initializes and sets the population, and then performs non-dominated sorting, selection, crossover, and mutation on the initial population to generate the first generation population. Afterwards, the parent population and the child population are merged into a new population and it is determined whether a new parent population has been generated. If not, a fast non-dominated sorting, calculation of congestion, elite strategy, and other operations are performed to generate a new parent population; otherwise, the generated parent population is subjected to selection, crossover, mutation, and other operations to generate a child population. Determine whether the evolutionary generation number Gen is equal to the maximum evolutionary generation number. If it is satisfied, the algorithm stops running, otherwise it continues running. ④ Obtain the Pareto optimal solution;

[0064] (4) Model verification and analysis: Verify the CNR and R established by the analysis im The mathematical model is used to calculate the error between the model and the actual value, R im is the spatial frequency, and the contrast-to-noise ratio (CNR), such as Figure 3 As shown in the figure, the multi-objective response surface regression model between process parameters and quality indicators was verified and analyzed, and the error between the CNR predicted value of the model and the actual CNR value was calculated, R im Predicted Value and R im The error between the actual values is compared and analyzed. It can be seen that the predicted values and the actual values are distributed in the same line area, the error is small, and the predicted values and the actual values can be consistent, which shows that the experimental method of the constructed multi-objective model is effective;

[0065] (5) The process parameters of the optimal solution in the Pareto solution set are selected for actual experiments, and the optimal DR detection process parameters are output, see Table 1.

[0066] Table 1 shows the optimal DR detection process parameters for different small diameter tubes;

[0067]

[0068] Advantages of this method: When micro-focus X-ray DR inspection of small-diameter pipe welds of aircraft engines is performed, the above steps can be used to study the influence of different combinations of multi-factor parameters on the spatial resolution and contrast-to-noise ratio of the inspection image, and to formulate optimized DR inspection process parameters. For the X-ray inspection of small-diameter pipe welds of aircraft engines of various specifications, the inspection process test cost can be saved. By using this method and conducting experiments according to the above steps (1) to (4), the optimized DR inspection parameters can be directly obtained, thereby improving the quality of the inspection image.

[0069] It should be noted that the terms in the specification and claims of the present invention and the above-mentioned drawings include a process, method, system, product or device that includes a series of steps or units, which are not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or are inherent to these processes, methods, products or devices.

[0070] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention pertains. The terms used in this specification of the present invention are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0071] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any ordinary technician in this industry can smoothly implement the present invention as shown in the drawings and described above. However, any equivalent changes, modifications and evolutions made by technicians familiar with this profession without departing from the scope of the technical solution of the present invention using the technical content disclosed above are all equivalent embodiments of the present invention. At the same time, any equivalent changes, modifications and evolutions made to the above embodiments based on the essential technology of the present invention are still within the scope of protection of the technical solution of the present invention.

Claims

1. A quality optimization method for micro-focus digital radiographic inspection of small diameter pipe welds, characterized in that: include, S1, determine the process parameters and quality indicators based on the characteristics of micro-focus X-ray DR inspection of small diameter pipe welds; S2, based on multivariate nonlinear response regression, a multi-objective response surface regression model between process parameters and quality indicators was established; S3, through the multi-objective response surface regression model between quality indicators and process parameters, the optimal solution in the Pareto solution set of the NSGA-II algorithm is obtained; S4, select the optimal solution in the Pareto solution set to conduct actual experiments and output the optimal DR detection process parameters; The quality indicators include CNR and SR; SR is spatial resolution, and CNR is contrast-to-noise ratio; The specific process of establishing the multi-objective response surface regression model between process parameters and quality indicators is as follows: S201, Determination of experimental factors and constraint intervals of response surface: The experimental factors of response surface include tube voltage , tube current , magnification and transillumination thickness ; Based on the actual system's detection capabilities and engineering experience, determine the value range of the process parameters that need to be optimized, thereby determining the constraint interval; S202, Response Surface Experimental Design: After level coding the experimental factors of the response surface, construct an experimental plan table to conduct the actual experiment and record the experimental data; S203, response surface regression fitting: performing regression fitting on the test data to obtain a regression equation between the quality index and the process parameters, thereby establishing a relationship model of the interaction between the process parameters and the quality index; The specific process of obtaining the optimal solution of the NSGA-II algorithm is as follows: S301, determining the constraint conditions and range of the solution of the regression equation between the quality index and the process parameter according to the actual test conditions; S302, determining the settings of NSGA-II algorithm parameters according to actual test conditions; S303, solving the multi-objective response surface regression model between process parameters and quality indicators using the NSGA-II algorithm, and obtaining the Pareto optimal solution set of the multi-objective model of the NSGA-II algorithm; Two optimal solutions from the Pareto optimal solution set of the multi-objective model of the NSGA-II algorithm are selected to verify and analyze the multi-objective response surface regression model between process parameters and quality indicators. The error between the predicted value and the actual value of the model is calculated and compared. If the predicted value is consistent with the actual value, it indicates that the test method of the component model is effective. If they do not match, continue to return to rebuild the multi-objective response surface regression model between process parameters and quality indicators until the results of verification that the predicted values are consistent with the actual values.

2. The method for optimizing the quality of micro-focus digital radiographic inspection of small-diameter pipe welds according to claim 1, characterized in that: The process parameters include tube voltage, tube current, magnification and transillumination thickness; the tube voltage is used to evaluate the contrast sensitivity of the detection image; the exposure is equal to the product of the tube current and the exposure time, and is used to evaluate the SNR of the detection image; the magnification is used to evaluate the unsharpness of the detection image; the transillumination thickness is used to select the optimal process parameters for small-diameter tubes with different wall thicknesses.

3. The method for optimizing the quality of micro-focus digital radiographic inspection of small-diameter pipe welds according to claim 1, characterized in that: The experimental design method of the response surface experimental design adopts Box-Behnken Design.

4. The method for optimizing the quality of micro-focus digital radiographic inspection of small-diameter pipe welds according to claim 1, characterized in that: The NSGA-II algorithm parameters include the maximum number of iterations, the population size, the optimal front-end individual sparseness and the fitness function deviation.

5. The method for optimizing the quality of micro-focus digital radiographic inspection of small-diameter pipe welds according to claim 1, characterized in that: The specific process of S303 is as follows: first, the population is initialized and set, and then the initialized population is subjected to non-dominated sorting, selection, crossover, and mutation to generate a parent population. After that, the parent population and the child population are merged into a new population and it is determined whether a new parent population has been generated. If not, a fast non-dominated sort, congestion calculation, and elite strategy are performed to generate a new parent population. Otherwise, the generated parent population is subjected to selection, crossover, and mutation to generate a child population. It is determined whether the evolutionary generation number Gen is equal to the maximum evolutionary generation number. If so, the algorithm stops running. Otherwise, it continues running until the Pareto optimal solution set of the multi-objective model of the NSGA-II algorithm is obtained.

6. A small diameter pipe weld micro-focus digital radiographic inspection quality optimization system, based on the digital radiographic inspection quality optimization method according to any one of claims 1 to 5, characterized in that: include, Parameter index determination module, used to determine process parameters and quality indicators based on the characteristics of micro-focus X-ray DR inspection of small-diameter pipe welds; Model building module, used to establish a multi-objective response surface regression model between process parameters and quality indicators based on multivariate nonlinear response regression; The calculation module is used to obtain the optimal solution in the Pareto solution set of the NSGA-II algorithm through a multi-objective response surface regression model between quality indicators and process parameters; The test module is used to select the optimal solution in the Pareto solution set for actual testing and output the optimal DR detection process parameters.

Images