Vehicle endurance test specification combined development method based on mixed entropy multi-index decision

Abstract

A kind of vehicle endurance test specification combination development method based on mixed entropy multi-index decision-making, comprising: collecting whole vehicle and user damage data, and based on load spectrum damage equivalence principle, construct user-test field damage equivalence correlation model, based on damage equivalence correlation model, construct multi-objective optimization model;Using NSGA-II algorithm to optimize and solve multi-objective optimization model, obtain Pareto non-inferior solution set;Introduce mixed entropy weight and CRITIC two methods, construct mixed entropy multi-index decision-making algorithm, determine the comprehensive contribution of each solution in Pareto non-inferior solution set by mixed entropy multi-index decision-making algorithm;Based on the comprehensive contribution of each solution in Pareto non-inferior solution set, obtain the solution with the highest comprehensive contribution, and obtain the vehicle endurance test specification based on the solution with the highest comprehensive contribution.The present application can realize the effective correlation between test field and user road in the process of endurance test specification formulation, thereby reducing test mileage and test duration, and reducing test cost.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle durability testing and inspection technology, specifically to a method for developing a combination of vehicle durability test specifications based on multi-index decision-making in a mixed entropy manner. Background Technology

[0002] The implementation method of vehicle durability testing usually involves physical vehicles undergoing long-term, high-intensity driving on reinforced test roads to quickly simulate various working conditions and load conditions experienced by vehicles in actual use. The purpose is to quickly expose vehicle design and manufacturing defects during the short vehicle development and design phase, and effectively avoid quality risks when new vehicles are launched.

[0003] The key to efficiently achieving durability verification lies in developing a complete and efficient durability testing specification. The strength of this specification is based on the severity of the actual usage scenarios of the target users. The physical realization of this specification strength relies on a combination of various test-strengthened road surfaces. How to effectively link the test field and user roads during the development of durability testing specifications, thereby reducing test mileage and duration, and lowering test costs, is a technical problem that needs to be solved by those skilled in the art. Summary of the Invention

[0004] In view of this, the present invention provides a method for developing vehicle durability test specifications based on multi-index decision-making of entropy, so as to realize the effective correlation between test sites and user roads in the process of durability test specification formulation, thereby reducing test mileage and test time, and reducing test costs.

[0005] A method for developing vehicle durability test specifications based on multi-index decision-making with entropy mixing includes:

[0006] Step S1: Collect damage data of the whole vehicle and the user, and construct a user-test field damage equivalent correlation model based on the load spectrum damage equivalence principle. Then, based on the damage equivalent correlation model, use the damage difference between the user and the test field as the objective function and the frequency cyclic matrix as the independent variable to construct a multi-objective optimization model.

[0007] Step S2: The NSGA-II algorithm is used to optimize and solve the multi-objective optimization model to obtain the Pareto non-dominated solution set;

[0008] Step S3: Based on the multi-decision optimization algorithm, two methods, hybrid entropy weight and CRITIC, are introduced to construct a hybrid entropy multi-index decision algorithm. The hybrid entropy multi-index decision algorithm is used to determine the comprehensive contribution of each solution in the Pareto non-dominated solution set.

[0009] Step S4: Based on the comprehensive contribution of each solution in the Pareto non-dominated solution set, obtain the solution with the highest comprehensive contribution, and obtain the vehicle durability test specification based on the solution with the highest comprehensive contribution.

[0010] The vehicle durability test specification combination development method based on multi-index decision-making based on entropy mixing provided by the present invention has the following beneficial effects:

[0011] (1) Based on the load spectrum damage equivalence principle and user association theory, this invention establishes a multi-objective optimization model with the difference between user and test field damage as the objective function and the frequency cyclic matrix as the independent variable, realizing the effective association between the test field and user roads. First, the NSGA-II algorithm is used to solve the Pareto non-dominated solution set of the multi-objective optimization model, and the mixed entropy multi-index decision algorithm is constructed through the two methods of mixed entropy weight and CRITIC, which can realize the optimal solution in the Pareto non-dominated solution set.

[0012] (2) This invention constructs a mixed entropy multi-index decision algorithm by using both the mixed entropy weight and CRITIC methods. This not only ensures the consistency of damage in the calculation channel, but also effectively takes into account factors such as damage distribution in the verification channel, test cycle, and mileage. Compared with the traditional CRITIC-TOPSIS method, this mixed entropy multi-index decision algorithm is beneficial for the fit of the endurance conditions of the test users and the balanced utilization of the enhanced road resources in the test field. At the same time, it can achieve the compression of test mileage, the reduction of connecting road mileage, and the shortening of the single test cycle.

[0013] (3) Through testing and verification of actual working conditions, compared with the vehicle durability test specifications obtained by traditional models, the test working mileage of the present invention is reduced by 12.12%, the connecting road mileage is reduced by 10.5%, and the test time is reduced by 10.23%, which effectively reduces the test cost. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the development method for combining vehicle durability test specifications based on multi-index decision-making based on entropy mixing, as provided in an embodiment of the present invention.

[0015] Figure 2 The image shows the statistical distribution of rain flow in the x-direction of the left front wheel compared to the method proposed in this invention and the traditional multi-objective decision-making algorithm.

[0016] Figure 3 This is a statistical diagram of the rain flow distribution in the y-direction of the left front wheel, comparing the method proposed in this invention with that of a traditional multi-objective decision-making algorithm.

[0017] Figure 4 The image shows the statistical distribution of rain flow in the z-direction of the left front wheel compared to the method proposed in this invention and the traditional multi-objective decision-making algorithm.

[0018] Figure 5 This is a comparison chart of the test metrics of the method proposed in this invention and traditional multi-objective decision-making algorithms. Detailed Implementation

[0019] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain embodiments of the present invention, and should not be construed as limiting the present invention.

[0020] Please see Figure 1 The embodiments of the present invention provide a method for developing a combination of vehicle durability test specifications based on multi-index decision-making with entropy mixing, including steps S1-S4:

[0021] Step S1: Collect damage data of the whole vehicle and the user, and construct a user-test field damage equivalent correlation model based on the load spectrum damage equivalence principle. Then, based on the damage equivalent correlation model, use the damage difference between the user and the test field as the objective function and the frequency cyclic matrix as the independent variable to construct a multi-objective optimization model.

[0022] The core of the equivalent correlation model is to construct a representation matrix for the number of cycles of various reinforced road surfaces tested, and to ensure that the load history statistical characteristics of the reinforced road surface spectrum in the test field match the load spectrum of user service. Among these, the vehicle damage matrix... for:

[0023]

[0024] in, , , , , , , , , Vehicle damage matrix The elements in, with For example, Indicates the first The first road to strengthening Damage to critical locations on the vehicle. To enhance the total number of roads, This represents the total number of critical locations on the vehicle.

[0025] User damage matrix for:

[0026]

[0027] in, , , User damage matrix The elements in, with For example, Indicates the first mileage corresponding to the target mileage under user conditions. Damage under various working conditions;

[0028] Frequency cyclic matrix in test specifications for:

[0029]

[0030] in, , , Frequency cyclic matrix The elements in, with For example, Indicates the first The number of cycles for each reinforced road;

[0031] The traditional user association equivalent model is shown in the following equation:

[0032]

[0033] After measuring the load conditions of typical user operating conditions and the reinforced road in the test field, the corresponding results were obtained using methods such as the fatigue damage equivalence principle, load spectrum extrapolation theory, and accelerated testing techniques. and Thus, the frequency cyclic matrix The solution is then performed. Finally, the frequency cyclic matrix can be verified using dimensions such as load and damage. The validity of the constructed results is used to complete the construction of the test field enhanced durability specifications.

[0034] In practical engineering applications, enhanced durability tests at test sites cannot achieve 100% equivalence to damage under typical user operating conditions. Therefore, this application uses the difference between user and test site damage as the objective function, employing a frequency cyclic matrix... Construct a multi-objective optimization model for the independent variables.

[0035] The objective function is expressed as follows:

[0036]

[0037] in, Indicates the first test site specification The damage difference between key locations of the vehicle and the user's working conditions. For the first reinforced road Damage to critical locations on the vehicle. For the second reinforced road Damage to critical locations on the vehicle. For the first The first road to strengthening Damage to critical locations on the vehicle. This indicates the number of cycles for the first reinforced road. This indicates the number of cycles for the second reinforced road. Indicates the first The number of cycles for each reinforced road. Indicates transpose. For the target mileage corresponding to the user conditions Damage under various working conditions.

[0038] The expression for the constructed multi-objective optimization model is: ;

[0039] This indicates taking the minimum value. This represents the damage difference between the first critical location of the vehicle under test track specifications and the user's operating conditions. This represents the damage difference between the second critical location of the vehicle under test track specifications and the user's operating conditions. Indicates the first test site specification The damage difference between key locations of the vehicle and the user's working conditions. This represents the total number of critical locations on the vehicle.

[0040] In this embodiment, to avoid excessively long testing times and high experimental costs, the multi-objective optimization model is subject to the following constraints:

[0041]

[0042]

[0043]

[0044]

[0045]

[0046] in, For the first The length of the reinforced road, For the first The length of the connecting pavement of the reinforced road To increase the total number of test cycles for the road; To limit the durability test mileage, in this embodiment, 47,000 km is used; In the first The time required to travel on the enhanced road. In the first The time required to travel on the connecting surfaces of the reinforced road. To limit the test time, in this embodiment, 69,000 min is used; For the first The rainflow count corresponds to the number of cycles under the load. For the SN curve and The corresponding number of failure cycles can be calculated based on Miner's linear damage accumulation criterion. .

[0047] Since the load on a vehicle during operation mainly comes from the vehicle's own weight and road excitation, and is transmitted to various parts of the vehicle through the wheels and suspension system, the key monitoring areas, to accurately characterize damage during vehicle operation, cover critical mechanical nodes such as the axle assembly, suspension system, steering tie rod assembly, and leaf spring connection points. Therefore, in this embodiment, the data acquisition matrix is ​​shown in Table 1, with four types of acquisition channels: force, displacement, acceleration, and stress. Six-component force sensors, displacement sensors, accelerometers, and strain sensors are deployed at the four wheel hubs, front and rear shock absorbers, front and rear axles, and leaf springs for data acquisition. Ultimately, a total of 24 force signal channels, 4 displacement and acceleration signal channels, and 5 stress channels are acquired.

[0048] Table 1

[0049]

[0050] in, F x _F r , F x _F l , F x _R r , F x _R l The x-forces are for the right front wheel, left front wheel, right rear wheel, and left rear wheel, respectively.

[0051] F y _F r 、F y _F l 、F y _R r 、F y _R l These are the y-forces of the right front wheel, left front wheel, right rear wheel, and left rear wheel, respectively.

[0052] F z _F r 、F z _F l 、F z _R r 、F z _R l These are the z-forces of the right front wheel, left front wheel, right rear wheel, and left rear wheel, respectively.

[0053] T x _F r 、T x _F l 、T x _R r 、T x _R l These are the x-direction torques of the right front wheel, left front wheel, right rear wheel, and left rear wheel, respectively.

[0054] T y _F r 、T y _F l 、T y _R r 、T y _R l These are the y-direction torques of the right front wheel, left front wheel, right rear wheel, and left rear wheel, respectively.

[0055] T z _F r 、T z _F l 、T z _R r 、T z _R lThese are the z-axis torques of the right front wheel, left front wheel, right rear wheel, and left rear wheel, respectively.

[0056] F r_dis 、F l_dis These are the displacements of the right front shock absorber and the left front shock absorber, respectively.

[0057] R r_dis 、R l_dis These are the displacements of the right rear shock absorber and the left rear shock absorber, respectively.

[0058] F r_acc 、F l_acc These are the deceleration on the right side of the front axle and the deceleration on the left side of the front axle, respectively.

[0059] R r_acc 、R l_acc These are the decelerations on the right and left sides of the rear axle, respectively.

[0060] Stab.Bar_strain For lateral stabilizer bar stress;

[0061] Leaf_Fr_strain, Leaf_Fl_strain, Leaf_Rr_strain, Leaf_Rl_strain These are the stresses at the front end of the right leaf spring, the front end of the left leaf spring, the rear end of the right leaf spring, and the rear end of the left leaf spring, respectively.

[0062] Regarding the damage data collection method at the vehicle test track: Based on the structural characteristics of light commercial vehicles and user application scenarios, 10 types of reinforced roads, including Belgian roads, washboard roads, pothole roads, and off-road roads, were selected for road spectrum collection. Furthermore, all driving directions at the test track were unidirectional, and the vehicle speed was the recommended value.

[0063] Step 3 concerns the method for collecting vehicle user damage data. The design target for the product's total lifespan durability mileage is defined as 240,000 kilometers over 10 years, equivalent to 95% user utilization. Based on the user durability design target, the calculated results of user durability data are extrapolated to obtain the total damage over the user's durability lifespan. Before collecting typical user operating condition road spectrum data, it is necessary to analyze and define the data collection scenarios for typical user operating conditions. The collection of typical user operating condition road spectrum data includes eight aspects: 1) Environmental condition definition: clearly defining four types of extreme scenarios: high temperature (≥35℃), extreme cold (≤-20℃), high humidity (relative humidity ≥70%), and high altitude (altitude ≥2500 meters); 2) Sample region selection: based on historical temperature and altitude data, defining the user sample region range that meets the above conditions; 3) Vehicle permanent residence determination: extracting GPS data from the last leg of the user's daily journey, analyzing the province, and statistically analyzing the province with the highest GPS percentage (>90%) as the permanent residence, and calculating the number of vehicles permanently stationed in each province; 4) Sample database construction: Based on the distribution of permanent residences, establish four types of user sample databases: high temperature, high cold, high humidity, and high altitude; 5) Sample quantity allocation: Based on the proportion of permanent vehicles in each province, randomly allocate the required sample quantity for each scenario in a stratified manner; 6) Data export and preprocessing: Export the raw data of the selected samples and classify the road types based on OpenStreetMap; 7) Operating condition characteristic analysis: Statistically analyze the proportion of road surface types, speed distribution, and driving mileage to quantify the actual user usage scenarios; 8) Standardized scenario determination: Based on the comprehensive environmental conditions and operating condition analysis results, define standardized scenarios for user durability testing.

[0064] Typical end-user road conditions include: highways, urban roads, national / provincial roads, and rough roads, with set proportions of 40%, 30%, 20%, and 10% respectively. The total mileage for a single user test cycle is 60,000 kilometers, covering cold, high-temperature, high-humidity, and high-altitude regions of China. Regarding vehicle test loads, all four road types in major cities are tested under full-load ballast conditions, consistent with the enhanced road conditions at the test track.

[0065] Step S2: The NSGA-II algorithm is used to optimize the multi-objective optimization model and obtain the Pareto non-dominated solution set.

[0066] The parameters for the NSGA-II algorithm are set as follows:

[0067] The Pareto ratio is set to 0.3, the population size to 100, the maximum number of generations to 200, the stopping generation to 100, and the function tolerance to 10. 210 In this embodiment, by substituting relevant test data, a solution set consisting of 30 Pareto non-dominated solutions is obtained through convergence.

[0068] Step S3: Based on the multi-decision optimization algorithm, two methods, mixed entropy weight and CRITIC, are introduced to construct a mixed entropy multi-index decision algorithm. The comprehensive contribution of each solution in the Pareto non-dominated solution set is determined by the mixed entropy multi-index decision algorithm.

[0069] In this embodiment, two different methods, hybrid entropy weighting and CRITIC, are used to calculate the weights of each indicator. This approach can take into account both the information entropy characteristics within the indicators and the differences and conflicts between indicators, thereby more comprehensively reflecting the objective situation of the data.

[0070] Step S3 specifically includes:

[0071] The first in the Pareto non-dominated solution set Each solution is labeled as a variable. Then, by combining decision indicators, a comprehensive evaluation decision matrix is ​​formed. The expression is:

[0072]

[0073] in, This is the first solution under the first evaluation metric. This is the first solution under the second evaluation metric. For the first The first solution under each evaluation metric. This is the second solution under the first evaluation metric. This is the second solution under the second evaluation metric. For the first The second solution under each evaluation metric. For the first evaluation indicator, the first One solution. For the second evaluation indicator One solution. For the first The first evaluation indicator One solution. The total number of solutions in the Pareto non-dominated solution set;

[0074] Comprehensive evaluation decision matrix Normalization is performed to obtain the normalized matrix. , The first in line, number Column elements The expression is:

[0075]

[0076] in, For the first The first evaluation indicator One solution;

[0077] No. The first evaluation indicator The proportion of each solution to the corresponding indicator for:

[0078]

[0079] No. The entropy value corresponding to each evaluation indicator and the coefficient of difference The expression is:

[0080]

[0081]

[0082] in, As an intermediate quantity, ;

[0083] Thus, the first Entropy weight of each evaluation indicator The expression is:

[0084]

[0085] Furthermore, calculate the first... Standard deviation of each evaluation indicator The expression is:

[0086]

[0087] in, For the first The average value of the solutions corresponding to each evaluation index;

[0088] No. Conflict measurement of evaluation indicators The expression is:

[0089]

[0090]

[0091] in, For the first The first evaluation indicator and the first The correlation coefficient of each evaluation indicator For the first The first evaluation indicator One solution. No. The average value of the solution corresponding to each evaluation index. For the first The first evaluation indicator One solution;

[0092] Then define the combined weights. for:

[0093]

[0094] Recalculate The corresponding weighted normalization matrix Optimal solution matrix And worst solution matrix The expression is:

[0095]

[0096]

[0097]

[0098] in, For the first evaluation indicator, the first The weighted normalization matrix corresponding to each solution For the second evaluation indicator The weighted normalization matrix corresponding to each solution This indicates taking the maximum value;

[0099] Then calculate The Middle Euclidean distance from the sequence corresponding to each quantity to the positive ideal solution set ,as well as The Middle Euclidean distance from the sequence corresponding to each quantity to the negative ideal solution set The expression is:

[0100]

[0101]

[0102] The smaller the value, the more related it is to the first... The closer the sequence corresponding to each quantity is to the positive ideal solution set, the further away it is from the positive ideal solution set; The smaller the value, the more related it is to the first... The closer the sequence corresponding to a quantity is to the negative ideal solution set, the further away it is from the negative ideal solution set.

[0103] Finally, the first The combined contribution of each solution The expression is:

[0104] .

[0105] Overall contribution The larger the value, the greater the overall contribution of the evaluated object to each evaluation indicator.

[0106] Step S4: Based on the comprehensive contribution of each solution in the Pareto non-dominated solution set, obtain the solution with the highest comprehensive contribution, and obtain the vehicle durability test specification based on the solution with the highest comprehensive contribution.

[0107] In this embodiment, a 30×6 decision matrix was constructed. Based on the calculated weight coefficients of the 12 indicators, a comprehensive contribution analysis was performed using a multi-indicator decision-making algorithm based on entropy mixing. The comprehensive contribution of each Pareto solution relative to the 12 indicators was calculated. ( q =1,2,3…,30), and the vehicle durability test specifications are obtained based on the solution with the highest contribution rate.

[0108] In this embodiment, the rainflow distribution in the x, y, and z directions of the left front wheel is statistically analyzed and compared with that of the traditional multi-objective decision-making algorithm (CRITIC-TOPSIS model). The results are as follows: Figures 2 to 4 As shown, by Figures 2 to 4 It can be seen that the rainflow count results generated by this invention are highly consistent with the actual values, indicating that the vehicle damage performance under the durability specification of this invention can meet the requirements. Furthermore, compared with the specification constructed by the traditional CRITIC-TOPSIS model, the specification established by this invention has a higher frequency of large-amplitude loads and a lower frequency of small-amplitude loads, which better matches the ideal load distribution in durability testing. In addition, since small-amplitude loads generally contribute less to fatigue damage, reducing the frequency of small-amplitude loads helps reduce invalid test cycles and thus reduce test costs. Secondly, since large-amplitude loads contribute more significantly to fatigue damage, appropriately increasing the frequency of large-amplitude loads can make the durability test closer to real-world usage scenarios, while also accelerating the test and improving test efficiency.

[0109] The method proposed in this invention is compared with the traditional multi-objective decision-making algorithm (CRITIC-TOPSIS model) in various indicators. Testing and verification are conducted through real-world operating scenarios, with the vehicle weight set to full load. Figure 5 It can be seen that the method proposed in this invention significantly improves the efficiency and test cost of durability test specifications. Specifically, the test mileage for the standard operating conditions established in this invention is 33,055 km, the connecting road mileage is 3,810 km, and the test duration is 51,600 min. Compared with the traditional CRITIC-TOPSIS model, the test operating mileage is reduced by 12.12%, the connecting road mileage is reduced by 10.5%, and the test duration is reduced by 10.23%.

[0110] In summary, the vehicle durability test specification combination development method based on multi-index decision-making based on entropy mixing, as described in the above embodiments, has the following beneficial effects:

[0111] (1) Based on the load spectrum damage equivalence principle and user association theory, this invention establishes a multi-objective optimization model with the difference between user and test field damage as the objective function and the frequency cyclic matrix as the independent variable, realizing the effective association between the test field and user roads. First, the NSGA-II algorithm is used to solve the Pareto non-dominated solution set of the multi-objective optimization model, and the mixed entropy multi-index decision algorithm is constructed through the two methods of mixed entropy weight and CRITIC, which can realize the optimal solution in the Pareto non-dominated solution set.

[0112] (2) This invention constructs a mixed entropy multi-index decision algorithm by using both the mixed entropy weight and CRITIC methods. This not only ensures the consistency of damage in the calculation channel, but also effectively takes into account factors such as damage distribution in the verification channel, test cycle, and mileage. Compared with the traditional CRITIC-TOPSIS method, this mixed entropy multi-index decision algorithm is beneficial for the fit of the endurance conditions of the test users and the balanced utilization of the enhanced road resources in the test field. At the same time, it can achieve the compression of test mileage, the reduction of connecting road mileage, and the shortening of the single test cycle.

[0113] (3) Through testing and verification of actual working conditions, compared with the vehicle durability test specifications obtained by traditional models, the test working mileage of the present invention is reduced by 12.12%, the connecting road mileage is reduced by 10.5%, and the test time is reduced by 10.23%, which effectively reduces the test cost.

[0114] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A vehicle endurance test specification combination development method based on entropy mixing multi-index decision, characterized in that, The method comprises the following steps: Step S1, collecting the whole vehicle and user damage data, and constructing a user-test field damage equivalent correlation model based on the load spectrum damage equivalent principle; and then, based on the damage equivalent correlation model, taking the user and test field damage difference as a target function and a frequency cycle matrix as an independent variable, a multi-objective optimization model is constructed; Step S2, the NSGA-II algorithm is used to optimize and solve the multi-objective optimization model to obtain a Pareto non-inferior solution set; Step S3, based on the multi-decision optimization algorithm, two methods of mixed entropy weight and CRITIC are introduced to construct a mixed entropy multi-index decision algorithm, and the comprehensive contribution of each solution in the Pareto non-inferior solution set is determined by the mixed entropy multi-index decision algorithm; Step S4, based on the comprehensive contribution of each solution in the Pareto non-inferior solution set, the solution with the highest comprehensive contribution is obtained, and the vehicle durability test specification is obtained based on the solution with the highest comprehensive contribution; In step S3, the expression of the target function in step S1 is: The first solution in the Pareto non-inferior solution set is marked as a variable , and a decision index is combined to form a comprehensive evaluation decision matrix , and the expression is: ​ wherein, is the 1st solution under the 1st evaluation index, is the 1st solution under the 2nd evaluation index, is the 1st solution under the 3rd evaluation index, is the 1st solution under the 4th evaluation index, is the 2nd solution under the 1st evaluation index, is the 2nd solution under the 2nd evaluation index, is the 2nd solution under the 3rd evaluation index, is the 2nd solution under the 4th evaluation index, is the 1st solution under the 1st evaluation index, is the 1st solution under the 2nd evaluation index, is the 1st solution under the 3rd evaluation index, is the 1st solution under the 4th evaluation index, is the 2nd solution under the 1st evaluation index, is the 2nd solution under the 2nd evaluation index, is the 2nd solution under the 3rd evaluation index, is the total number of solutions in the Pareto non-inferior solution set; The decision matrix is evaluated comprehensively The normalized matrix is obtained by normalizing , The expression of the element in the first row, the first column of the normalized matrix is ​ wherein, is the first evaluation index under the first solution; No. The first evaluation indicator The proportion of each solution to the corresponding indicator for: No. The entropy value corresponding to each evaluation indicator and the coefficient of difference The expression is: wherein is an intermediate quantity, ; Further, the entropy weight of the first evaluation index is obtained , and the expression is as follows . The standard deviation of the evaluation index is calculated The expression is:​ wherein, is the average value of the solution for the evaluation index. The conflict measure of the first evaluation index is expressed as: in, For the first The first evaluation indicator and the first The correlation coefficient of each evaluation indicator For the first The first evaluation indicator One solution. No. The average value of the solution corresponding to each evaluation index. For the first The first evaluation indicator One solution; The combined weight is then defined as is: recalculation corresponding weighted norm matrix optimal solution matrix and worst solution matrix , expressed as: wherein, is the weighted normalized matrix corresponding to the jth solution under the first evaluation index, is the weighted normalized matrix corresponding to the jth solution under the first evaluation index, is the weighted normalized matrix corresponding to the jth solution under the second evaluation index, is the weighted normalized matrix corresponding to the jth solution under the second evaluation index, denotes taking the maximum value; Then the Euclidean distance of the sequence corresponding to the first quantity in the middle to the positive ideal solution set is calculated The expression is:​​​​​ Finally, the overall contribution of the nth solution is obtained expressed by the formula​ 。 2. The vehicle endurance test specification combined development method based on entropy mixing multi-index decision according to claim 1, characterized by, In step S1, the limited condition of the multi-objective optimization model is: wherein, represents the damage difference of the th vehicle key position and the user condition under the test field specification, is the damage of the th vehicle key position on the first reinforced road, is the damage of the th vehicle key position on the second reinforced road, is the damage of the th vehicle key position on the th reinforced road, represents the number of cycles of the first reinforced road, represents the number of cycles of the second reinforced road, represents the number of cycles of the th reinforced road, represents the transpose, is the damage of the th condition corresponding to the target mileage under the user condition.

3. The vehicle endurance test specification combined development method based on entropy mixing multi-index decision according to claim 2, characterized in that, In step S1, the expression of the multi-objective optimization model constructed is: ; represents the minimum value, represents the damage difference between the first vehicle key position and the user working condition under the test field specification, represents the damage difference between the second vehicle key position and the user working condition under the test field specification, represents the damage difference between the first vehicle key position and the user working condition under the test field specification, is the total number of vehicle key positions.

4. The vehicle endurance test specification combined development method based on entropy mixing multi-index decision according to claim 3, characterized by, In step S1, the collection of the whole vehicle and user damage data specifically comprises: wherein, is the number of times of traveling on the strengthened road, is the length of the road surface connected to the strengthened road, is the total number of cycles of the test on the strengthened road, is the limited durability test distance, is the time required for traveling on the strengthened road, is the time required for traveling on the road surface connected to the strengthened road, is the limited test time, is the number of cycles under the load corresponding to the is the number of cycles of failure on the S-N curve corresponding to .

5. The vehicle endurance test specification combined development method based on entropy mixing multi-index decision according to claim 1, characterized by, Four types of collection channels of force, displacement, acceleration and stress are set, and six-component force sensor, displacement sensor, accelerometer and strain sensor are arranged at the positions of four-wheel cores, front and rear shock absorbers, front and rear axles and leaf springs for data collection. In step S2, the parameter settings of the NSGA-II algorithm are as follows:

6. The vehicle endurance test specification combined development method based on entropy mixing multi-index decision according to claim 1, characterized by, ​ The Pareto ratio is set to 0.3, the population size is set to 100, the maximum genetic generation is set to 200, the stop generation is set to 100, and the function tolerance is set to 10 210 .

Images