A torque distribution method for a dual-motor drive system based on cumulative fatigue damage equalization

Abstract

This application relates to the technical field of vehicle control, and in particular to a torque distribution method for a dual-motor drive system based on cumulative fatigue damage balancing. The method includes the following steps: data collection, including the driver's required torque, current gear information, and the current speed and torque of all motors; fatigue damage calculation, which calculates the instantaneous damage rate of the motors using a nonlinear damage model and integrates to obtain the cumulative fatigue coefficient; multi-objective optimization analysis, which calculates the life balancing cost and efficiency cost using the fatigue coefficient, and minimizes the output torque of each motor using weighted coefficients and the total cost; and torque output, where the vehicle controller distributes torque commands to each motor. The application of this application can realize a paradigm shift from optimal instantaneous energy consumption to optimal reliability throughout the entire life cycle, thereby significantly improving the overall durability and reliability of the powertrain while ensuring economy, and reducing maintenance costs and downtime risks caused by premature or uneven wear.

Description

Technical Field

[0001] This application relates to the technical field of vehicle control, and in particular to a torque distribution method for a dual-motor drive system based on cumulative fatigue damage balancing. Background Technology

[0002] In existing technologies, the torque distribution strategy of multi-motor drive systems mainly revolves around the single objective of optimal system efficiency (lowest energy consumption). The principle is that the vehicle controller (VCU) allocates the total required torque to the two motors according to the driver's required torque, combined with the efficiency MAPs of the two motors at different speeds, torques and corresponding gears, through table lookup and optimization algorithms, so that the system operates in the most efficient range.

[0003] However, a control method that solely pursues optimal system efficiency can easily lead to problems such as mismatched lifespans of drive chain components, reduced overall system reliability, and increased lifecycle costs. Summary of the Invention

[0004] To address the aforementioned issues, this application provides a torque distribution method for a dual-motor drive system based on cumulative fatigue damage balancing.

[0005] The torque distribution method for a dual-motor drive system based on cumulative fatigue damage balancing provided in this application adopts the following technical solution: A torque distribution method for a dual-motor drive system based on cumulative fatigue damage balancing includes the following steps: S1. Collect data, including the torque required by the driver, current gear information, and the current speed and torque of all motors; S2. Fatigue damage calculation: The instantaneous damage rate of the motor is calculated by using a nonlinear damage model and the cumulative fatigue coefficient is obtained by integration. S3. Multi-objective optimization analysis: calculate the life balance cost and efficiency cost through fatigue coefficient, and solve the output torque of each motor by minimizing the weight coefficient and total cost. S4, Torque Output: Torque commands assigned to each motor by the vehicle controller.

[0006] By adopting the above technical solution, this application overcomes the problems of mismatched lifespans of drivetrain components, reduced overall system reliability, and increased life-cycle costs caused by the single pursuit of optimal system efficiency in existing technologies, through the introduction of a life-cycle balance optimization objective. The application of this application enables a paradigm shift from optimal instantaneous energy consumption to optimal life-cycle reliability, thereby significantly improving the overall durability and reliability of the powertrain while ensuring economic efficiency, and reducing maintenance costs and downtime risks caused by premature or uneven wear.

[0007] As a preferred embodiment, in S2, when the motor speed is in the constant torque region of the motor external characteristic curve: based on the real-time torque T, the torque-damage coefficient nonlinear curve is queried to obtain the instantaneous damage coefficient f, and the coefficient is integrated to obtain the cumulative torque damage value Itrq in this region; When the motor speed is in the constant power region of the motor external characteristic curve: based on the real-time power P, look up the power-damage coefficient nonlinear curve to obtain the instantaneous damage coefficient k, and integrate the coefficient to obtain the cumulative power damage value Ipwr in this region. Calculate the cumulative mileage ODO to obtain the average damage coefficient of the motor D=(Itrq+Ipwr) / ODO. The average damage coefficient of motor one is D1, and the average damage coefficient of motor two is D2.

[0008] By adopting the above technical solution, the current damage status of the motor can be detected in real time by comparing the motor's usage with the motor's damage curve, and the damage status can be reflected in the form of data, which facilitates subsequent data processing.

[0009] Preferably, S3 includes torque distribution ratio calculation. S31. Calculate the basic torque distribution ratio αbase. Calculate the difference measure ΔD = (D1 - D2)^2 between the two motor damage states, and compare ΔD with the difference thresholds X0, X1, and X2, where 0 < 0 < 0. <X0<X1<X2, When ΔD≤X0, we assume the damage is balanced and let αbase=1; When X0 < ΔD ≤ X1, the damage deviation is small, and αbase is adjusted from 1 to 0.6 to suppress the load on the path with greater damage. When X1 < ΔD ≤ X2, the damage deviation is large, and αbase is adjusted from 0.6 to 0 to significantly reduce the burden on high-damage paths; When ΔD>X2, the damage deviation is too large. Let αbase=0, and drive it alone by the motor with less damage.

[0010] By adopting the above technical solution, the current degree of damage to the two motors is compared first. The motor with less damage is assigned to bear more output torque, rather than simply driving the motor in a way that prioritizes efficiency. This is to avoid the motor with more damage still bearing more output work, which would accelerate the motor's damage and significantly shorten its service life.

[0011] As a preferred embodiment, the torque distribution ratio calculation also includes, S32. Calculate the basic allocation ratio α. Introduce a motor available torque adjustment factor q, and the maximum available torque of motor one at the current speed is T1k, and the maximum available torque of motor two at the current speed is T2k; If D1>D2, then q=T1k / (T1k+T2k); If D1≤D2, then q=T2k / (T1k+T2k); α = αbase × q.

[0012] By adopting the above technical solution, taking into account the degree of damage to the motor and its own torque output capability, the output ratio of the two motors is allocated to achieve the optimal output state and ensure the service life of the two motors.

[0013] As a preferred embodiment, the torque distribution ratio calculation also includes, S33. Calculate the feasible interval of the basic allocation ratio α. Let the total required torque be Tz, the available torque of the motor with greater damage be Td, and the available torque of the motor with less damage be Tx. Then α must satisfy: Torque distribution for motors with significant damage: α×Tz≤Td; Torque distribution for motors with minimal damage: (1-α)×Tz≤Tx; Therefore, the feasible interval for α is obtained: The minimum value of α is αmin = max(0, 1 - Tx / Tz); The maximum value of α is αmax = min(1, Td / Tz).

[0014] As a preferred embodiment, the torque distribution ratio calculation also includes, S34. Determine the final basic allocation ratio α and construct the search space. Set a search radius δ, which defines the initial search range centered at α, i.e., the initial search interval [α-δ, α+δ]. The intersection of the initial search interval and the feasible interval [αmin, αmax] is the effective search interval. α Lower bound of the effective search space: αlow = max(αmin, α-δ); α: Upper limit of effective search space: αhigh = min(αmax, α + δ); Any value of α within this interval is called αi. All αi are traversed with a certain step size Δα. For each αi, the torque T1i of motor one is calculated as αi × Tz, and the torque T2i of motor two is calculated as (1-αi) × Tz. If there is no intersection between [α-δ, α+δ] and the feasible interval [αmin, αmax], directly assign α to αmin or αmax, and the search radius δ is 0.

[0015] As a preferred option, the computational efficiency cost includes, S35. Using the calculated search space, traverse all feasible torque distribution points (T1, T2); Calculate the overall system efficiency η=(T1·η1+T2·η2) / (T1+T2), and obtain the efficiency penalty term Feff(η)=1–η.

[0016] As a preferred option, the computational lifetime balancing cost includes, S36. Determine the region of the motor's external characteristics where the torque distribution point is located. If the torque distribution point is in the constant torque region, look up the torque-damage coefficient nonlinear curve to obtain the damage coefficient r. The instantaneous damage rate is the damage coefficient r / vehicle speed v. If the torque distribution point is in the constant power region, look up the power-damage coefficient nonlinear curve to obtain the instantaneous damage rate r / v. The instantaneous damage rate of motor one is r1 / v, and the instantaneous damage rate of motor two is r2 / v. The instantaneous damage difference at the torque distribution point is Δr=(r1 / v-r2 / v)^2. The life balance penalty term Flife(D1,D2,T1,T2)=β1×ΔD+β2×Δr, where β1 and β2 represent the weighting coefficients of historical damage difference and current damage increment difference, respectively.

[0017] Preferably, the output torque of each motor is solved by minimizing the weighting coefficients and the total cost, including: S37. Construct the total cost function C: C = Weff × Feff(η) + Wlife × Flife(D1,D2,T1,T2) = Weff × (1-η) + Wlife × (β1 × ΔD + β2 × Δr), where the settings of Weff and Wlife reflect the degree of preference for the two objectives. The VCU uses an optimization algorithm to find the (T1,T2) combination that minimizes the total cost C and controls the motor output accordingly.

[0018] By employing the above technical solution, the system collects torque data from each drive path, calculates the instantaneous damage rate using a nonlinear damage model (where the damage rate increases exponentially with torque after exceeding a threshold), and integrates the results to obtain the cumulative fatigue coefficients (D1, D2). When making torque allocation decisions, the VCU not only calculates the system efficiency of each candidate scheme but also its corresponding lifespan balancing cost. Finally, a weighted total cost function is used to evaluate all feasible schemes, and the scheme with the minimum total cost is selected as the final output, thereby proactively balancing component lifespan while ensuring economic efficiency.

[0019] In summary, this application includes the following beneficial technical effects: This application achieves a leap from "energy management" to "health management" by balancing the two objectives of "optimal instantaneous efficiency" and "optimal life cycle cost". Attached Figure Description

[0020] Figure 1This is a schematic diagram of the motor external characteristic curve and damage calculation partition in the embodiment; Figure 2 This is a schematic diagram of the damage coefficient curve of the motor in the constant torque region in the embodiment; Figure 3 This is a schematic diagram of the damage coefficient curve of the motor in the constant power region in the embodiment; Figure 4 This is a schematic diagram of the torque distribution ratio mapping relationship based on damage differences in the embodiment; Figure 5 This is a schematic diagram illustrating the intersection between the search interval and the feasible interval in the embodiment; Figure 6 This is a schematic diagram illustrating the relationship between the search interval and the feasible interval in the embodiment where they do not intersect. Detailed Implementation

[0021] The present application will be further described in detail below with reference to all the accompanying drawings.

[0022] Example

[0023] This application discloses a torque distribution method for a dual-motor drive system based on cumulative fatigue damage balancing. The vehicle controller (VCU) includes a fatigue damage calculation module, a multi-objective optimization decision module, and a torque distribution execution module. The fatigue damage calculation module includes a threshold judgment unit, a nonlinear instantaneous damage rate calculation unit, and an integral accumulation unit. The multi-objective optimization decision module includes an efficiency cost calculation unit, a life balance cost calculation unit, and a weighting coefficient and total cost minimization solution unit. The specific control steps are as follows: S1. Collect data, including the torque required by the driver, the current gear information, and the current speed and torque of all motors.

[0024] S2. Fatigue damage calculation, refer to... Figure 1 The instantaneous damage rate of the motor is calculated by using a nonlinear damage model and the cumulative fatigue coefficient is obtained by integration.

[0025] Reference Figure 2 When the motor speed is in the constant torque region of the motor's external characteristic curve: based on the real-time torque T, query the torque-damage coefficient nonlinear curve to obtain the instantaneous damage coefficient f, and integrate this coefficient to obtain the cumulative torque damage value Itrq for this region.

[0026] Reference Figure 3 When the motor speed is in the constant power region of the motor's external characteristic curve: based on the real-time power P, query the power-damage coefficient nonlinear curve to obtain the instantaneous damage coefficient k, and integrate this coefficient to obtain the cumulative power damage value Ipwr in this region.

[0027] Calculate the cumulative mileage ODO, and obtain the average damage coefficient D of the motor as D = (Itrq + Ipwr) / ODO. The average damage coefficient of Motor 1 is D1, and the average damage coefficient of Motor 2 is D2.

[0028] S3. Multi-objective optimization analysis. Refer to Figure 4 to analyze the output torque of the two motors.

[0029] Torque distribution ratio calculation. S31. Calculate the basic torque distribution ratio αbase.

[0030] Calculate the difference metric value ΔD of the damage states of the two motors as ΔD = (D1 - D2)^2, and compare ΔD with the difference thresholds X0, X1, X2, where 0 < X0 < X1 < X2. The VCU determines the basic torque distribution ratio αbase according to the interval where ΔD falls, by looking up a table or using a mapping function.

[0031] When ΔD ≤ X0, it indicates that the damage of the two motors is balanced, and let αbase = 1; When X0 < ΔD ≤ X1, it indicates that the damage deviation of the two motors is small, and αbase is adjusted from 1 towards 0.6 to suppress the load on the path with greater damage; When X1 < ΔD ≤ X2, it indicates that the damage deviation of the two motors is large, and αbase is adjusted from 0.6 towards 0 to significantly reduce the burden on the high-damage path; When ΔD > X2, it indicates that the damage deviation of the two motors is too large, and let αbase = 0, and the motor with less damage drives alone to protect the motor with greater damage.

[0032] S32. Calculate the basic distribution ratio α.

[0033] Due to the available torque limit of the motor, introduce the motor available torque adjustment factor q to adjust the basic distribution ratio α to make it more in line with the actual capabilities of the motor. The maximum available torque of Motor 1 at the current speed is T1k, and the maximum available torque of Motor 2 at the current speed is T2k (the maximum available torque is determined by the motor external characteristic curve); If D1 > D2, it indicates that Motor 1 is the motor with greater damage, then q = T1k / (T1k + T2k); If D1 ≤ D2, it indicates that Motor 2 is the motor with greater damage, then q = T2k / (T1k + T2k); α = αbase × q.

[0034] S33. Calculate the feasible interval of the basic distribution ratio α.

[0035] Due to the limitation of the available torque of the motor, the practically feasible range of α is physically constrained. Let the total required torque be Tz, the available torque of the motor with greater damage be Td, and the available torque of the motor with less damage be Tx. Then α must satisfy: Torque distribution for motors with significant damage: α×Tz≤Td; Torque distribution for motors with minimal damage: (1-α)×Tz≤Tx; Therefore, the feasible interval for α is obtained: The minimum value of α is αmin = max(0, 1 - Tx / Tz); The maximum value of α is αmax = min(1, Td / Tz).

[0036] S34. Determine the final basic allocation ratio α and construct the search space.

[0037] Reference Figure 5 Let δ be the search radius. This radius defines the initial search range centered at α, i.e., the initial search interval [α-δ, α+δ]. The intersection of the initial search interval and the feasible interval [αmin, αmax] is the effective search interval. α Lower bound of the effective search space: αlow = max(αmin, α-δ); α: Upper limit of effective search space: αhigh = min(αmax, α + δ); Any value of α within this interval is collectively referred to as αi. All αi are iterated over with a certain step size Δα (e.g., 0.01). For each αi, the torque T1i = αi × Tz of motor one and the torque T2i = (1 - αi) × Tz of motor two are calculated.

[0038] Reference Figure 6 If there is no intersection between [α-δ, α+δ] and the feasible interval [αmin, αmax], directly assign α to αmin or αmax, set the search radius δ to 0, calculate the torque according to the boundary value and output it.

[0039] S35. Calculation of efficiency cost. The VCU uses the calculated search space and a step size of 10% of Tz to traverse all feasible torque distribution points (T1, T2).

[0040] Calculate the overall system efficiency η = (T1·η1 + T2·η2) / (T1 + T2), and obtain the efficiency penalty term Feff(η) = 1 – η S36. Calculate the life balance cost using the fatigue coefficient.

[0041] Determine the region of the motor's external characteristics where the torque distribution point is located. If the torque distribution point is in the constant torque region, consult the torque-damage coefficient nonlinear curve to obtain the damage coefficient r. The instantaneous damage rate is the damage coefficient r / vehicle speed v. If the torque distribution point is in the constant power region, consult the power-damage coefficient nonlinear curve to obtain the instantaneous damage rate r / v. The instantaneous damage rate of motor one is r1 / v, and the instantaneous damage rate of motor two is r2 / v.

[0042] The instantaneous damage difference at the torque distribution point is Δr=(r1 / v-r2 / v)^2.

[0043] The life balance penalty term Flife(D1,D2,T1,T2)=β1×ΔD+β2×Δr, where β1 and β2 represent the weighting coefficients of historical damage difference and current damage increment difference, respectively.

[0044] S37. Solve for the output torque of each motor by minimizing the weighting coefficients and the total cost. Construct the total cost function C: C = Weff × Feff(η) + Wlife × Flife(D1,D2,T1,T2) = Weff × (1-η) + Wlife × (β1 × ΔD + β2 × Δr), where the settings of Weff and Wlife reflect the degree of preference for the two objectives. The VCU uses an optimization algorithm to find the (T1,T2) combination that minimizes the total cost C and controls the motor output accordingly.

[0045] S4, Torque Output: Torque commands assigned to each motor by the vehicle controller.

[0046] For example: The total required torque Tz is 800 Nm, D1=0.6, D2=0.4, and the available torque of the motor is: T1k=500 Nm, T2k=400 Nm.

[0047] Calculate the damage difference and αbase: ΔD = (0.6 - 0.4)^2 = 0.04. Assuming X0 < ΔD ≤ X1, αbase = 0.8 is obtained according to the mapping rule.

[0048] The adjustment factor q = T1k / (T1k+T2k) = 500 / (500+400) = 0.5556.

[0049] The basic allocation ratio is calculated as α = αbase × q = 0.8 × 0.5556 = 0.4445.

[0050] Due to the limitation of the available torque of the motor, the practically feasible range of α must satisfy: Torque distribution for motor 1 (significant damage): α×800≤500→α≤0.625; Motor 2 (less damaged) torque distribution: (1-α)×800≤400→α≥1-400 / 800=0.5.

[0051] Therefore, the feasible interval for α is: αmin=max(0,1–T2k / Tz)=max(0,1-400 / 800)=0.5; αmax=min(1,T1k / Tz)=min(1,500 / 800)=0.625.

[0052] Construct a search space centered at α: with α=0.4445 as the center and a search radius δ=0.1, calculate the search interval as [0.3445, 0.5445]. This intersects with the feasible interval [0.5, 0.625] at [0.5, 0.5445].

[0053] With ΔT=10%*800Nm=8Nm as the torque step size, and the corresponding proportional step size Δα=ΔT / 800=0.01, the torque distribution combinations are traversed.

[0054] The total cost of traversing all torque distribution combinations: Option 1: α = 0.5, T1 = 400 Nm, T2 = 400 Nm Check efficiency MAP: Motor 1: 400 Nm, 2000 rpm → η1 = 0.89; Motor 2: 400 Nm, 2000 rpm → η2 = 0.89; The overall system efficiency η = (400 × 0.89 + 400 × 0.89) / 800 = 0.89.

[0055] The efficiency penalty term Feff = 1 - 0.89 = 0.11.

[0056] Assume the current instantaneous damage rates are r1 / v = 0.002 and r2 / v = 0.005; Instantaneous damage difference: Δr = (0.002 - 0.005)^2 = 0.000009; Historical damage difference: ΔD=0.04.

[0057] Lifetime equilibrium penalty term (assuming β1=0.6, β2=0.4): Flife=0.6×0.04+0.4×0.000009=0.024+0.0000036=0.0240036; Total cost (assuming Weff=0.7, Wlife=0.3): C=0.7×0.11+0.3×0.0240036=0.077+0.00720108=0.08420108.

[0058] Option 2: α=0.51, T1=408Nm, T2=392Nm.

[0059] The overall system efficiency η = (408 × 0.895 + 392 × 0.888) / 800 = 0.892; The efficiency penalty term Feff = 1 - 0.892 = 0.108; Instantaneous damage rates: r1 / v = 0.0021, r2 / v = 0.0048; Instantaneous damage difference: Δr = (0.0021 - 0.0048)^2 = 0.00000729; Historical damage difference: ΔD = 0.04 (unchanged).

[0060] Lifespan balance penalty: Flife=0.6×0.04+0.4×0.00000729=0.024+0.000002916=0.024002916.

[0061] Total cost: C=0.7×0.108+0.3×0.024002916=0.0756+0.007200875=0.082800875.

[0062] Option 3: α=0.52, T1=416Nm, T2=384Nm.

[0063] The overall system efficiency η = 0.894; The efficiency penalty term Feff = 1 - 0.894 = 0.106; Instantaneous damage rates: r1 / v = 0.0022, r2 / v = 0.0046; Instantaneous damage difference: Δr = (0.0022 - 0.0046)^2 = 0.00000576; Historical damage difference: ΔD = 0.04 (unchanged).

[0064] Lifespan balance penalty: Flife=0.6×0.04+0.4×0.00000576=0.024+0.000002304=0.024002304.

[0065] Total cost: C=0.7×0.106+0.3×0.024002304=0.0742+0.007200691=0.081400691.

[0066] Similarly, calculate the total cost C corresponding to α=0.53 and α=0.54, and select the one with the smallest C as the optimal allocation solution.

[0067] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for torque distribution of a dual-motor drive system based on accumulated fatigue damage equalization, characterized in that: Includes the following steps, S1. Collect data, including the torque required by the driver, current gear information, and the current speed and torque of all motors; S2. Fatigue damage calculation: The instantaneous damage rate of the motor is calculated by using a nonlinear damage model and the cumulative fatigue coefficient is obtained by integration. S3. Multi-objective optimization analysis: calculate the life balance cost and efficiency cost through fatigue coefficient, and solve the output torque of each motor by minimizing the weight coefficient and total cost. S4, Torque Output: Torque commands distributed by the vehicle controller to each motor. In S2, when the motor speed is in the constant torque region of the motor external characteristic curve: based on the real-time torque T, query the torque-damage coefficient nonlinear curve to obtain the instantaneous damage coefficient f, and integrate the coefficient to obtain the cumulative torque damage value Itrq in this region; When the motor speed is in the constant power region of the motor external characteristic curve: based on the real-time power P, look up the power-damage coefficient nonlinear curve to obtain the instantaneous damage coefficient k, and integrate the coefficient to obtain the cumulative power damage value Ipwr in this region. Calculate the cumulative mileage ODO to obtain the average damage coefficient of the motor D=(Itrq+Ipwr) / ODO. The average damage coefficient of motor one is D1, and the average damage coefficient of motor two is D2. The computational efficiency cost includes, S35. Using the calculated search space, traverse all feasible torque distribution points (T1, T2); Calculate the overall system efficiency η=(T1·η1+T2·η2) / (T1+T2), and obtain the efficiency penalty term Feff(η)=1–η, where η1 is the efficiency of motor one and η2 is the efficiency of motor two. The cost of calculating lifetime balancing includes, S36. Determine the region of the motor's external characteristics where the torque distribution point is located. If the torque distribution point is in the constant torque region, look up the torque-damage coefficient nonlinear curve to obtain the damage coefficient r. The instantaneous damage rate is the damage coefficient r / vehicle speed v. If the torque distribution point is in the constant power region, look up the power-damage coefficient nonlinear curve to obtain the instantaneous damage rate r / v. The instantaneous damage rate of motor one is r1 / v, and the instantaneous damage rate of motor two is r2 / v. The instantaneous damage difference at the torque distribution point is Δr=(r1 / v-r2 / v)^2. The life balance penalty term Flife(D1,D2,T1,T2)=β1×ΔD+β2×Δr, where β1 and β2 represent the weighting coefficients of historical damage difference and current damage increment difference, respectively.

2. The torque distribution method for a dual-motor drive system based on cumulative fatigue damage equalization according to claim 1, characterized in that: S3 includes torque distribution ratio calculation. S31. Calculate the basic torque distribution ratio αbase. Calculate the difference measure ΔD = (D1 - D2)^2 between the two motor damage states, and compare ΔD with the difference thresholds X0, X1, and X2, where 0 < 0 < 0. <X0<X1<X2, When ΔD≤X0, we assume the damage is balanced and let αbase=1; When X0 < ΔD ≤ X1, the damage deviation is small, and αbase is adjusted from 1 to 0.6 to suppress the load on the path with greater damage. When X1 < ΔD ≤ X2, the damage deviation is large, and αbase is adjusted from 0.6 to 0 to significantly reduce the burden on high-damage paths; When ΔD>X2, the damage deviation is too large. Let αbase=0, and drive it alone by the motor with less damage.

3. The torque distribution method for a dual-motor drive system based on cumulative fatigue damage equalization according to claim 2, characterized in that: Torque distribution ratio calculation also includes, S32. Calculate the basic allocation ratio α. Introduce a motor available torque adjustment factor q, and the maximum available torque of motor one at the current speed is T1k, and the maximum available torque of motor two at the current speed is T2k; If D1>D2, then q=T1k / (T1k+T2k); If D1≤D2, then q=T2k / (T1k+T2k); α = αbase × q.

4. The torque distribution method for a dual-motor drive system based on cumulative fatigue damage equalization according to claim 3, characterized in that: Torque distribution ratio calculation also includes, S33. Calculate the feasible interval of the basic allocation ratio α. Let the total required torque be Tz, the available torque of the motor with greater damage be Td, and the available torque of the motor with less damage be Tx. Then α must satisfy: Torque distribution for motors with significant damage: α×Tz≤Td; Torque distribution for motors with minimal damage: (1-α)×Tz≤Tx; Therefore, the feasible interval for α is obtained: The minimum value of α is αmin = max(0, 1 - Tx / Tz); The maximum value of α is αmax = min(1, Td / Tz).

5. The torque distribution method for a dual-motor drive system based on cumulative fatigue damage equalization according to claim 4, characterized in that: Torque distribution ratio calculation also includes, S34. Determine the final basic allocation ratio α and construct the search space. Set a search radius δ, which defines the initial search range centered at α, i.e., the initial search interval [α-δ, α+δ]. The intersection of the initial search interval and the feasible interval [αmin, αmax] is the effective search interval. α Lower bound of the effective search space: αlow = max(αmin, α-δ); α: Upper limit of effective search space: αhigh = min(αmax, α + δ); Any value of α within this interval is called αi. All αi are traversed with a certain step size Δα. For each αi, the torque T1i of motor one is calculated as αi × Tz, and the torque T2i of motor two is calculated as (1-αi) × Tz. If there is no intersection between [α-δ, α+δ] and the feasible interval [αmin, αmax], directly assign α to αmin or αmax, and the search radius δ is 0.

6. The torque distribution method for a dual-motor drive system based on cumulative fatigue damage equalization according to claim 1, characterized in that: The solution for minimizing the output torque of each motor using weighting coefficients and total cost includes: S37. Construct the total cost function C: C = Weff × Feff(η) + Wlife × Flife(D1,D2,T1,T2) = Weff × (1-η) + Wlife × (β1 × ΔD + β2 × Δr), where the settings of Weff and Wlife reflect the degree of preference for the two objectives. The VCU uses an optimization algorithm to find the (T1,T2) combination that minimizes the total cost C and controls the motor output accordingly.

Images