New energy automobile power assembly load spectrum construction method
By combining torque and speed counting, extrapolating the enhancement coefficient and Gaussian kernel function, performing hierarchical simplification, and optimizing with the Cuckoo search algorithm, the problem of not being able to simultaneously evaluate the performance of bearings and gears in traditional methods has been solved. This has enabled efficient and low-cost load spectrum construction, improving the accuracy and efficiency of powertrain testing for new energy vehicles.
Patent Information
- Application Number
- CN202511450790.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-11
- Publication Date
- 2025-11-21
AI Technical Summary
Traditional methods cannot simultaneously evaluate the performance of bearings and gears when constructing load spectra for new energy vehicle powertrain bench tests, resulting in the need for multiple tests and increased costs.
The method employs joint counting of torque-speed and torque-speed difference, introduces a strengthening coefficient and Gaussian kernel function for load spectrum extrapolation, hierarchical simplification and linear cumulative damage rule, and combines the cuckoo optimization algorithm to optimize parameters, thus integrating the damage assessment of bearings and gears.
The number of tests was reduced, the test cycle was shortened, the efficiency and accuracy of constructing the load spectrum of new energy vehicle powertrains were improved, and the cost was reduced.
Smart Images
Figure CN120995884A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of reliability test of new energy vehicle powertrain system, and particularly relates to a new energy vehicle powertrain load spectrum construction method. BACKGROUND
[0002] In the development process of the new energy vehicle powertrain, it is crucial to ensure the reliability of bearings and gears. In the traditional method of constructing a bench test load spectrum, only a single part can be subjected to fatigue damage equivalent conversion, which not only requires multiple tests, but also greatly increases the cost.
[0003] Therefore, there is an urgent need for a powertrain load spectrum construction scheme that can simultaneously evaluate the performance of bearings and gears to improve efficiency and reduce the cost of a new energy vehicle powertrain load spectrum construction method to overcome the deficiencies in current practical applications. SUMMARY
[0004] The present application aims to provide a new energy vehicle powertrain load spectrum construction method to solve the problems raised in the background.
[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme: a new energy vehicle powertrain load spectrum construction method, the method comprising:
[0006] Obtaining target user load data, the load data including motor torque, speed, and wheel speed signal data;
[0007] Based on different load condition data, torque-speed and torque-speed difference joint counting is performed, and the calculation results are superimposed based on different load conditions;
[0008] Introducing a reinforcement coefficient and a Gaussian kernel function to extrapolate the full life cycle load spectrum;
[0009] Classifying and simplifying the extrapolated load spectrum to generate a bench test spectrum;
[0010] Based on the linear cumulative damage rule, damage calculation is performed on the bench test spectrum and the extrapolated load spectrum;
[0011] Based on the cuckoo optimization algorithm, the parameters are optimized, and the errors in the optimization process are corrected.
[0012] As a further scheme of the present application, the step of obtaining target user load data specifically comprises:
[0013] Collecting motor torque, speed, and wheel speed signal data, and removing abnormal data based on the Laplace criterion;
[0014] Based on a preset interval, torque signal data and speed signal data are removed;
[0015] The missing data is filled based on a piecewise interpolation method.
[0016] Resampling is performed according to time specified frequency interpolation.
[0017] As a further scheme of the present application, the different load conditions include a forward driving condition, a reverse driving condition and a reverse running condition.
[0018] As a further scheme of the present application, the step of introducing the strengthening coefficient and the Gaussian kernel function for the life cycle load spectrum extrapolation specifically includes:
[0019] A unit mileage damage target value is determined based on a Weibull distribution, and a total damage target is calculated based on a target life mileage and the unit mileage damage target value.
[0020] A load spectrum strengthening coefficient is calculated based on the total damage target.
[0021] The Gaussian kernel function is extrapolated to generate an extrapolated load spectrum.
[0022] As a further scheme of the present application, the step of grading and simplifying the extrapolated load spectrum to generate a bench test spectrum includes: dividing the torque load amplitude and the rotating speed in the extrapolated load spectrum into multiple grades according to an unequal interval method, and simplifying the load spectrum.
[0023] As a further scheme of the present application, the step of calculating damage based on a linear cumulative damage rule for the bench test spectrum and the extrapolated load spectrum specifically includes:
[0024] Total fatigue damage of the bearing and the gear is calculated.
[0025] A damage form of the gear and the bearing is determined.
[0026] Damage matrices of the bench test spectrum and the extrapolated load spectrum are determined based on the damage form of the gear and the bearing and the total fatigue damage of the bearing and the gear.
[0027] A target function is constructed based on the damage matrices of the bench test spectrum and the extrapolated load spectrum.
[0028] As a further scheme of the present application, the damage form of the gear and the bearing includes bearing contact failure, gear contact failure and gear bending failure.
[0029] As a further scheme of the present application, the damage corresponding to the damage form of the gear and the bearing in the extrapolated load spectrum is respectively:
[0030] Bearing contact failure damage m1=3
[0031] Gear contact failure , m2 = 6.54;
[0032] Gear bending failure , m3 = 8.68;
[0033] wherein, is torque, unit N·m; , n is rotating speed, unit r / min; t is time, unit h.
[0034] As a further scheme of the present application, the target function is: ;
[0035] wherein, is the damage of the bench test spectrum, is the damage of the extrapolated load spectrum.
[0036] As a further scheme of the present application, the step of correcting the error in the optimization process specifically comprises:
[0037] calculating the frequency common divisor of each level load in the loading cycle, and designing single loading cycle damage;
[0038] calculating the correction cycle number of the bearing and the gear according to the optimized damage matrix and the damage matrix of the extrapolated load spectrum;
[0039] adjusting the actual loading cycle of the bearing and the gear based on the correction cycle number.
[0040] Compared with the prior art, the present application has the beneficial effects that: the present application realizes the construction of the synchronous fatigue damage equivalent load spectrum of the shaft and gear through the cuckoo optimization algorithm, integrates various damage evaluations of the bearing and the gear, reduces the test number and shortens the cycle. BRIEF DESCRIPTION OF DRAWINGS
[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application.
[0042] Figure 1 A flowchart of a new energy automobile power assembly load spectrum construction method provided by the embodiments of the present application.
[0043] Figure 2 A flowchart of the step of acquiring target user load data provided by the embodiments of the present application.
[0044] Figure 3 A flowchart of the step of introducing the reinforcement coefficient and the Gaussian kernel function to perform the full life cycle load spectrum extrapolation provided by the embodiments of the present application.
[0045] Figure 4 A flow chart of a damage calculation step of a linear cumulative damage rule based on a bench test spectrum and an extrapolated load spectrum is provided for the embodiment of the present application.
[0046] Figure 5 A flow chart of a step of correcting errors in the optimization process is provided for the embodiment of the present application. DETAILED DESCRIPTION
[0047] In order to make the technical problems to be solved by the present application, technical solutions and beneficial effects more clear, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.
[0048] Figure 1 A flow chart of a new energy vehicle power assembly load spectrum construction method is provided, and the new energy vehicle power assembly load spectrum construction method in the embodiment of the present application includes steps S100 to S600:
[0049] Step S100, target user load data is acquired, and the load data includes motor torque, speed, and wheel speed signal data;
[0050] Step S200, torque-speed and torque-speed difference joint counting is performed based on different load condition data, and the calculation results are superimposed based on different load conditions;
[0051] Step S300, a reinforcement coefficient and a Gaussian kernel function are introduced to perform life cycle load spectrum extrapolation;
[0052] Step S400, the load spectrum after extrapolation is classified and simplified to generate a bench test spectrum;
[0053] Step S500, damage calculation is performed on the bench test spectrum and the load spectrum after extrapolation based on a linear cumulative damage rule;
[0054] Step S600, parameters are optimized based on a cuckoo optimization algorithm, and errors in the optimization process are corrected.
[0055] As shown in Figure 2 As a preferred embodiment of the present application, the step of acquiring target user load data specifically includes:
[0056] Step S101, motor torque, speed, and wheel speed signal data are collected, and abnormal data is removed based on the Laplace criterion;
[0057] Step S102, torque signal data and speed signal data are removed based on a preset interval;
[0058] Step S103, filling the missing data based on the piecewise interpolation method;
[0059] Step S104, resampling according to the time specified frequency interpolation.
[0060] In the embodiment, the target user load data cleaning includes:
[0061] Collecting motor torque, speed and wheel speed signals, and removing abnormal data through the Lyapunov criterion;
[0062] Removing high-frequency noise of the torque signal, removing data less than ±1Nm in the torque signal, and compressing the torque and speed signals with the determination condition of [torque, speed]<[±1, ±1].
[0063] Removing abnormal data outside the torque and speed requirement interval, and filling the missing data by piecewise interpolation.
[0064] Data resampling includes:
[0065] Resampling by time specified frequency interpolation, extracting non-repetitive time, and solving the signal time alignment problem.
[0066] Interpolating at a uniform sampling frequency of 10hz per second.
[0067] As a preferred embodiment of the present application, the different load conditions include forward driving condition, reverse driving condition and reverse running condition.
[0068] In the embodiment, for the problem of opposite signs of left and right wheel speeds at low speed, the speed data less than 0.01 is assigned as 0, and the monthly data for statistical counting is obtained, and the typical load condition data is classified, and the definitions of each condition are shown in the following table.
[0069]
[0070] The torque-speed and torque-speed difference joint counting method is adopted to count the three types of typical conditions respectively, and the torque-speed joint counting results under the three types of typical conditions are obtained, and the load counting model of the bearing and gear parts is established:
[0071] The data classified according to the user typical load condition data after preprocessing is taken as input.
[0072] The required torque-speed data interval is divided into 128*128 level intervals, and the median value and boundary data of each small interval are taken.
[0073] The number of data in each speed-torque small interval in the input data is recorded as the frequency on the median value in this small interval, and is kept in the form of a matrix, and the T-n counting matrix is as follows:
[0074] ;
[0075] The count matrix is as follows:
[0076] ;
[0077] Wherein, represents the torque, unit N·m; represents the rotating speed, unit r / min; represents the rotating speed difference of left and right half shafts; represents the frequency corresponding to the torque, rotating speed (rotating speed difference).
[0078] The load superposition process includes:
[0079] After completing the multi-dimensional load joint counting, the load data under different typical working conditions needs to be superposed to integrate the load characteristics under different working conditions.
[0080] The superposed matrix is as follows:
[0081] ;
[0082] ;
[0083] As shown in Figure 3 , as a preferred embodiment of the present application, the step of introducing the strengthening coefficient and the Gaussian kernel function for life cycle load spectrum extrapolation specifically includes:
[0084] Step S301, determining the unit mileage damage target value based on the Weibull distribution, and calculating the total damage target based on the target life mileage and the unit mileage damage target value;
[0085] Step S302, calculating the load spectrum strengthening coefficient based on the total damage target;
[0086] Step S303, extrapolating the Gaussian kernel function to generate an extrapolated load spectrum.
[0087] In this embodiment, according to the fatigue engineering experience, the fatigue life / fatigue damage generally obeys the Weibull distribution. The probability density function of the Weibull distribution is:
[0088] ;
[0089] In the formula: respectively represent the shape and scale parameters.
[0090] The user monthly unit mileage damage is taken as the strength index, which should represent the original load unit mileage damage strength of 90% percentile users. The Weibull distribution probability curve is fitted through the original unit mileage damage, and the value corresponding to P=0.9 in the probability curve is taken as the 90% percentile user unit mileage damage target value :
[0091] ;
[0092] The distribution goodness-of-fit test method (Anderson-Darling) is adopted, based on the Weibull distribution model, combined with the least square method for parameter estimation, and the goodness-of-fit of the distribution is determined by the distribution goodness-of-fit test method:
[0093] ;
[0094] The distribution goodness-of-fit test method is tested by comparing and the critical value of the corresponding distribution cluster, at the significance level , the original hypothesis is accepted or rejected , wherein the smaller the AD value, the better the distribution fitting effect.
[0095] Mileage extrapolation and strengthening extrapolation process:
[0096] Combined with the target life mileage (such as = 240,000 km), the total damage target of the vehicle life cycle is calculated :
[0097] ;
[0098] According to the actual user total damage , the load spectrum strengthening coefficient K is calculated:
[0099] ;
[0100] The Gaussian kernel function extrapolation process includes:
[0101] Suppose that the random variable x follows an unknown probability density function f(x), and the distribution of the function f(x) is estimated by using the samples x1,..., xn drawn from it. Define a smooth function K(·) as the kernel function, and h is the smooth parameter (bandwidth). The bandwidth h=(b-a) is the length of the interval, and the number of samples falling into the interval [a, b] is k, then the relationship between the kernel function K(·) and the number of samples is:
[0102] ;
[0103] Then the probability density function For:
[0104] ;
[0105] The Gaussian kernel function used is:
[0106] ;
[0107] The corresponding two-dimensional kernel density estimation is:
[0108] ;
[0109] For another factor affecting the probability density function under the kernel density estimation principle, the bandwidth, there is an optimal bandwidth selection. Here, the adaptive bandwidth form is selected:
[0110] An adaptive factor is introduced to make the bandwidth change with the change of data, and get a more close to the actual distribution of the extrapolation results:
[0111] ;
[0112] In the formula, is the adaptive factor; is the probability density value at point ; N is the number of sample data; g is calculated by
[0113] .
[0114] The final adaptive kernel density estimation is:
[0115] ;
[0116] In the formula, N is the number of sample data.
[0117] After combining the mileage extrapolation, reinforcement extrapolation and Gaussian kernel function extrapolation, a set of extrapolated frequency data at each level of speed torque frequency data with expansion in frequency and speed torque can be obtained:
[0118] ;
[0119] ;
[0120] Repeat the above steps to get the and the respective extrapolation coefficients , the load spectrum after extrapolation for the calculation of target damage .
[0121] The extrapolated matrix form is:
[0122] ;
[0123] ;
[0124] As a preferred embodiment of the present invention, the step of grading and simplifying the extrapolated load spectrum to generate the bench test spectrum includes: dividing the torque load amplitude and speed in the extrapolated load spectrum into multiple levels according to the unequal interval method to simplify the load spectrum.
[0125] In this embodiment, the extrapolated load spectrum is reconstructed into a bench test spectrum. The torque load amplitude and speed in the extrapolated load spectrum are divided into multiple levels using an unequal interval method to simplify the load spectrum. For example, a five-level spectrum is selected as the test spectrum, and the level requirements should be:
[0126] ;
[0127] ;
[0128] ;
[0129] ;
[0130] ;
[0131] The initial values of T2, T3, and T4 are random numbers drawn from the initial interval. The frequency corresponding to the fifth level of torque. Initial value set to This is done in a format that ensures the initial frequency is not too low, thus reducing the number of iterations. Five-stage rotation speed. Set the initial value to a random number that satisfies the constraints.
[0132] Taking forward drive as an example, the matrix after initial input load classification is as follows:
[0133] .
[0134] like Figure 4 As shown, in a preferred embodiment of the present invention, the step of calculating the damage based on the bench test spectrum and the extrapolated load spectrum according to the linear cumulative damage rule specifically includes:
[0135] Step S501: Calculate the total fatigue damage of the bearings and gears;
[0136] Step S502: Determine the damage type of the gears and bearings;
[0137] Step S503: Determine the damage matrix of the bench test spectrum and the damage matrix of the extrapolated load spectrum based on the damage modes of gears and bearings and the total fatigue damage of bearings and gears.
[0138] Step S504, based on the damage matrix of the bench test spectrum and the damage matrix of the extrapolated load spectrum, a target function is constructed.
[0139] In this embodiment, based on the linear cumulative damage rule and the linear cumulative damage principle, the pseudo-damage is calculated by using the load-frequency life analysis method, and the fatigue strength S has the following relationship:
[0140] ;
[0141] In the formula, It is generally taken as 5 according to the material properties; is the fatigue life of the part (expressed in revolutions).
[0142] The total fatigue damage D of the gear is expressed as:
[0143] ;
[0144] In the formula is the number of revolutions (meshing frequency) of the torque interval in the observation period. The total damage represents the cumulative process of each individual damage .
[0145] The total fatigue damage D of the bearing is expressed as:
[0146] ;
[0147] In the formula: is the individual damage, is the fatigue life (failure revolution) at a fixed torque and a fixed rotational speed , and M is the number of revolutions (rotational revolutions) of the torque interval in the observation period.
[0148] The extrapolated load spectrum takes three common damage forms as the target damage for the calculation of the target function when constructing the bench test spectrum:
[0149] The bearing contact failure damage , m1=3;
[0150] The gear contact failure , m2=6.54;
[0151] The gear bending failure , m3=8.68;
[0152] Wherein, is the torque, unit N·m; , n is the rotational speed, unit r / min; t is the time, unit h.
[0153] The extrapolated load spectrum is reconstructed into a bench test spectrum, while considering the load characteristic difference between shaft parts and gear parts, a cuckoo search optimization algorithm is used to find the design parameter combination of minimizing cumulative damage D, so that the bench test spectrum can more accurately reflect the damage accumulation process under actual working conditions.
[0154] The damage of the corresponding torque frequency in the five-level spectrum is calculated according to three failure modes, and the test spectrum damage matrix { }, j = 1, 2, 3.
[0155] As a preferred embodiment of the application, in order to achieve the purpose that the test spectrum damage and the damage results of each type of failure mode after extrapolation are small, the objective function is: ;
[0156] Wherein, is the damage of the bench test spectrum, is the damage of the extrapolated load spectrum, (T, n, t, m).
[0157] In order to achieve the target and speed up the optimization speed, the method of optimizing the amplitude value at the same time is adopted in the actual code, and the difference value used for optimization is set to:
[0158] ;
[0159] The average value of the difference value is:
[0160] ;
[0161] The radiation value of the difference value is: ;
[0162] The final optimization target value is:
[0163] .
[0164] The constraint relationship construction process is:
[0165] The input power is:
[0166] ;
[0167] Wherein, is the torque of each level of the test spectrum, is the speed of each level of the test spectrum, is the maximum input power.
[0168] The total duration is: ;
[0169] Wherein is the frequency of each level of the test spectrum, For each level of rotational speed corresponding to the duration, 500 N·m;
[0170] Rotational speed: Where n max = 15000 r / min.
[0171] In order to avoid ending the iteration process before the optimal solution is obtained, a maximum value is set as the number of times of early ending of iteration, which is initially set to 5000000.
[0172] The final optimization target considers that when the error of the iteration result is small enough, the optimization is stopped, and the vehicle parameters at this time are output.
[0173] Based on the above process, the user's full life cycle random load spectrum under three typical working conditions is equivalent to the bench test spectrum, and the output result can be used as the loading input basis for durability or reliability test.
[0174] As shown in Figure 5 , as a preferred embodiment of the present application, the step of correcting the error in the optimization process specifically includes:
[0175] Step S601, calculate the frequency of each level of load in the loading period, design single loading period damage;
[0176] Step S602, calculate the correction cycle number according to the damage matrix after optimization and the damage matrix of the extrapolated load spectrum;
[0177] Step S603, adjust the actual loading period based on the correction cycle number.
[0178] In this embodiment, in order to eliminate the error of the optimization result relative to the target damage, the result obtained by optimization is corrected.
[0179] Design loading period:
[0180] By finding the greatest common divisor of , five loading levels in a period are designed :
[0181] , i = 1, 2, 3, 4, 5.
[0182] Bring into the calculation formula of the total fatigue damage D of the bearing and gear respectively:
[0183] , i = 1, 2, 3, 4, 5.
[0184] Obtain the damage of single loading cycle bearing And the damage of gear :
[0185] ;
[0186] ;
[0187] Calculate the error damage:
[0188] Record the three kinds of damage obtained by global optimization as , and the target damage , the error of the three kinds of damage obtained by global optimization compared with the target damage can be expressed as :
[0189] ;
[0190] The number of loading cycles required to correct the error:
[0191] ;
[0192] Therefore, when carrying out fatigue damage bench test of the three kinds of damage, the actual loading cycle of each kind of damage is:
[0193] , i=1, 2, 3, 4, 5.
[0194] If is positive, the number of loading cycles is less ; if it is negative, the number of loading cycles is more .
[0195] By adopting the cuckoo optimization algorithm, the optimal design parameter combination minimizing the cumulative fatigue damage D is found through global optimization. This method simultaneously considers the load characteristics difference between shaft parts and gear parts in the process of converting the load spectrum into the bench test loading spectrum, and realizes the one-time synchronous evaluation of bearing contact fatigue damage, gear contact fatigue damage and gear bending fatigue damage. Compared with the traditional method which needs to carry out three independent tests, the fatigue test period is shortened by 70%-80%, which greatly saves time and cost.
[0196] In the conversion process from 128-level detailed load spectrum to 5-level simplified loading spectrum, the invention allows to set fixed load boundaries, and by separating the torque load amplitude and speed in the data according to the unequal interval method, the torque amplitude and speed value of each level can be controlled, so that the whole process is more flexible and controllable.
[0197] To address the potential errors introduced by the global optimization algorithm, this invention introduces a periodic loading mechanism. By dynamically switching load levels and designing the loading cycle based on the least common divisor of the loading frequencies of different load levels, the loading cycle can be adjusted periodically for different types of damage during bench testing, thereby achieving error correction. This method solves the problem of non-equivalent damage transformation caused by global optimization errors, ensuring the accuracy of equivalent damage transformation.
[0198] In summary, this invention constructs the equivalent load spectrum of synchronous fatigue damage of bearings and gears through the Cuckoo Search algorithm, integrates multiple damage assessments of bearings and gears, reduces the number of tests and shortens the cycle, and ensures the accuracy of error correction and equivalent damage transformation through the periodic loading system, thus significantly improving the efficiency and accuracy of reliability testing of new energy vehicle products.
[0199] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for constructing a load spectrum of a powertrain of a new energy vehicle, characterized in that, The method comprises: acquiring target user load data, the load data including motor torque, rotating speed, and wheel speed signal data; jointly counting torque-rotating speed and torque-rotating speed difference based on different load working condition data, superimposing the calculation results based on different load working conditions; introducing reinforcement coefficients and Gaussian kernel functions to extrapolate the full life cycle load spectrum; grading and simplifying the extrapolated load spectrum to generate a bench test spectrum; calculating damage based on a linear cumulative damage rule for the bench test spectrum and the extrapolated load spectrum; optimizing parameters based on a cuckoo optimization algorithm and correcting errors in the optimization process.
2. The method according to claim 1, characterized in that, The step of acquiring target user load data specifically comprises: collecting motor torque, rotating speed, and wheel speed signal data, and eliminating abnormal data based on the Laplace criterion; eliminating torque signal data and rotating speed signal data based on a preset interval; filling in missing data based on a piecewise interpolation method; resampling according to a time-specified frequency interpolation.
3. The method according to claim 1, characterized in that, The different load working conditions include forward driving, reverse driving, and reverse running conditions.
4. The method according to claim 1 or 3, characterized in that, The step of introducing reinforcement coefficients and Gaussian kernel functions to extrapolate the full life cycle load spectrum specifically comprises: determining a unit mileage damage target value based on a Weibull distribution, and calculating an overall damage target based on a target service life mileage and the unit mileage damage target value; calculating a load spectrum reinforcement coefficient based on the overall damage target; extrapolating a Gaussian kernel function to generate an extrapolated load spectrum.
5. The new energy vehicle power assembly load spectrum construction method according to claim 1, characterized in that, The step of grading and simplifying the extrapolated load spectrum to generate a bench test spectrum comprises: dividing torque load amplitudes and rotating speeds in the extrapolated load spectrum into multiple grades according to a non-equal interval method, and simplifying the load spectrum.
6. The new energy vehicle power assembly load spectrum construction method according to claim 1, characterized in that, The step of calculating damage based on a linear cumulative damage rule for the bench test spectrum and the extrapolated load spectrum specifically comprises: calculating total fatigue damage of bearings and gears; determining damage forms of the gears and bearings; determining damage matrices of the bench test spectrum and the extrapolated load spectrum based on the damage forms of the gears and bearings and the total fatigue damage of the bearings and gears; constructing an objective function based on the damage matrices of the bench test spectrum and the extrapolated load spectrum.
7. The new energy vehicle power assembly load spectrum construction method according to claim 6, characterized in that, The damage forms of the gears and bearings include bearing contact failure, gear contact failure, and gear bending failure.
8. The new energy vehicle power assembly load spectrum construction method according to claim 7, characterized in that, In the extrapolated load spectrum, the damages corresponding to the damage forms of the gears and bearings are respectively: Bearing contact failure damage m1 = 3; Gear contact failure m2 = 6.54; Gear bending failure m3 = 8.68; wherein, is the torque, in N-m; n is the rotational speed, in r / min; and t is the time, in h.
9. The new energy vehicle power assembly load spectrum construction method according to claim 6 or 8, characterized in that, The objective function is: ; wherein, is the damage for the bench test spectrum, is the damage for the extrapolated load spectrum.
10. The method according to claim 1, wherein, The step of correcting errors in the optimization process specifically comprises: calculating frequency common divisors of each grade of load in a loading period, and designing single loading period damage; calculating a correction period number based on the optimized damage matrix and the damage matrix of the extrapolated load spectrum; adjusting actual loading periods of the bearings and gears based on the correction period number.