Durable rack design method and system

Abstract

The invention provides a durable rack design method and system, and relates to the field of durable rack design, and the method comprises the steps: obtaining a plurality of samples of a durable rack to construct an original sample set, calculating the evaluation score of each sample, sequentially selecting the samples according to the evaluation scores, constructing a first sample set, and obtaining the remaining samples; screening residual samples according to the evaluation function and the selected probability to construct a second sample set, constructing a secondary sample set according to the first sample set and a preset sample set, calculating an evaluation score of each sample in the secondary sample set, and judging whether all evaluation scores converge or not; and if yes, screening the secondary sample set according to a preset condition to obtain a target sample, and designing the durable rack according to the sample data of the target sample. The technical problem that a durable rack in the prior art is not high in overall integration level and narrow in test coverage range and cannot adapt to multiple test items is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of durable bench design, and particularly relates to a durable bench design method and system. BACKGROUND

[0002] With the acceleration of new energy vehicle product technology iteration, the test parameters such as prototype size, ambient temperature test requirements, water cooling test requirements and test duration continue to change, which leads to the following problems. On the one hand, the environmental simulation capability of the existing test equipment is insufficient: it cannot meet the higher requirements of the product iteration on the key indicators such as temperature rise rate, water cooling rate and transmission efficiency, it is difficult to adapt to the dynamically changing test scene, and there is a technical gap between the environmental simulation parameters and the actual test requirements. On the other hand, the test equipment of the current new energy vehicle test room has the defects of single function and insufficient coordination: different test projects (such as temperature and humidity cycle, vibration durability, electrical performance test, etc.) need to rely on independent special benches, which leads to the fragmentation of the test process. The specific performance is that the whole cycle test of a single component (such as a motor controller) needs to be transferred between multiple benches (12-15 times), which not only prolongs the overall test time, but also causes test backlog at key nodes due to the mismatch of test time and unbalanced resource scheduling of each bench, and cannot realize efficient continuous testing.

[0003] In the prior art, on the one hand, although the integration of part of the test is realized (only the brake and drive test is integrated), there are problems of low overall integration and narrow test coverage: only the integration of two types of tests for specific components is carried out, the integration demand of multiple types of tests (temperature and humidity, vibration, electrical performance, etc.) of other components (such as motor controllers, electrical systems, etc.) in the new energy vehicle test process is not considered comprehensively, the efficiency optimization effect of the whole test process is limited, and the integrated test pain point of multiple projects and multiple components cannot be solved. On the other hand, the existing equipment does not form a hardware interface and control system suitable for multiple test projects: because the special test room equipment is not added, the bench compatible interface and collaborative control system are not expanded, it is impossible to complete multiple test projects without disassembling and transferring the tested components, and the test connection needs to be realized by transferring between benches, which further aggravates the process fragmentation and efficiency loss, and lacks the hardware and control basis to support "one-stop" continuous testing. SUMMARY

[0004] Based on this, the purpose of the present application is to provide a durable bench design method and system to solve the technical problem that the existing durable bench has low overall integration and narrow test coverage, which cannot adapt to multiple test projects.

[0005] In one aspect, the present application provides a durable bench design method, comprising: obtaining a plurality of samples of the durability bench to construct an original sample set according to sample data, substituting the sample data into an evaluation function to calculate an evaluation score of each sample, arranging the calculated evaluation scores in descending order to sequentially select a first preset number of samples according to the evaluation scores to construct a first sample set, and obtaining remaining samples in combination with a total number of samples of the original sample set; screening the remaining samples according to the evaluation function and the selected probability to obtain a second preset number of samples to construct a second sample set, constructing a secondary sample set of a target number according to the first sample set, the second sample set, and a preset sample set, and calculating an evaluation score of each sample in the secondary sample set according to the evaluation function; judging whether the evaluation scores of all samples in the secondary sample set converge; If not, removing samples with non-converging evaluation scores and obtaining a removal number, and randomly selecting a number of samples corresponding to the removal number from the original sample set to obtain a make-up sample, updating the secondary sample set according to the make-up sample to maintain the target number of the secondary sample set, and returning to the step of calculating the evaluation score of each sample in the secondary sample set according to the evaluation function until the evaluation scores of all samples in the secondary sample set converge; If yes, screening the secondary sample set according to a preset condition to obtain a target sample, and designing the durability bench according to sample data of the target sample.

[0006] In addition, the durability bench design method according to the above-mentioned application can have the following additional technical features: Further, the original sample set includes a plurality of samples, each sample includes a plurality of sample data, the sample data includes a plurality of sample data, and the plurality of sample data includes type A sample data, type B sample data, and type C sample data; The step of substituting the sample data into the evaluation function to calculate the evaluation score of each sample further includes: randomly sorting the samples in the original sample set, each sample including a plurality of sample data; dividing the sorted samples into a plurality of groups with each group including two adjacent samples; crossing and exchanging one type of sample data in each group of samples to obtain exchanged samples; substituting the sample data of the exchanged samples into the evaluation function to calculate the evaluation score of each sample.

[0007] Further, the step of crossing and exchanging one type of sample data in each group of samples to obtain exchanged samples includes: selecting one type of sample data to be crossed and exchanged in a group of samples, and performing data mutation on the type of sample data to obtain mutated sample data; crossing and exchanging the mutated sample data to obtain exchanged samples; The methods for data mutation include: Compared with the original sample data, the rate of change of a single sample data after mutation is no more than 10%, and it gradually decreases with the number of iterations until it reaches 0%.

[0008] Further, the step of filtering the remaining samples according to the evaluation function and the selection probability to obtain a second preset number of samples to construct a second sample set includes: Roulette wheel selection is used as the selection method to obtain a second preset number of samples. The selection method includes: The fitness of all individuals is obtained by summing the evaluation scores of all individuals in the remaining samples according to the evaluation function, and the probability of each individual being selected is calculated based on the fitness. Generate and obtain random numbers, calculate the cumulative probability starting from the low-scoring samples, and iterate through the cumulative probabilities based on the random numbers; The cumulative probability of the first random number not less than the current random number is obtained based on the traversal results and used as the target probability. The target sample is then obtained based on the target probability. Determine whether the number of target samples meets the second preset number; If not, return to the step of summing the evaluation scores of all individuals in the remaining samples according to the evaluation function to obtain the fitness of all individuals, until the target sample size meets the second preset number.

[0009] Furthermore, the expression for the evaluation function is: Y=ω1· R heating +ω2· η gearbox +ω3· η cooling ; In the formula: Y represents the evaluation function; R heating For the temperature rise efficiency of the incubator; η gearbox For gearbox transmission efficiency; η cooling ω1, ω2, and ω3 are the cooling efficiency of the water chiller; where: R heating The calculation formula is: ; In the formula: P net This refers to the total power of the incubator; ρ This refers to the air density inside the incubator. C p The specific heat capacity of the air inside the incubator; V K1 is the volume of the temperature chamber; K1 is the insulation coefficient; ΔT represents the temperature difference between the medium temperature and room temperature. η gearbox The calculation formula is: ; In the formula: d is the gear diameter; N is the number of gears; μ The coefficient of friction; η cooling The calculation formula is: ; In the formula: m flow Water flow rate; C p,water ΔT represents the specific heat capacity of water; ∆T represents the temperature difference between the medium temperature and room temperature. P pump This refers to the power of the water chiller.

[0010] Another aspect of the present invention provides a durability test bench design system for implementing the above-mentioned durability test bench design method, wherein the method is specifically applied to a control module; the system includes an environmental chamber, two gearboxes located diagonally opposite to the environmental chamber, and two dynamometers located at the power output end of the gearboxes, the two dynamometers being diagonally arranged; two test pieces are provided in the environmental chamber, and the gearboxes are connected to the test piece differentials via drive shafts; The system also includes a load cabinet for simulating inverter load, a water chiller, a power analyzer, and a high-voltage power supply. The load cabinet is connected to a gearbox to serve as the load of the load cabinet. The water chiller is connected to two test pieces to provide cooling for the operation of the test pieces. The power analyzer is communicatively connected to the control module and the two test pieces. The high-voltage power supply is connected to the two test pieces to supply power to the two test pieces. The system also includes a control module, which includes a main control cabinet and a low-voltage power supply. The low-voltage power supply powers the main control cabinet. The control module is communicatively connected to other devices, including an environmental chamber, test pieces, a load cabinet, a water chiller, a high-voltage power supply, and a power analyzer.

[0011] The aforementioned durability test bench design method and system, through hierarchical screening and convergence verification, can accurately identify core samples that "can support multi-experiment integration and adapt to iterative needs," avoiding test bench design based on invalid or low-value samples. This eliminates the need for redundant development of multiple dedicated test benches; a single highly integrated test bench designed based on the target samples can cover multiple testing scenarios, significantly reducing test bench development costs and resource waste. Secondly, through convergence judgment and iterative mechanisms for supplementary updates, it ensures that the final target samples all meet the condition of "evaluation score convergence," meaning that the test bench parameters corresponding to the sample data have consistency and stability. This allows the durability test bench designed based on these samples to avoid resource mismatch problems caused by fluctuations in test bench parameters. Furthermore, this application uses dynamic sample optimization, where supplementary samples are selected from the original samples while maintaining the target number of secondary sample sets. Different sample data from the original sample set can be flexibly incorporated, enabling the resulting test bench to adapt to multi-dimensional parameter changes after product iteration. This allows it to meet higher environmental simulation requirements without frequent test bench modifications, solving the pain point of mismatch between environmental simulation capabilities and iterative needs. Attached Figure Description

[0012] Figure 1 This is a flowchart illustrating the steps of the durability test bench design method in an embodiment of the present invention. Figure 2 This is a flowchart of steps S1011-S1014 in an embodiment of the present invention; Figure 3 This is a flowchart of the steps of the preferred selection method in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the process of obtaining the target probability through cumulative probability in an embodiment of the present invention; Figure 5 This is a schematic diagram of the durability test bench design system in an embodiment of the present invention; Explanation of key component symbols:

[0013] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0014] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0015] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0016] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0017] To address the technical problems of low overall integration and narrow test coverage in existing durability test benches, which prevent them from adapting to multiple test projects, this application provides a durability test bench design method and system. Through hierarchical screening and convergence verification, it can accurately identify core samples that "can support multi-test integration and adapt to iterative needs," avoiding test bench design based on invalid or low-value samples. This eliminates the need to repeatedly develop multiple dedicated test benches; a single highly integrated test bench designed based on the target samples can cover multiple test scenarios, significantly reducing test bench development costs and resource waste. Secondly, through convergence judgment and iterative update mechanisms, it ensures that the final target samples all meet the "evaluation score convergence" condition, meaning that the test bench parameters corresponding to the sample data have consistency and stability. This ensures that the durability test bench designed based on these samples avoids resource mismatch problems caused by fluctuations in test bench parameters. Furthermore, this application optimizes the sample dynamically, that is, the replacement sample is selected from the original sample while maintaining the target number of the secondary sample set. Different sample data from the original sample set can be flexibly included, so that the resulting test bench can adapt to the multi-dimensional parameter changes after product iteration. It can meet higher environmental simulation requirements without frequent test bench modifications, thus solving the pain point of mismatch between environmental simulation capabilities and iteration needs.

[0018] To facilitate understanding of the present invention, several embodiments are given below. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that the disclosure of the present invention will be more thorough and complete.

[0019] Example 1 Please see Figure 1 The diagram shows a durability test bench design method in the first embodiment of the present invention, including steps S101-S107: S101. Obtain multiple samples from the durability test bench to construct an original sample set based on the sample data. Substitute the sample data into the evaluation function to calculate the evaluation score of each sample. Arrange the calculated multiple evaluation scores in descending order to select a first preset number of samples according to the evaluation scores to construct a first sample set. Combine the total number of samples in the original sample set to obtain the remaining samples.

[0020] In this embodiment, the original sample set includes multiple samples, each sample includes multiple sample data, and the sample data includes multiple types of sample data, including type A sample data, type B sample data, and type C sample data.

[0021] As a concrete example, such as Figure 2 As shown, before the step of substituting the sample data into the evaluation function to calculate the evaluation score for each sample, steps S1011-S1014 are included: S1011. Randomly sort the samples in the original sample set, with each sample including data from multiple classes.

[0022] Taking the drag test stand as an example, the test piece length is 800mm ( L axis > 800 mm, V> 4 m 3 The load-bearing temperature cycling test requirement is the operating baseline. m flow > 10 l / min, ΔT> 50 ℃ The iteration count is 100, the initial sample size is 50, and the bench design method includes: Sample data from various inverter durability test benches and dual-mode environment durability test benches were collected to construct an original sample set. The sample data included wheelbase. L axis Gear diameter d Number of gears N Incubator volume V Total power of the incubator P net Cooling water flow rate m flow Water cooling power P pump and the temperature difference ∆ between the medium temperature and room temperature T Among them, wheelbase L axis Gear diameter d Number of gears N Data for Class A samples; incubator volume V Total power of the incubator P netThis is data for Class B samples; cooling water flow rate. m flow Water cooling power P pump The temperature difference between the medium and room temperature ∆ T This is Class C sample data. Further, the medium includes cooling water and air, which can be selected based on the actual situation. Specifically, the sample data is shown in Table 1.

[0023] Table 1:

[0024] S1012. Divide the sorted samples into multiple groups by grouping adjacent samples into two groups.

[0025] As shown in Table 1, samples 1 and 2 are grouped together, samples 3 and 4 are grouped together, and so on, to obtain multiple groups of samples.

[0026] S1013. Cross-interchange one type of sample data in each group of samples to obtain the interchanged sample.

[0027] Select one type of sample data from a set of samples to be cross-interchanged, and perform data mutation on this type of sample data to obtain mutated sample data; then perform cross-interchange on the mutated sample data to obtain the interchanged sample; wherein the data mutation method includes ensuring that the change rate of a single sample data is no greater than 10% compared with the sample data before mutation. It should be further noted that in this embodiment, the cross-interchange is performed by matching high scores with low scores in the early stage (e.g., first and last, second and second to last), and by matching according to score order in the later stage (e.g., first and second, third and fourth), with 60% of the cycle count in the early stage and 40% in the later stage.

[0028] Specifically, as shown in Table 1, wheelbase L axis Gear diameter d Number of gears N Data for Class A samples; incubator volume V Total power of the incubator P net This is data for Class B samples; cooling water flow rate. m flow Water cooling power P pump The temperature difference between the medium and room temperature (25℃) ∆ T This is data from category C samples. a 1. a 2 and a 3 represents Class A sample data; a 4 and a 5 represents Class B sample data; a6. a 7 and a 8 represents the data for Class C samples.

[0029] In the first group constructed from samples 1 and 2, samples 1 and 2 are cross-exchanged to obtain new samples 1 and 2. Then, the data of the Class B samples in the new samples 1 and 2 are mutated and cross-exchanged to obtain the swapped samples. The mutation process is as follows: the total power of the incubator in the new sample 1 P net The original 95kW was autonomously mutated to 90kW due to the temperature chamber volume. V Total power of the incubator P net For similar sample data, there exists a corresponding mathematical relationship; therefore, the temperature chamber volume... V The value varies with the total power of the temperature chamber P net Changes occur due to variations; The temperature chamber is a cube, and the formula for calculating the insulation power minus the volume is: P’ net =6K 1 ·V (2 / 3) ·△T ; In the formula, P’ net K represents the insulation power; K1 is the insulation coefficient; ΔT represents the temperature difference between the medium temperature and the ambient temperature (25℃).

[0030] Similarly, in the second group constructed from samples 3 and 4, samples 3 and 4 are cross-exchanged to obtain new samples 3 and 4. Then, the type A sample data in the new samples 3 and 4 are mutated and cross-exchanged to obtain the swapped samples. The mutation process specifically involves: the axis distance in the new sample 3... L axis The original 1140mm diameter was autonomously modified to 1120mm, at which point the corresponding gear diameter... d =224mm, due to wheelbase L axis Gear diameter d Number of gears N For the data in category A, there exists a corresponding mathematical relationship; therefore, the number of gears... N The value varies with the axis distance L axisThe variation occurs due to the gearbox's variation, which can be calculated using the diameter-wheelbase formula. It should be further clarified that a sample includes Class A, Class B, and Class C sample data. In this embodiment, Class A sample data corresponds to the gearbox, Class B to the temperature chamber, and Class C to the water chiller. Regarding autonomous variation, each class of sample data can vary at most once. For example, if Class A sample data variates, its sample data can only vary by one of its components. Furthermore, the variation data for the gearbox is generally the wheelbase or gear diameter; the number of gears does not participate in the variation.

[0031] The formula for calculating the diameter-wheelbase is as follows: ; In the formula, d The diameter of the gear; L axis This refers to the wheelbase; N Let N be the number of gears. It should be further noted that in this calculation formula, only the gear diameter d participates in the variation; the number of gears N does not.

[0032] Specifically, the swapped samples are shown in Table 2.

[0033] Table 2:

[0034] S1014. Substitute the sample data of the swapped samples into the evaluation function to calculate the evaluation score for each sample.

[0035] Furthermore, in the technical solution of this application, the expression of the evaluation function is: Y=ω1· R heating +ω2· η gearbox +ω3· η cooling ; In the formula: Y represents the evaluation function; R heating For the temperature rise efficiency of the incubator; η gearbox For gearbox transmission efficiency; η cooling The cooling efficiency of the water chiller is ω1, ω2, and ω3 are coefficients. In this embodiment, ω1, ω2, and ω3 are taken as 0.4, 0.3, and 0.3, respectively. R heating The calculation formula is: ; In the formula: P net This refers to the total power of the incubator; ρThis refers to the air density inside the incubator. C p The specific heat capacity of the air inside the incubator; V K1 is the volume of the temperature chamber; K1 is the insulation coefficient; ΔT represents the temperature difference between the medium temperature and the ambient temperature (25℃). To match the actual power temperature rise system, in this embodiment, K1=60.

[0036] η gearbox The calculation formula is: ; In the formula: d is the gear diameter; N is the number of gears; μ The coefficient of friction; η cooling The calculation formula is: ; In the formula: m flow Water flow rate; C p,water ∆T represents the specific heat capacity of water; ∆T represents the temperature difference between the medium temperature and room temperature (25℃). P pump This refers to the power of the water chiller.

[0037] In this embodiment, the following can be calculated based on the data in Table 2: Y=ω1· R heating +ω2· η gearbox +ω3· η cooling =0.4×1.18+0.3×0.93+0.3×0.91=1.024.

[0038] in: according to We can obtain: R heating = (105*1000-6*60*8) 2 / 3 *65) / (1.2*1005*8)=1.18℃ / s; according to We can obtain: η gearbox = (1 - 0.01 * 3 / 0.45) = 0.93; according to We can obtain: η cooling =10*4186*65 / (60*50000)=0.91.

[0039] It should be further noted that the evaluation function Y is generally used for rough calculations to provide a reference direction for engineering design, and is not used for specific fine calculations.

[0040] S102. Filter the remaining samples according to the evaluation function and the selection probability to obtain a second preset number of samples to construct a second sample set.

[0041] Furthermore, in this embodiment, as Figure 3 As shown, roulette wheel is used as the selection method to obtain a second preset number of samples. The selection method includes: S1021. Summing up the evaluation scores of all individuals in the remaining samples according to the evaluation function to obtain the fitness of all individuals, and calculating the probability of each individual being selected based on the fitness.

[0042] The top 10% of data sets with the highest evaluation scores are retained, and the remaining 90% of data sets are used to construct the remaining sample. The fitness of all individuals in the remaining sample is then summed using the following formula: ; In the formula, n This represents the population size of the remaining samples.

[0043] As a concrete example, the probability P of each individual being selected is... i ) and its fitness W total Proportional, that is: ; In the formula, P( i ) represents the probability that each individual is selected.

[0044] S1022. Generate and obtain random numbers, calculate the cumulative probability starting from the low-scoring samples, and iterate through the cumulative probabilities based on the random numbers.

[0045] like Figure 4 As shown, in this embodiment, the generated random number Random(x)∈(0,1) is used to calculate the cumulative probability q starting from the low-scoring sample. i ), iterate through the cumulative probability q( i ), and then select the first one that satisfies q( i The cumulative probability of ≥Random(x) is used as the target probability. Target samples are obtained based on the target probability, and this process is repeated until the new generation population is full. In this embodiment, it is until the second sample set is full.

[0046] Specifically, the cumulative probability q( i The formula for calculating ) is: ; In the formula,n This represents the population size of the remaining samples.

[0047] S1023. Obtain the first cumulative probability that is not less than the current random number based on the traversal results, and use it as the target probability. Obtain the target sample based on the target probability.

[0048] like Figure 4 As shown, taking Random(x) as 0.6 as an example, i.e., Random(x) = 0.6, the cumulative probability q( i We can see that 0.6 is between 0.51 and 0.77. Therefore, the cumulative probability of the first value not less than 0.6 is 0.77, and the corresponding sample is 49. Therefore, sample 49 is the target sample.

[0049] S1024. Determine whether the number of target samples meets the second preset number.

[0050] If the target sample size does not meet the second preset number, return to step S1021 until the target sample size meets the second preset number. If the target sample size meets the second preset number, proceed to step S103.

[0051] S103. Construct a secondary sample set of the target number based on the first sample set, the second sample set, and the preset sample set.

[0052] Taking an original sample set containing 50 samples as an example, a first sample set is constructed by using 10% of the sample size as the first preset quantity, i.e., the first sample set contains 5 samples, leaving 45 samples. From the remaining 45 samples, 25 samples are selected to construct a second sample set, i.e., the second preset quantity is 25. Then, 20 samples are selected from the known bench design parameters as preset samples, i.e., the preset quantity is 20. By combining the 5 samples from the first sample set, the 25 samples from the second sample set, and the 20 preset samples, a new set of 50 samples is obtained, i.e., the target quantity is 50. Therefore, in this embodiment, the "target quantity of the secondary sample set" can be understood as the target quantity of the secondary sample set being 50.

[0053] S104. Calculate the evaluation score for each sample in the quadratic sample set based on the evaluation function.

[0054] S105. Determine whether the evaluation scores of all samples in the quadratic sample set have converged.

[0055] If the evaluation scores of all samples in the secondary sample set do not converge, then step S106 is executed, and the process returns to step S104 until the evaluation scores of all samples in the secondary sample set converge; if the evaluation scores of all samples in the secondary sample set converge, then step S107 is executed. It should be further noted that, in this embodiment, convergence means that the evaluation scores are stable within a certain range.

[0056] S106. Remove samples whose evaluation scores do not converge and obtain the number of samples to be removed. Randomly select a number of samples from the original sample set that correspond to the number of samples to be removed to obtain replacement samples. Update the secondary sample set according to the replacement samples so that the secondary sample set maintains the target number.

[0057] As a concrete example, if there are non-convergent sample evaluation scores in the secondary sample set, the corresponding samples are obtained based on these non-convergent evaluation scores and then removed. For instance, if there are four non-convergent sample evaluation scores in the secondary sample set, these four non-convergent samples are identified based on their scores and then removed. Since four samples are removed, to maintain the number of 50 samples in the secondary sample set, four new samples are randomly selected from the remaining 45 samples in the original sample set as replacement samples. These replacement samples are then used to fill the gaps left by the four removed non-convergent samples until the evaluation scores of all samples in the secondary sample set converge.

[0058] S107. Select a secondary sample set according to preset conditions to obtain the target sample, and design a durable test bench based on the sample data of the target sample.

[0059] When the evaluation scores of all samples in the secondary sample set converge, the target samples are obtained by screening according to preset conditions. For example, taking the load temperature cycling test as an example, the corresponding conditions are the preset conditions, which include... R heating >15, η gearbox >95%, m flow >10L / min, ∆T>100℃; the optimal sample data is selected based on these preset conditions and then used as the target sample.

[0060] In summary, the durability test bench design method in the above embodiments of the present invention, through hierarchical screening and convergence verification, can accurately identify core samples that "can support multi-experiment integration and adapt to iteration requirements," avoiding test bench design based on invalid or low-value samples. This eliminates the need for repeated development of multiple dedicated test benches; a single highly integrated test bench designed based on the target samples can cover multiple testing scenarios, significantly reducing test bench development costs and resource waste. Secondly, through convergence judgment and iterative mechanisms for supplementary updates, it ensures that the final target samples all meet the condition of "evaluation score convergence," meaning that the test bench parameters corresponding to the sample data have consistency and stability. This allows the durability test bench designed based on these samples to avoid resource mismatch problems caused by fluctuations in test bench parameters. Furthermore, this application uses dynamic sample optimization, where supplementary samples are selected from the original samples while maintaining the target number of secondary sample sets. Different sample data from the original sample set can be flexibly incorporated, enabling the resulting test bench to adapt to multi-dimensional parameter changes after product iteration. This allows it to meet higher environmental simulation requirements without frequent test bench modifications, solving the pain point of mismatch between environmental simulation capabilities and iteration needs. Example 2 Please see Figure 5 The image shows a durability test bench design system according to the second embodiment of the present invention. The system includes an environmental chamber 1009, two gearboxes 1004 located diagonally opposite each other in the environmental chamber 1009 for 1:1 towing transmission, and two dynamometers located at the power output end of the gearboxes 1004. The two dynamometers are diagonally arranged. In this embodiment, the dynamometers are compensated motors 1005. Two test pieces are provided in the environmental chamber 1009. The gearboxes 1004 are connected to the test piece differentials through a transmission shaft 1008. The dynamometers and the gearboxes 1004 form a four-quadrant motion simulation module for the test piece motor, including all working conditions such as motor braking, driving, and power generation.

[0061] The system also includes a load cabinet 1007 for simulating inverter loads, which is connected to a gearbox 1004 as its load. The system also includes a water chiller 1006, a power analyzer 1002, and a high-voltage power supply 1003. The water chiller 1006 is connected to two test pieces to provide cooling for their operation; the power analyzer 1002 is communicatively connected to the control module 1001 and the two test pieces; the high-voltage power supply 1003 is connected to the two test pieces to supply power. The system also includes a control module 1001, which includes a main control cabinet and a low-voltage power supply. The low-voltage power supply supplies power to the main control cabinet. The control module 1001 is communicatively connected to other components, including an environmental chamber 1009, test pieces, load cabinet 1007, water chiller 1006, high-voltage power supply 1003, and power analyzer 1002.

[0062] In this embodiment, gearboxes 1004, diagonally positioned on both sides of the environmental chamber 1009, are used for 1:1 towing transmission; a water chiller 1006 is used to provide cooling in real-time simulation of the vehicle assembly operation; a load cabinet 1007 is used to simulate the inverter load; a low-voltage power supply is also used to power the assembly oil pump and shift motor; an early failure power analyzer 1002 is used to analyze vibration data, and a genetic algorithm is used for optimizing the efficiency design of the main components under limited cost and power conditions. Furthermore, the water chiller 1006 includes two main branches, each with independently controllable temperature; each main branch is further divided into three sub-branches, and the flow rate of each sub-branch is adjustable.

[0063] As a specific example, the environmental chamber 1009 is located in the center of the test room. The test bench is located inside the environmental chamber 1009 and is equipped with mounting claws for mounting the sample. The test bench is equipped with a three-point suspension support to simulate the layout of the engine compartment of a vehicle. The suspension support is fixed on the mounting claws. The mounting claws consist of four crossbeams and three vertical beams. The positions of the crossbeams and vertical beams can be adjusted according to the size of the test piece to adapt to the sample.

[0064] Furthermore, load cabinet 1007 serves as a dummy load to simulate the motor. Three load cabinets are installed on the right side, using inductive loads to simulate the motor. This allows for the simultaneous testing of three dual inverter controllers or six single inverter controllers for endurance testing.

[0065] The high-voltage power supply 1003 consists of a PDU cabinet and a battery simulator, which can meet the needs of 3 dual inverter controllers or 6 single inverter controllers. The battery simulator outputs DC power to the PDU cabinet, and the PDU cabinet actively controls the contactor to open and close. When the assembly is in the power generation mode, the electrical energy is fed back to the battery simulator through the DC bus, and then fed back to the grid through the transformer.

[0066] The low-voltage power supply also provides low-voltage power to the test piece for communication between the host computer and the test piece. It can simultaneously power 3 dual inverter controllers or 6 single inverter controllers. Each interface integrates overvoltage / overcurrent protection and is designed with a quick-connect plug for easy power access.

[0067] The main control cabinet integrates various sensor signals of the test piece, test piece control board signals, gearbox vibration signals, dynamometer operating status signals, water chiller flow and temperature control / monitoring signals, environmental chamber temperature and humidity control / monitoring signals, low-voltage power supply status signals, high-voltage power supply status signals, and battery simulator status signals. Combined with the power analyzer & DETLA1002, it controls / monitors all the signals required for the test and records them to a shared disk without entering the test room.

[0068] The power analyzer & DETLA1002 enables fault early warning and health management by analyzing minute changes in vibration signals. Furthermore, an efficiency optimization design genetic algorithm is used to iteratively optimize design schemes and testing costs under constraints of cost and power.

[0069] In summary, the durability test bench design system in the above embodiments of the present invention, through hierarchical screening and convergence verification, can accurately identify core samples that "can support multi-experiment integration and adapt to iteration requirements," avoiding test bench design based on invalid or low-value samples. This eliminates the need to repeatedly develop multiple dedicated test benches; a single highly integrated test bench designed based on the target samples can cover multiple testing scenarios, significantly reducing test bench development costs and resource waste. Secondly, through convergence judgment and iterative mechanisms for supplementary updates, it ensures that the final target samples all meet the condition of "evaluation score convergence," meaning that the test bench parameters corresponding to the sample data have consistency and stability. This allows the durability test bench designed based on these samples to avoid resource mismatch problems caused by fluctuations in test bench parameters. Furthermore, this application uses dynamic sample optimization, where supplementary samples are selected from the original samples while maintaining the target number of secondary sample sets. Different sample data from the original sample set can be flexibly incorporated, enabling the resulting test bench to adapt to multi-dimensional parameter changes after product iteration. This allows it to meet higher environmental simulation requirements without frequent test bench modifications, solving the pain point of mismatch between environmental simulation capabilities and iteration needs.

[0070] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0071] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. 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 modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for designing a durable test bench, characterized in that, include: Multiple samples from the durability test bench are obtained to construct an original sample set based on the sample data. The sample data is substituted into the evaluation function to calculate the evaluation score of each sample. The calculated multiple evaluation scores are arranged in descending order to select a first preset number of samples in order of evaluation scores to construct a first sample set. The remaining samples are obtained by combining the total number of samples in the original sample set. The remaining samples are filtered according to the evaluation function and the selection probability to obtain a second preset number of samples to construct a second sample set. A target number of secondary sample sets are constructed according to the first sample set, the second sample set and the preset sample set. The evaluation score of each sample in the secondary sample set is calculated according to the evaluation function. Determine whether the evaluation scores of all samples in the quadratic sample set converge; If not, remove samples whose evaluation scores do not converge and obtain the number of samples to be removed. Then, randomly select a number of samples from the original sample set corresponding to the number of samples to be removed to obtain replacement samples. Update the secondary sample set according to the replacement samples to maintain the target number of samples in the secondary sample set. Then, return to the step of calculating the evaluation score of each sample in the secondary sample set according to the evaluation function until the evaluation scores of all samples in the secondary sample set converge. If so, the secondary sample set is filtered according to preset conditions to obtain the target sample, and the durable test bench is designed based on the sample data of the target sample.

2. The durability test bench design method according to claim 1, characterized in that, The original sample set includes multiple samples, each sample includes multiple sample data, and the sample data includes multiple classes of sample data, including class A sample data, class B sample data, and class C sample data; Before the step of substituting the sample data into the evaluation function to calculate the evaluation score for each sample, the following steps are also included: The samples in the original sample set are randomly sorted, and each sample includes data from multiple classes. The sorted samples are divided into multiple groups by grouping adjacent samples together. In each sample group, one type of sample data is cross-exchanged to obtain the cross-exchanged sample; The sample data of the swapped samples are substituted into the evaluation function to calculate the evaluation score for each sample.

3. The durability test bench design method according to claim 2, characterized in that, The steps for cross-interchanging one class of sample data in each sample group to obtain the cross-interposed sample include: Select one type of sample data from a set of samples to be cross-exchanged, and perform data mutation on this type of sample data to obtain the mutated sample data; The mutated sample data are cross-interchanged to obtain the interchanged sample; The methods for data mutation include: Compared with the original sample data, the rate of change of a single sample data after mutation is no more than 10%, and it gradually decreases with the number of iterations until it reaches 0%.

4. The durability test bench design method according to claim 1, characterized in that, The steps of filtering the remaining samples according to the evaluation function and the selection probability to obtain a second preset number of samples to construct a second sample set include: Roulette wheel selection is used as the selection method to obtain a second preset number of samples. The selection method includes: The fitness of all individuals is obtained by summing the evaluation scores of all individuals in the remaining samples according to the evaluation function, and the probability of each individual being selected is calculated based on the fitness. Generate and obtain random numbers, calculate the cumulative probability starting from the low-scoring samples, and iterate through the cumulative probabilities based on the random numbers; The cumulative probability of the first random number not less than the current random number is obtained based on the traversal results and used as the target probability. The target sample is then obtained based on the target probability. Determine whether the number of target samples meets the second preset number; If not, return to the step of summing the evaluation scores of all individuals in the remaining samples according to the evaluation function to obtain the fitness of all individuals, until the target sample size meets the second preset number.

5. The durability test bench design method according to claim 1, characterized in that, The expression for the evaluation function is: Y=ω1· R heating +ω2· η gearbox +ω3· η cooling ; In the formula: Y represents the evaluation function; R heating For the temperature rise efficiency of the incubator; η gearbox For gearbox transmission efficiency; η cooling ω1, ω2, and ω3 are the cooling efficiency of the water chiller; where: R heating The calculation formula is: ； In the formula: P net This refers to the total power of the incubator; ρ This refers to the air density inside the incubator. C p The specific heat capacity of the air inside the incubator; V K1 is the volume of the temperature chamber; K1 is the insulation coefficient; ΔT represents the temperature difference between the medium temperature and room temperature. η gearbox The calculation formula is: ； In the formula: d is the gear diameter; N is the number of gears; μ The coefficient of friction; η cooling The calculation formula is: ； In the formula: m flow Water flow rate; C p,water ΔT represents the specific heat capacity of water; ∆T represents the temperature difference between the medium temperature and room temperature. P pump This refers to the power of the water chiller.

6. A durable test bench design system, characterized in that, The system is used to implement the durability test bench design method according to any one of claims 1-5, wherein the method is specifically applied to the control module; the system includes an environmental chamber, two gearboxes located diagonally opposite to the environmental chamber, and two dynamometers located at the power output end of the gearboxes, wherein the two dynamometers are diagonally arranged; two test pieces are provided in the environmental chamber, and the gearboxes are connected to the test piece differentials via drive shafts; The system also includes a load cabinet for simulating inverter load, a water chiller, a power analyzer, and a high-voltage power supply. The load cabinet is connected to a gearbox to serve as the load of the load cabinet. The water chiller is connected to two test pieces to provide cooling for the operation of the test pieces. The power analyzer is communicatively connected to the control module and the two test pieces. The high-voltage power supply is connected to the two test pieces to supply power to the two test pieces. The system also includes a control module, which includes a main control cabinet and a low-voltage power supply. The low-voltage power supply powers the main control cabinet. The control module is communicatively connected to other devices, including an environmental chamber, test pieces, a load cabinet, a water chiller, a high-voltage power supply, and a power analyzer.

Images