Multi-objective collaborative optimization method for aircraft assembly drilling process parameters

Abstract

The invention discloses a multi-objective collaborative optimization method for aircraft assembly hole-making process parameters, which is suitable for reasonably selecting and matching key process parameters influencing hole-making precision, and comprises the following steps: determining a required optimization objective according to a manufacturing index of a hole-making process, identifying a process parameter comprehensively influencing the quality index as a global variable; establishing a multi-parameter variable single factor experiment design scheme, and determining the influence trend of each single variable on the optimization target; on the basis of statistical analysis, obtaining the influence weight of each process parameter and the interaction of each process parameter on the optimization target, and screening key process variables on the basis of weight sorting; establishing a technological parameter multi-order response surface model surrounding an optimization target, and performing multi-target optimization on the established regression model through an improved genetic algorithm; and determining a global optimal aircraft assembly drilling process parameter domain, and verifying the accuracy of the global optimal aircraft assembly drilling process parameter domain.

Description

Technical Field

[0001] The present invention relates to the field of automated hole making technology, but is not limited to it, and in particular to a multi-objective collaborative optimization method for aircraft assembly hole making process parameters. Background Art

[0002] In aviation automated assembly, in order to ensure the quality of the overall processing holes of the material, ensure assembly continuity, and improve hole-making efficiency, aviation laminated materials are generally drilled in multiple layers at one time, and no longer processed separately. However, compared with single-layer material hole-making, the processing mechanism of laminated materials is more complicated and is affected by many factors in the actual processing process.

[0003] Considering the diversity of manufacturing indicators for different positions or different products of the same product, the process plan needs to be dynamically adjusted to meet the requirements of different indicators. In addition, due to the complex, time-varying, nonlinear, and strongly coupled relationship between process parameters and quality indicators, it is particularly difficult to reasonably obtain a set of optimal process parameters. However, it is difficult to quickly achieve the optimization of process parameter domains for multiple manufacturing indicators at the same time in the existing technology. Summary of the invention

[0004] The purpose of the present invention is: in order to solve the above problems, an embodiment of the present invention provides a multi-objective collaborative optimization method for aircraft assembly hole making process parameters to solve the problem in the prior art that it is difficult to quickly achieve simultaneous optimization of process parameter domains for multiple manufacturing indicators.

[0005] The technical solution of the present invention is as follows: The embodiment of the present invention provides a multi-objective collaborative optimization method for aircraft assembly hole making process parameters, comprising:

[0006] Step 1, according to the manufacturing index of the aircraft assembly hole making process, multiple optimization targets of the aircraft assembly hole making are determined, and the process parameters corresponding to each optimization target are globally identified;

[0007] Step 2: Establish a single-factor experimental scheme for all process parameters identified in step 1, quantitatively characterize the experimental results through each single-factor experiment, determine the influence of each single process parameter on its corresponding optimization objectives, and calibrate the value boundary of each process parameter;

[0008] Step 3: Based on statistical analysis, the influence weight of each process parameter and the interaction between process parameters on each optimization target is obtained, and key process variables are selected based on the weight ranking;

[0009] Step 4, establishing a multi-order response surface model of process parameters that meets multiple optimization objectives of aircraft assembly hole making, and optimizing multiple optimization objectives by performing multiple iterations of the multi-order response surface model based on an improved genetic algorithm to determine the global optimal hole making process parameter domain;

[0010] Step 5, verifying the accuracy of the global optimal hole-making process parameter domain determined in step 4, and obtaining the optimal process parameter combination that simultaneously meets multiple optimization objectives.

[0011] Optionally, in the above-mentioned multi-objective collaborative optimization method for aircraft assembly hole making process parameters, a single factor experiment scheme is implemented in step 2, and the adjustment form of the process parameters in different single factor experiments requires that the aircraft assembly hole making system has an adjustable and controllable hardware module, including:

[0012] The pressure foot module is used to adjust the pressure foot to fit the surface of the hole-making product according to the preset clamping force value, and the pressure value is 10-3000N;

[0013] The spindle module is used to adjust the rotation speed of the hole-making spindle according to the preset value. The speed change delay is no more than 10ms, and the value range is 1000-30000rpm;

[0014] The spindle module is also used to adjust the feed speed of the hole-making spindle in the axial direction according to a preset value, with a speed change delay of no more than 10ms and a value range of 10-300mm / min.

[0015] Optionally, in the multi-objective collaborative optimization method for aircraft assembly hole making process parameters as described above,

[0016] The optimization target includes at least one of the following: morphology accuracy, hole position accuracy, processing efficiency, automatic hole making rate, and hole position layout; wherein the morphology accuracy includes at least one of hole making diameter, countersink diameter, and countersink size discreteness.

[0017] Optionally, in the multi-objective collaborative optimization method for aircraft assembly hole making process parameters as described above,

[0018] The process parameters include at least one of the following: clamping force, spindle speed, spindle feed speed, product stack thickness, stack bonding surface gap distance, product material, product curvature, tool length, tool material, tool cutting angle, tool tip angle; the spindle speed, drilling speed and countersinking speed are divided according to time periods, and the time periods include drilling feed speed and countersinking feed speed.

[0019] Optionally, in the above-mentioned multi-objective collaborative optimization method for aircraft assembly hole making process parameters, step 2 comprises:

[0020] Step 21, based on the control variable method, for each process parameter, only one of the process parameters is changed in each single factor experiment, and the other process parameters remain unchanged;

[0021] Step 22, repeating step 21 for all process parameters in sequence until a single factor experiment for each process parameter is completed;

[0022] Step 23, through the experimental results of step 21 and step 22, obtain the influence law of each single process parameter on each optimization target, and obtain the value boundary of each process parameter according to the requirements of each optimization target.

[0023] Optionally, in the above-mentioned multi-objective collaborative optimization method for aircraft assembly hole making process parameters, step 3 comprises:

[0024] Step 31, performing a normality test on the experimental data obtained through the single factor experiment in step 2 to ensure that the experimental data of all single factor experiments meet the analysis conditions;

[0025] Step 32, using the AVOVA method to calculate the F value of each process parameter and the interaction between the process parameters and the P value used to represent the probability of significance; wherein the F value is the ratio of the mean square sum to the mean square sum of errors;

[0026] Step 33, determine the size of the P value and the standard threshold of the normality test. If P<threshold, it is considered that the process parameter has a significant impact on the optimization target, and the size of the F value corresponding to the P value represents the weight;

[0027] Step 34, based on the P value of each process parameter and the interaction between the process parameters, determine whether each process parameter has a significant impact on each optimization target. If no significant impact is produced, the corresponding process parameter does not participate in the subsequent optimization steps; if a significant impact is produced, the influence weight is determined according to the F value corresponding to the P value, and the influence weight is sorted. If the influence weight accounts for less than 1%, it can be ignored, and the corresponding process parameter does not participate in the subsequent optimization steps.

[0028] Optionally, in the above-mentioned multi-objective collaborative optimization method for aircraft assembly hole making process parameters, the method of establishing a multi-order response surface model of process parameters in step 4 includes:

[0029] The Box-Behnken experimental design method was used to design the response surface experimental plan. The process parameters were set as response factors, and the optimization target was set as the response result. The multivariate multi-order response surface model established was:

[0030]

[0031] Among them, a0 is a constant term; a i is the process parameter x i The coefficient of the first-order term; a ij is the process parameter x i and x j The interaction coefficient ofii is the process parameter x i The quadratic coefficient of x i is the i-th process parameter, x j is the jth process parameter; n is the number of process parameters; Y is the parameter characterizing the shape accuracy of the hole profile of the aircraft assembly;

[0032] A response surface experiment database is formed through the response surface experiment, including: experimental data of each group of response surface experiments and corresponding Y data.

[0033] Optionally, in the above-mentioned multi-objective collaborative optimization method for aircraft assembly hole making process parameters, the method of implementing optimization of multiple optimization objectives by performing multiple iterations based on the improved genetic algorithm in step 4 includes:

[0034] Step 41, analyzing the influence of each process parameter on each optimization target through regression analysis, and determining whether the influence of each process parameter on each optimization target is a positive effect or a negative effect;

[0035] Step 42, testing the significance of different process parameters for each optimization target through variance analysis, quantifying the weight of the significance of the process parameters, and using this as a basis for simplifying the multivariate multi-order response surface model;

[0036] Step 43, based on the response surface experiment database formed by the response surface experiment, the process parameters of the single optimization target with the largest contribution are first optimized by improving the genetic algorithm, and then the process parameters of the single optimization target with the secondary contribution are optimized on the optimal solution set, and so on, until all the process parameters corresponding to the single optimization targets obtain the corresponding optimization solution sets, thereby realizing the global optimal hole-making process parameter domain.

[0037] Optionally, in the above-mentioned multi-objective collaborative optimization method for aircraft assembly hole making process parameters, the way of verifying the hole making process parameter domain in step 5 is: verifying the accuracy of the response surface regression equation fitting by testing the residual normal plot of the standardized effect, including:

[0038] Determine whether the residual distribution after fitting the multi-order response surface model of process parameters is uniform. If each sample data point satisfies the standard diagonal linear distribution and the distance deviation value is within 5%, it is judged to pass the residual normal plot test and the multi-order response surface model is valid.

[0039] Beneficial effects of the present invention: In view of the problem that the aircraft assembly hole making process scheme is difficult to quickly achieve the optimization of the process parameter domain for multiple manufacturing indicators at the same time, the embodiment of the present invention provides a multi-objective collaborative optimization method for aircraft assembly hole making process parameters. The method uses a multi-level experimental design to identify the response characteristics and contribution weights of each process parameter layer by layer, and constructs a prediction model that takes into account multiple optimization objectives to achieve accurate locking of the hole making process parameter domain. The multi-objective collaborative optimization method provided by the embodiment of the present invention can solve the problem of multi-indicator collaborative optimization at a lower experimental cost, and use intelligent algorithms and optimization technologies to ensure the manufacturing accuracy of the aircraft assembly hole making process. It is an efficient, intelligent, and highly flexible analysis method based on data-driven. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings are used to provide a further understanding of the technical solution of the present invention and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the technical solution of the present invention and do not constitute a limitation on the technical solution of the present invention. ;

[0041] Figure 1 A flowchart of a multi-objective collaborative optimization method for aircraft assembly hole making process parameters is provided for an embodiment of the present invention;

[0042] Figure 2 It is a contour map of response surfaces of various parameters based on response surface optimization in an embodiment of the present invention;

[0043] Figure 3 : It is the residual normal probability diagram of the discreteness of the dimple size after hole making in the embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solution and advantages of the present invention more clear, the embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other arbitrarily without conflict.

[0045] The above background technology has already explained the technical difficulties of hole making in aviation laminated materials, and the difficulty in the prior art to quickly achieve the optimization of process parameter domains for multiple manufacturing indicators at the same time.

[0046] The traditional optimization method usually relies on manual experience to continuously adjust the process parameters to obtain the ideal process parameter combination, but this method is relatively blind, which will greatly prolong the time of process verification, increase the risk of product service, and cause cost increase and waste of raw materials. Therefore, how to scientifically and quickly obtain a process parameter combination that meets the multiple quality requirements of the assembly hole making process has great practical significance for the automated manufacturing of aircraft. In response to this demand, an embodiment of the present invention provides a multi-objective collaborative optimization method for aircraft assembly hole making process parameters.

[0047] The present invention provides the following specific embodiments which can be combined with each other. The same or similar concepts or processes may not be described in detail in some embodiments.

[0048] Figure 1 A flowchart of a multi-objective collaborative optimization method for aircraft assembly hole making process parameters provided by an embodiment of the present disclosure. Figure 1 As shown, the multi-objective collaborative optimization method provided by the embodiment of the present invention includes the following steps:

[0049] Step 1, according to the manufacturing index of the aircraft assembly hole making process, multiple optimization targets of the aircraft assembly hole making are determined, and the process parameters corresponding to each optimization target are globally identified;

[0050] Step 2: Establish a single-factor experimental scheme for all process parameters identified in step 1, quantitatively characterize the experimental results through each single-factor experiment, determine the influence of each single process parameter on its corresponding optimization objectives, and calibrate the value boundary of each process parameter;

[0051] Step 3: Based on statistical analysis, the influence weight of each process parameter and the interaction between process parameters on each optimization target is obtained, and key process variables are selected based on the weight ranking;

[0052] Step 4, establishing a multi-order response surface model of process parameters that meets multiple optimization objectives of aircraft assembly hole making, and optimizing multiple optimization objectives by performing multiple iterations of the multi-order response surface model based on an improved genetic algorithm to determine the global optimal hole making process parameter domain;

[0053] Step 5, verifying the accuracy of the global optimal hole-making process parameter domain determined in step 4, and obtaining the optimal process parameter combination that simultaneously meets multiple optimization objectives.

[0054] In order to implement each single factor experimental scheme in step 2 of the embodiment of the present invention, the process parameters need to be adjusted in different single factor experiments. Therefore, based on the adjustment requirements of the process parameters, the aircraft assembly hole making system is required to have an adjustable and controllable hardware module, specifically including:

[0055] The pressure foot module is used to adjust the pressure foot to fit the surface of the hole-making product according to the preset clamping force value, and the pressure value is 10-3000N;

[0056] The spindle module is used to adjust the rotation speed of the hole-making spindle according to the preset value. The speed change delay is no more than 10ms, and the value range is 1000-30000rpm;

[0057] The spindle module is also used to adjust the feed speed of the hole-making spindle in the axial direction according to the preset value. The speed change delay is no more than 10ms, and the value range is 10-300mm / min.

[0058] In one implementation of an embodiment of the invention, the optimization target determined in step 1 includes at least one of the following: morphology accuracy, hole position accuracy, processing efficiency, automatic hole making rate, and hole position layout; wherein the morphology accuracy includes at least one of the hole making diameter, countersink diameter, and countersink size discreteness.

[0059] Accordingly, the process parameters determined in the implementation include at least one of the following: clamping force, spindle speed, spindle feed speed, product stack thickness, stack gap distance, product material, product curvature, tool length, tool material, tool cutting angle, tool tip angle. It should be noted that the spindle speed is divided into drilling speed and countersinking speed according to the time period, and similarly, the time period includes drilling feed speed and countersinking feed speed.

[0060] In one implementation of the embodiment of the invention, the implementation process of the above step 2 may include:

[0061] Step 21, based on the control variable method, for each process parameter, only one of the process parameters is changed in each single factor experiment, and the other process parameters remain unchanged;

[0062] Step 22, repeating step 21 for all process parameters in sequence until a single factor experiment for each process parameter is completed;

[0063] Step 23, through the experimental results of step 21 and step 22, obtain the influence law of each single process parameter on each optimization target, and obtain the value boundary of each process parameter according to the requirements of each optimization target.

[0064] In an optional implementation of an embodiment of the invention, the statistical analysis method used in the above step 3 may include but is not limited to variance analysis, regression analysis, and range analysis to quantify the influence of each process parameter and its interaction on the optimization target, and screen key process variables based on weight ranking.

[0065] In one implementation of the embodiment of the invention, the implementation of step 3 is described by taking regression analysis as an example. The above step 3 includes:

[0066] Step 31, performing a normality test on the experimental data obtained through the single factor experiment in step 2 (including the quantitative characterization of the experimental results by the process parameters of each group of experiments) to ensure that the experimental data structure of all single factor experiments meets the analysis conditions;

[0067] Step 32, using the AVOVA method to calculate the F value of each process parameter and the interaction between the process parameters and the P value used to represent the probability of significance; wherein the F value is the ratio of the mean square sum to the mean square sum of errors;

[0068] Step 33, determine the size of the P value and the standard threshold of the normality test. If P<threshold, it is considered that the process parameter has a significant impact on the optimization target, and the size of the F value corresponding to the P value represents the weight;

[0069] Step 34, based on the P value of each process parameter and the interaction between the process parameters, determine whether each process parameter has a significant impact on each optimization target. If no significant impact is produced, the corresponding process parameter does not participate in the subsequent optimization steps; if a significant impact is produced, the influence weight is determined according to the F value corresponding to the P value, and the influence weight is sorted. If the influence weight accounts for less than 1%, it can be ignored, and the corresponding process parameter does not participate in the subsequent optimization steps.

[0070] In one implementation of the embodiment of the invention, the method of establishing the multi-order response surface model of process parameters in the above step 4 may include:

[0071] The Box-Behnken experimental design method was used to design the response surface experimental plan. The process parameters were set as response factors, and the optimization target was set as the response result. The multivariate multi-order response surface model established was:

[0072]

[0073] Among them, a0 is a constant term; a i is the process parameter x i The coefficient of the first-order term; a ij is the process parameter x i and x j The interaction coefficient of ii is the process parameter x i The quadratic coefficient of x i is the i-th process parameter, x j is the jth process parameter; n is the number of process parameters; Y is the parameter characterizing the shape accuracy of the hole profile of the aircraft assembly;

[0074] In this step, a response surface experiment database is formed through response surface experiments, including: experimental data of each group of response surface experiments and corresponding Y data.

[0075] Furthermore, in this implementation, the optimization method of implementing multiple optimization objectives through multiple iterations based on the improved genetic algorithm in step 4 may include:

[0076] Step 41, analyzing the influence of each process parameter on each optimization target through regression analysis, and determining whether the influence of each process parameter on each optimization target is a positive effect or a negative effect;

[0077] Step 42, testing the significance of different process parameters for each optimization target through variance analysis, quantifying the weight of the significance of the process parameters, and using this as a basis for simplifying the multivariate multi-order response surface model;

[0078] Step 43, based on the response surface experiment database formed by the response surface experiment, the process parameters of the single optimization target with the largest contribution are first optimized by improving the genetic algorithm, and then the process parameters of the single optimization target with the secondary contribution are optimized on the optimal solution set, and so on, until all the process parameters corresponding to the single optimization targets obtain the corresponding optimization solution sets, thereby realizing the global optimal hole-making process parameter domain.

[0079] In one implementation of the embodiment of the invention, the method of verifying the hole-making process parameter domain in the above step 5 is: verifying the accuracy of the response surface regression equation fitting by testing the residual normal plot of the standardized effect, including:

[0080] Determine whether the residual distribution after fitting the multi-order response surface model of process parameters is uniform. If each sample data point satisfies the standard diagonal linear distribution and the distance deviation value is within 5%, it is judged to pass the residual normal plot test and the multi-order response surface model is valid.

[0081] In view of the problem that it is difficult to quickly achieve the optimization of process parameter domains for multiple manufacturing indicators at the same time in the aircraft assembly hole making process scheme, an embodiment of the present invention provides a multi-objective collaborative optimization method for aircraft assembly hole making process parameters. The method uses a multi-level experimental design to identify the response characteristics and contribution weights of each process parameter layer by layer, and constructs a prediction model that takes into account multiple optimization objectives to achieve accurate locking of the hole making process parameter domain. The multi-objective collaborative optimization method provided by the embodiment of the present invention can solve the problem of multi-indicator collaborative optimization at a lower experimental cost, and use intelligent algorithms and optimization technologies to ensure the manufacturing accuracy of the aircraft assembly hole making process. It is an efficient, intelligent, and highly flexible analysis method based on data-driven.

[0082] The following is a schematic illustration of the implementation of the multi-objective collaborative optimization method for aircraft assembly hole making process parameters provided by an embodiment of the present invention through a specific implementation case.

[0083] This implementation case takes the shape accuracy of the hole profile after hole making in aircraft assembly as the response index, and uses various process parameters that affect the response index, including clamping force, drilling speed, countersinking speed, hole making feed speed, and countersinking feed speed, as response factors to conduct multi-objective collaborative optimization. Figure 1As shown, the method provided by the embodiment of the present invention includes the following steps:

[0084] Step 1: According to the manufacturing index of the aircraft assembly hole making process, the optimization target of the process parameters is determined, and the process parameters corresponding to each optimization target are globally identified;

[0085] The optimization targets in this embodiment include, for example: hole diameter, countersink diameter, and countersink size discreteness; the process parameters include: clamping force, drilling speed, countersink speed, drilling feed speed, and countersink feed speed.

[0086] Step 2: Establish a single-factor experimental scheme for all process parameters identified in step 1, and determine the preliminary value boundaries of the five process parameters in the aircraft assembly hole making process and the influence trend of each process parameter on its corresponding optimization target through quantitative characterization of the experimental results of each single-factor experiment;

[0087] Step 3: Based on the results of the single-factor test, the measurement results of various optimization targets were subjected to regression analysis using SPSS software to obtain the influence weight of each process parameter and its interaction on each optimization target, and the key process variables were selected based on the weight ranking;

[0088] Step 4: Based on the measurement and analysis results of the single factor test, a second-order response surface model of process parameters is established to meet multiple optimization objectives of aircraft assembly holes, and multiple iterations are performed based on the hierarchical sequence method to achieve optimization of multiple optimization objectives;

[0089] Step 5: Determine the global optimal process parameter domain and verify its accuracy to obtain the optimal process parameter combination that simultaneously meets multiple manufacturing indicators.

[0090] The implementation methods of each step in this implementation case are analyzed in detail as follows.

[0091] In step 2 of this implementation case, the single factor experimental scheme is shown in Table 1. The values ​​of each process parameter are tested according to the maximum range that the current parameters can cover. Characterization tests are performed for each group of experiments, including hole diameter, countersink diameter, and countersink size dispersion.

[0092] Table 1 Process parameters of single factor test scheme

[0093]

[0094] In step 3 of this implementation case, SPSS software was used to perform regression analysis on the test results, and significance analysis was performed on each process parameter. Hypothesis testing was performed based on the regression model to obtain the influence weight of each response variable. Table 2 shows the regression analysis results of the countersink diameter test data, and Table 3 shows the regression analysis results of the countersink size dispersion.

[0095] Table 2 Linear regression analysis results of countersink diameter (n=76)

[0096]

[0097] Table 2 Linear regression analysis results of countersink diameter (n=76)

[0098]

[0099] Dependent variable: Y

[0100] *p<0.05 **p<0.01

[0101] Table 3 Linear regression analysis results of countersink size dispersion (n = 76)

[0102]

[0103] Dependent variable: Y

[0104] *p<0.05 **p<0.01

[0105] The main evaluation indicators in the table are F value and P value; F value is the ratio of the mean square sum to the error mean square sum, that is, the significant impact of the process parameters on the optimization target. The lower the value, the less significant the impact of the factor. The P value is the significance probability, which is used to measure the probability of whether to reject the null hypothesis. The lower the P value, the lower the probability, indicating that the evidence for rejecting the null hypothesis is more sufficient, indicating that the impact factor has a more significant impact on the response value. Based on statistical significance, P>0.05 indicates that the process parameter has no significant impact on the result, P<0.05 indicates that the parameter has a significant impact on the result, and P<0.01 indicates that the parameter has an extremely significant impact on the result. According to the above analysis results, the significance ranking of the five variables that affect the countersink diameter response target is: clamping force>countersink speed>drilling speed>drilling feed speed>countersink feed speed, among which the clamping force has the highest influence weight. Similarly, the significance ranking of the five variables affecting the discreteness of countersink size is: countersink speed > drilling speed > countersink feed speed > drilling feed speed > clamping force, among which the countersink speed has the highest influence weight.

[0106] In step 4 of this implementation case, the implementation method of establishing a multi-order response surface model of process parameters is described as follows:

[0107] In order to achieve multi-objective optimization of response indicators, a second-order response surface model of aircraft assembly hole shape accuracy characterization parameters with respect to process parameters is established. Based on the Box-Behnken experimental design method, response surface modeling is performed on five parameter variables to obtain a database for model fitting. The specific method is: three-level values ​​are taken for the response factors, and the three levels of the response factors are represented by "-1", "0" and "+1". The "0" level represents the middle value, the "-1" level represents the value that deviates negatively from the middle value, and the "+1" level represents the value that deviates positively from the middle value. A five-factor three-level experimental plan is established, as shown in the following table:

[0108] 5-factor 3-level response surface analysis experimental design table

[0109]

[0110] Experimental design and results of optimizing countersinking precision using response surface method

[0111]

[0112]

[0113] Combined with the above modeling sample data, the sub-regression equation based on Taylor expansion is selected for modeling:

[0114]

[0115] In the formula, a0 is a constant term; a i is the process parameter x i The coefficient of the first-order term; a ij is the process parameter x i and x j The interaction coefficient of ii is the process parameter x i The quadratic coefficient of x i is the i-th process parameter, x j is the jth process parameter; n is the number of process parameters; Y is the parameter characterizing the shape accuracy of the hole surface of the aircraft assembly. A response surface experiment database is formed through the response surface experiment, including: experimental data of each group of response surface experiments and corresponding Y data.

[0116] In the specific implementation, multiple iterative fittings were performed through the Design expert data analysis software to obtain the quadratic polynomial regression equation of the countersink size dispersion on the clamping force, drilling speed, countersink speed, drilling feed speed, and countersink feed speed:

[0117] Y=0.14+0.004937*A+0.005812*B+0.006187*C-0.00475*D-0.004188*E+0.012*AB-0.0085*AC+0 .016*AD-0.001*AE+0.00675*BC-0.00275*BD-0.005*BE+0.004*CD-0.02*CE-0.011*DE-0.041*A 2 -0.039*B 2 -0.055*C 2 -0.024*D 2 -0.016*E 2 .

[0118] Figure 2 The following table shows the response surface contour map of each parameter optimized based on the response surface in the embodiment of the present invention. The model significance and accuracy obtained by the response surface analysis are as follows:

[0119] Analysis results of empirical model and regression coefficient of countersink size dispersion

[0120]

[0121]

[0122] Note: P < 0.01 is extremely significant, indicated by **, P < 0.05 is significant, indicated by *, P > 0.05 is not significant, indicated by ns.

[0123] The regression model P < 0.0001 (extremely significant), and its lack of fit term P = 0.3291 > 0.05 (not significant), indicating that the model has a good degree of fit and can predict the corresponding regression value of the regression equation. 2 =0.9656, adjusted R 2=0.9380 (greater than 0.8000), indicating that 93.80% of the data can be explained by the model, indicating that the actual value of the model is not much different from the predicted value, indicating that the reliability of the equation is high. The size of the F value is an important indicator for evaluating the degree of influence of each variable on the response value. The larger the F value, the higher the contribution of the relevant model components to the response. When the significance test probability P<0.05, it reveals that the variable has a significant effect on the response value and has mathematical and statistical significance. By analyzing the relevant data, it can be seen that the first-order countersinking speed and drilling speed have extremely significant effects on the diameter (P<0.01), and the clamping force, countersinking feed speed, and drilling feed speed have significant effects on the diameter (P<0.05). The main effect relationship of each factor is analyzed as follows: C>B>A>D>E, that is, countersinking speed>drilling speed>clamping force>drilling feed speed>countersinking feed speed. The quadratic interactions AB, AD, and CE have extremely significant effects on the diameter (P < 0.01), and AC and DE have significant effects on the diameter (P < 0.05). Among the interactions between variables, the interaction between the clamping force and the countersink feed speed (AE) is not significant and can be excluded from the model fitting; the interaction between the drilling speed and the drilling feed speed (BD) is not significant and can be excluded from the model fitting; similarly, BE and CD are not significant and are not included in the model fitting.

[0124] The insignificant items in the above analysis results are simplified to obtain the response surface prediction model:

[0125] Y=0.14+0.004937×A+0.005812×B+0.006187×C-0.00475×D-0.004188×E+0.012×AB-0.0085×AC+0.016×AD-0.001×CE-0.011×DE-0.041×A 2 -0.039×B 2 -0.055×C 2 -0.024×D 2 -0.016×E 2

[0126] In step 4 of the implementation case, the method of performing multi-objective optimization on the established second-order response surface model by using the improved genetic algorithm is as follows:

[0127] The optimization of aircraft assembly hole making process parameters is a multi-objective optimization problem with multiple factors interacting. During the optimization process, the hole making quality evaluation indicators include hole making diameter, countersink diameter, countersink size discreteness, etc. It is difficult to achieve the optimality of the above indicators at the same time, and it is difficult to convert them into a single-objective optimization problem through weight coefficients. Therefore, the hole making quality characterization parameters are sorted by the hierarchical sequence method, and the improved genetic algorithm is used to optimize the objective function with large contribution first, and then the objective function with secondary contribution is optimized on its optimal solution set, and so on, until all objective functions are optimized, and finally the globally optimal hole making process parameter domain and the optimal hole making quality characterization parameter domain are determined.

[0128] Process parameter domain

[0129]

[0130] The specific expression of the multi-objective optimization model is:

[0131]

[0132] In this example, C1=7mm is used as the optimization target of the cavity diameter Q; C2=4mm is used as the optimization target of the aperture H; and the discreteness of the cavity diameter Q tends to 0 as the optimization target. Based on the constraints of the above decision variables, the three-objective optimization is performed, and the predicted optimal conditions are:

[0133]

[0134] According to the actual experimental conditions, the conditions were corrected to 301N, 8875rpm, 5980rpm, 188mm / min, and 138mm / min. Under this optimal condition, after three parallel tests, the actual countersink fluctuation range was 0.014±0.0015mm, which was within 5% of the predicted value of 0.010mm. This confirmed the good correlation between the predicted and experimental values, and indicated that the optimal machining process optimized by the response surface methodology was reasonable.

[0135] Table 2 shows the test results and prediction results of hole quality characterization performance.

[0136] The method of verifying the hole-making process parameter domain in step 5 of this implementation case is: the residual normal probability plot can be used to verify whether the residual is normally distributed. Figure 3 The residual normal probability diagram of the discrepancy of the countersink size after hole making in the embodiment of the present invention is shown in FIG. In the residual normal probability diagram, each data scatter point is distributed as a straight line, which shows that the model fitting error value is extremely small, indicating the effectiveness, accuracy and reliability of the linear regression model.

[0137] Although the embodiments disclosed in the present invention are as above, the contents are only embodiments adopted to facilitate understanding of the present invention and are not intended to limit the present invention. Any technician in the field to which the present invention belongs can make any modifications and changes in the form and details of implementation without departing from the spirit and scope disclosed in the present invention, but the patent protection scope of the present invention shall still be subject to the scope defined in the attached claims.

Claims

1. A multi-objective collaborative optimization method for aircraft assembly hole making process parameters, characterized in that: include: Step 1, according to the manufacturing index of the aircraft assembly hole making process, multiple optimization targets of the aircraft assembly hole making are determined, and the process parameters corresponding to each optimization target are globally identified; Step 2: Establish a single-factor experimental scheme for all process parameters identified in step 1, quantitatively characterize the experimental results through each single-factor experiment, determine the influence of each single process parameter on its corresponding optimization objectives, and calibrate the value boundary of each process parameter; Step 3: Based on statistical analysis, the influence weight of each process parameter and the interaction between process parameters on each optimization target is obtained, and key process variables are selected based on the weight ranking; Step 4, establishing a multi-order response surface model of process parameters that meets multiple optimization objectives of aircraft assembly hole making, and optimizing multiple optimization objectives by performing multiple iterations of the multi-order response surface model based on an improved genetic algorithm to determine the global optimal hole making process parameter domain; Step 5, verifying the accuracy of the global optimal hole-making process parameter domain determined in step 4, and obtaining the optimal process parameter combination that simultaneously meets multiple optimization objectives.

2. The multi-objective collaborative optimization method for aircraft assembly hole making process parameters according to claim 1 is characterized in that: In step 2, a single factor experiment scheme is implemented. The adjustment of process parameters in different single factor experiments requires that the aircraft assembly hole making system has an adjustable and controllable hardware module, including: The pressure foot module is used to adjust the pressure foot to fit the surface of the hole-making product according to the preset clamping force value, and the pressure value is 10-3000N; The spindle module is used to adjust the rotation speed of the hole-making spindle according to the preset value. The speed change delay is no more than 10ms, and the value range is 1000-30000rpm; The spindle module is also used to adjust the feed speed of the hole-making spindle in the axial direction according to a preset value, with a speed change delay of no more than 10ms and a value range of 10-300mm / min.

3. The multi-objective collaborative optimization method for aircraft assembly hole making process parameters according to claim 1 is characterized in that: The optimization target includes at least one of the following: morphology accuracy, hole position accuracy, processing efficiency, automatic hole making rate, and hole position layout; wherein the morphology accuracy includes at least one of hole making diameter, countersink diameter, and countersink size discreteness.

4. The multi-objective collaborative optimization method for aircraft assembly hole making process parameters according to claim 1 is characterized in that: The process parameters include at least one of the following: clamping force, spindle speed, spindle feed speed, product stack thickness, stack bonding surface gap distance, product material, product curvature, tool length, tool material, tool cutting angle, tool tip angle; the spindle speed, drilling speed and countersinking speed are divided according to time periods, and the time periods include drilling feed speed and countersinking feed speed.

5. The multi-objective collaborative optimization method for aircraft assembly hole making process parameters according to claim 1 is characterized in that: The step 2 comprises: Step 21, based on the control variable method, for each process parameter, only one of the process parameters is changed in each single factor experiment, and the other process parameters remain unchanged; Step 22, repeating step 21 for all process parameters in sequence until a single factor experiment for each process parameter is completed; Step 23, through the experimental results of step 21 and step 22, obtain the influence law of each single process parameter on each optimization target, and obtain the value boundary of each process parameter according to the requirements of each optimization target.

6. The multi-objective collaborative optimization method for aircraft assembly hole making process parameters according to claim 1 is characterized in that: The step 3 comprises: Step 31, performing a normality test on the experimental data obtained through the single factor experiment in step 2 to ensure that the experimental data of all single factor experiments meet the analysis conditions; Step 32, using the AVOVA method to calculate the F value of each process parameter and the interaction between the process parameters and the P value used to represent the probability of significance; wherein the F value is the ratio of the mean square sum to the mean square sum of errors; Step 33, determine the size of the P value and the standard threshold of the normality test. If P<threshold, it is considered that the process parameter has a significant impact on the optimization target, and the size of the F value corresponding to the P value represents the weight; Step 34, based on the P value of each process parameter and the interaction between the process parameters, determine whether each process parameter has a significant impact on each optimization target. If no significant impact is produced, the corresponding process parameter does not participate in the subsequent optimization steps; if a significant impact is produced, the influence weight is determined according to the F value corresponding to the P value, and the influence weight is sorted. If the influence weight accounts for less than 1%, it can be ignored, and the corresponding process parameter does not participate in the subsequent optimization steps.

7. The multi-objective collaborative optimization method for aircraft assembly hole making process parameters according to claim 1 is characterized in that: The method of establishing the multi-order response surface model of process parameters in step 4 includes: The Box-Behnken experimental design method was used to design the response surface experimental plan. The process parameters were set as response factors, and the optimization target was set as the response result. The multivariate multi-order response surface model established was: Among them, a0 is a constant term; a i is the process parameter x i The coefficient of the first-order term; a ij is the process parameter x i and x j The interaction coefficient of ii is the process parameter x i The quadratic coefficient of x i is the i-th process parameter, x j is the jth process parameter; n is the number of process parameters; Y is the parameter characterizing the shape accuracy of the hole profile of the aircraft assembly; A response surface experiment database is formed through the response surface experiment, including: experimental data of each group of response surface experiments and corresponding Y data.

8. The multi-objective collaborative optimization method for aircraft assembly hole making process parameters according to claim 7 is characterized in that: The method of performing multiple iterations based on the improved genetic algorithm in step 4 to achieve optimization of multiple optimization objectives includes: Step 41, analyzing the influence of each process parameter on each optimization target through regression analysis, and determining whether the influence of each process parameter on each optimization target is a positive effect or a negative effect; Step 42, testing the significance of different process parameters for each optimization target through variance analysis, quantifying the weight of the significance of the process parameters, and using this as a basis for simplifying the multivariate multi-order response surface model; Step 43, based on the response surface experiment database formed by the response surface experiment, the process parameters of the single optimization target with the largest contribution are first optimized by improving the genetic algorithm, and then the process parameters of the single optimization target with the secondary contribution are optimized on the optimal solution set, and so on, until all the process parameters corresponding to the single optimization targets obtain the corresponding optimization solution sets, thereby realizing the global optimal hole-making process parameter domain.

9. The multi-objective collaborative optimization method for aircraft assembly hole making process parameters according to claim 1, characterized in that: The method of verifying the hole-making process parameter domain in step 5 is to verify the accuracy of the response surface regression equation fitting by testing the residual normal plot of the standardized effect, including: Determine whether the residual distribution after fitting the multi-order response surface model of process parameters is uniform. If each sample data point satisfies the standard diagonal linear distribution and the distance deviation value is within 5%, it is judged to pass the residual normal plot test and the multi-order response surface model is valid.

Images