A Parameter Optimization Method for Hybrid Powertrain Based on Rearward Simulation of Wheel-End Conditions

Abstract

The present invention provides a method for optimizing the parameters of a hybrid powertrain based on rearward simulation of wheel-end conditions, including: preprocessing sample data and obtaining sample wheel-end power and vehicle speed data, and calculating to obtain wheel-end data; extracting the dynamic feature matrix and static feature matrix of the wheel-end data and performing feature clustering; constructing a rearward simulation model of the hybrid system to obtain the simulated fuel consumption; establishing a multi-objective optimization data model based on additional cost and fuel-saving rate; using the NSGA-II algorithm to solve the established multi-objective optimization mathematical model, and the solved powertrain parameters include the motor, engine, battery capacity, axle speed ratio, transmission model, and tire model to achieve the optimal power matching decision.

Description

Technical Field

[0001] The present invention relates to a method for optimizing the parameters of a hybrid powertrain based on rearward simulation of wheel-side working conditions, and pertains to the technical field of automotive intelligent networking. Background Art

[0002] With the rapid development of the automotive industry and the continuous progress of new energy vehicle technologies, power matching decisions play a crucial role in the vehicle design process. Traditional methods often rely on experimental tests and empirical judgments, which are not only time-consuming and laborious but also difficult to comprehensively cover the performance under various working conditions. To overcome this difficulty, modern vehicle design increasingly adopts simulation-based methods for power matching decisions.

[0003] The prior art uses a vehicle simulation model for power matching decisions by combining a cycle working condition road spectrum with AVL Cruise software. Generally, power boundaries are derived using dynamic performance indicators such as maximum vehicle speed, acceleration duration, and maximum climbing gradient to determine the feasible regions for engine, motor, and battery selection. A vehicle simulation model is built using AVL Cruise software, and the cycle working condition road spectrum is used as the input to simulate the economic and dynamic performance results of different matching parameters. Some optimization algorithms are combined to make power matching decisions for the simulation results. There are the following disadvantages:

[0004] (1) The cycle working condition road spectrum is generally a general working condition (such as c-wtvc, etc.) or a self-constructed working condition. Usually, it is synthesized based on the recombination of motion segments. The selection and recombination of motion segments will inevitably cause deviations between the constructed working condition and the characteristics of the original working condition;

[0005] (2) Usually, the typical cycle working condition road spectrum is only a time-vehicle speed spectrum. The characteristics of altitude and gradient are generally reflected in the speed spectrum in the form of acceleration conversion. However, hybrid vehicles need to combine gradient and braking cycle information for battery energy management strategies and balance matching verification. It is difficult to accurately match the motor and battery with a single time-vehicle speed spectrum;

[0006] (3) When building the AVL Cruise simulation model, parameters such as the total vehicle and cargo weight and the vehicle resistance coefficient need to be input. However, the actual vehicle load and vehicle resistance in the market are not fixed parameter values affected by the environment. Therefore, there are matching deviations in the simulation with fixed parameters. Summary of the Invention

[0007] To solve the problems existing in the prior art, the present invention proposes a method for optimizing the parameters of a hybrid powertrain based on rearward simulation of wheel-side working conditions. The specific technical solution is as follows:

[0008] A method for optimizing the parameters of a hybrid powertrain based on rearward simulation of wheel-side working conditions includes the following steps:

[0009] : Preprocess the sample data, obtain the sample wheel-end power and vehicle speed data, and calculate to obtain the wheel-end data;

[0010] : Extract the dynamic feature matrix and static feature matrix of the wheel-end data and perform feature clustering;

[0011] : Build a backward simulation model of the hybrid system and obtain the simulated fuel consumption;

[0012] : Establish a multi-objective optimization data model based on the additional cost and fuel-saving rate;

[0013] : Use the NSGA-II algorithm to solve the established multi-objective optimization mathematical model.

[0014] Preferably, the specific method is:

[0015] Based on the time of the rotational speed signal, use the linear interpolation method to obtain the corresponding values of the vehicle speed, torque, load, and slope signals, align the vehicle speed, rotational speed, torque, load, and slope data of the sample data, identify the driving condition and braking condition, and calculate the corresponding wheel-end power:

[0016] ;

[0017] represents the wheel-end power at the moment of braking condition, satisfying the vehicle speed , and the braking signal ; is the vehicle mass; is the gravitational acceleration; is the longitudinal slope, calculated from the altitudes at moment and moment; is the air density; is the air resistance coefficient; is the frontal area; is the rotating mass conversion coefficient; is the driving acceleration;

[0018] represents the wheel-end power at the moment of driving condition, satisfying the vehicle speed and the torque ;

[0019] represents the net engine output power at the moment of driving condition, Indicates The fan power at the moment when the driving condition is present, Indicates The air pump power at the moment when the driving condition is present, is the overall vehicle transmission system comprehensive efficiency at the current moment;

[0020] Obtain the wheel-end data of a single sample through the wheel-end power: .

[0021] Furthermore, the is calculated through the engine speed and engine torque at time t:

[0022] .

[0023] Furthermore, the The static feature matrix in is a two-dimensional distribution matrix of wheel-end power - vehicle speed, and the vehicle speed range is 0 to 120 , with an interval of 5 , and the wheel-end power range is -600 to 400 , with an interval of 50 ; the The dynamic feature matrix in is a state transition probability matrix of wheel-end features, specifically: segment the operation segments according to continuous driving conditions and continuous braking conditions, and extract the average power, duration, and power standard deviation as features for each operation segment. The average power range is -600 to 400 , with an interval of 100 , the duration range is 0 to 300 , with an interval of 60 , and the power standard deviation range is 0 to 200, with an interval of 50, to obtain 200 states and form a state transition probability matrix.

[0024] Furthermore, the method for feature clustering in the is as follows:

[0025] Apply the method to cluster the static feature matrix of the samples, and use the silhouette coefficient evaluation method to select the value to obtain types of typical working conditions, and then apply the method to cluster the dynamic feature matrix of each type of sample in the type. After two-layer clustering, types of typical working conditions are obtained; use the wheel-end data of the clustering center samples in each class to represent the features of the types of typical working conditions.

[0026] Furthermore, the backward simulation model of the hybrid system includes a wheel-side power distribution function based on an optimal energy management strategy. Input wheel-side data into the wheel-side power distribution function based on the optimal energy management strategy and configuration parameters of the motor, engine, battery capacity, axle ratio, transmission model, and tire model to calculate the simulation fuel consumption. The wheel-side power distribution function based on the optimal energy management strategy is as follows:

[0027] Under driving conditions, the expression for power distribution is as follows:

[0028] ;

[0029] In the formula: is the wheel-side driving power of the vehicle at time is the driving power allocated to the engine at time is the transmission efficiency of the engine power system, is the driving power allocated to the motor at time is the transmission efficiency of the motor power system, is the maximum power generated by the motor at time is the peak power of the motor at time is the rated power of the motor at time is the value of the battery at time value; is the lowest value allowed for normal operation of the battery; and are respectively the engine speed and engine torque corresponding to the lowest instantaneous fuel consumption rate of the engine under the conditions of simultaneously satisfying the motor and battery constraints and the feasible region of the transmission gear at time

[0030] Under braking conditions, the expression for power distribution is as follows:

[0031] ;

[0032] In the formula: is the wheel-side braking power of the vehicle at time is the braking power allocated to the motor at time is the The power required by mechanical braking when the moment exceeds the motor battery recycling capacity, is the maximum power recovered by the motor at the moment, is the highest value during normal battery operation;

[0033] Battery remaining charge calculation function:

[0034] ;

[0035] In the formula: is the motor efficiency at the moment, is the battery efficiency, represents the battery capacity;

[0036] The motor speed and torque are calculated through the vehicle speed at the current moment and the motor power as follows:

[0037] ;

[0038] In the formula: is the motor speed at the moment, is the gear position of the transmission at the moment, is the drive axle ratio at the moment, is the tire radius.

[0039] Furthermore, the feasible region condition of the transmission gear position is:

[0040] ;

[0041] In the formula: is the engine speed at the moment, , , and are the lower and upper limits of the shift speed respectively.

[0042] Furthermore, the simulation function for calculating the simulation fuel consumption in the is:

[0043] ;

[0044] In the formula: is the fuel consumption per 100 kilometers of the hybrid vehicle, is the fuel density, is the mileage of vehicle operation, is the engine torque at the moment, At the engine speed at a moment and the engine torque the fuel consumption rate.

[0045] Furthermore, the multi-objective optimization data model in

[0046] ;

[0047] In the formula: is the maximum power of the motor, is the maximum power of the engine, is the cost function, is the motor cost, is the battery cost, is the cost other than the motor and the battery;

[0048] The fuel saving rate function is:

[0049] ;

[0050] In the formula: is the market statistical fuel consumption data.

[0051] The present invention uses the wheel-side power and vehicle speed time-series working conditions obtained by big data clustering as inputs for hybrid powertrain parameter matching. The wheel-side power includes the power requirements under braking and driving conditions, and truly reflects the information characteristics of the total mass, slope, resistance, road and traffic conditions of the original working conditions. Based on the vehicle dynamics balance formula and the principle of the optimal energy management strategy, a wheel-side power distribution model and a backward simulation model of the hybrid system are constructed to solve the problems of the motion chain and power flow decomposition of the hybrid powertrain, and the fuel consumption values under different parameter combinations are obtained by simulation. A multi-objective function with low additional cost and high fuel saving rate is established, and the NSGA-II algorithm is applied to optimize and solve the powertrain parameters. The solved powertrain parameters include the motor, engine, battery capacity, axle speed ratio, transmission model, and tire model to achieve the optimal power matching decision. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is a flowchart of a method for optimizing hybrid powertrain parameters based on wheel-side working conditions backward simulation according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0054] As Figure 1 shown, a method for optimizing the parameters of a hybrid powertrain based on rearward simulation of wheel-side working conditions includes the following steps:

[0055] : Preprocess the sample data, obtain the sample wheel-side power and vehicle speed data, and calculate and obtain the wheel-side data;

[0056] : Extract the dynamic feature matrix and static feature matrix of the wheel-side data and perform feature clustering;

[0057] : Build a rearward simulation model of the hybrid system and obtain the simulation fuel consumption;

[0058] : Establish a multi-objective optimization data model based on additional cost and fuel-saving rate;

[0059] : Use the NSGA-II algorithm to solve the established multi-objective optimization mathematical model.

[0060] The data preprocessing in:

[0061] Taking a single vehicle as a sample, align the vehicle speed, rotation speed, torque, load, slope and other data of the sample data, identify the driving condition and braking condition, and calculate the corresponding wheel-side power:

[0062] ;

[0063] Denote the wheel-side power at the moment of braking condition, satisfying the vehicle speed , and the braking signal ; is the vehicle mass ( ); is the acceleration due to gravity ( ); is the longitudinal slope ( ), calculated from the altitude at moment and moment; is the air density ( ); is the air resistance coefficient; is the frontal area ( ); is the rotating mass conversion coefficient; is the driving acceleration ( );

[0064] denotes the wheel-end power at the moment of driving condition, satisfying the vehicle speed and the torque ;

[0065] denotes the net engine output power at the moment of driving condition, denotes the fan power at the moment of driving condition, which is obtained by looking up the table according to the current fan speed ; denotes the air pump power at the moment of driving condition, is the overall efficiency of the vehicle transmission system at the current moment;

[0066] The wheel-end data of a single sample is obtained through the wheel-end power: .

[0067] Since the acquisition frequencies of different signals are different, it is necessary to align the time axes of each signal. The acquisition frequency of the rotational speed signal is 1 s / time. Based on the time of the rotational speed signal, the corresponding values of the vehicle speed, torque, load, and slope signals are obtained by linear interpolation to align the time axes of all signals.

[0068] The is calculated through the engine rotational speed at time t and the engine torque :

[0069] .

[0070] Specifically, the static feature matrix in the is a two-dimensional distribution matrix of wheel-end power - vehicle speed. The vehicle speed range is 0 - 120 , with an interval of 5 , and the wheel-end power range is -600 - 400 , with an interval of 50 , to obtain the following matrix:

[0071] ;

[0072] where denotes that the vehicle speed range is within and the frequency proportion within the power range satisfies ;

[0073] The dynamic feature matrix in is the state transition probability matrix of the wheel-end features. Specifically, the operation segments are segmented according to continuous driving conditions and continuous braking conditions, and the average power, duration, and power standard deviation are extracted as features for each operation segment. The average power range is -600 to 400 , 100 is the interval, the duration range is 0 to 300 , 60 is the interval, the power standard deviation range is 0 to 200, 50 is the interval, and 200 states are obtained to form the state transition probability matrix:

[0074] ;

[0075] wherein represents the state , represents the probability of transitioning from the state to , and satisfies .

[0076] Furthermore, in the , the method for feature clustering is as follows: Apply the method to cluster the static feature matrix of the samples, and use the silhouette coefficient evaluation method to select the value to obtain types of typical working conditions. Then apply the method to cluster the dynamic feature matrix of each type of sample in the type. After two-layer clustering, types of typical working conditions are obtained; use the wheel-end data of the clustering center samples in each class to represent the features of types of typical working conditions.

[0077] Specifically, the backward simulation model of the hybrid system in the includes a wheel-end power distribution function based on an optimal energy management strategy. Input the wheel-end data and the configuration parameters of the motor, engine, battery capacity, axle ratio, transmission model, and tire model into the wheel-end power distribution function based on the optimal energy management strategy to calculate the simulation fuel consumption. The wheel-end power distribution function based on the optimal energy management strategy is as follows:

[0078] Under the driving condition and satisfying the motor and battery constraint conditions, power distribution is performed based on the principle of the lowest instantaneous consumption of the engine, and its expression is as follows:

[0079] ;

[0080] In the formula: is the driving power of the vehicle wheel side at time ), is the driving power allocated to the engine at time ), is the transmission efficiency of the engine power system, is the driving power allocated to the motor at time ), is the transmission efficiency of the motor power system, is the maximum power generated by the motor at time ), is the peak power of the motor at time ), is the rated power of the motor at time ), is the value of the battery at time is the lowest value allowed for the normal operation of the battery; and are respectively the engine speed ( ) and the engine torque ( ) corresponding to the operating point with the lowest instantaneous fuel consumption rate of the engine under the conditions of simultaneously satisfying the motor and battery constraints and the gearbox gear feasible region at time

[0081] The gearbox gear feasible region condition is:

[0082] ;

[0083] In the formula: is the engine speed at time , , and are respectively the lower limit and the upper limit of the shift speed.

[0084] Under the braking condition and satisfying the motor and battery constraint conditions, power distribution is carried out on the principle of recovering as much energy as possible, and its expression is as follows:

[0085] ;

[0086] In the formula: is the braking power of the vehicle wheel side at time ), is the braking power allocated to the motor at time ), is the power that needs to be provided by mechanical braking when the braking power at time exceeds the battery recovery capacity of the motor is the maximum power recovered by the motor at time ), which is related to the current battery power and the duration of continuous operation of the motor peak power is the highest value when the battery is operating normally;

[0087] The function for calculating the remaining battery power (assuming the initial battery power is 50%, = 50%):

[0088] ;

[0089] In the formula: is the motor efficiency at time , which is obtained by looking up the table according to the motor speed and torque; is the efficiency of the battery represents the battery capacity ( );

[0090] The motor speed and torque are calculated from the vehicle speed at the current time and the motor power :

[0091] ;

[0092] In the formula: is the motor speed at time , is the gear position of the transmission at time , is the drive axle ratio at time , is the tire radius

[0093] The simulation function for calculating the simulation fuel consumption in S3 is:

[0094] ;

[0095] In the formula: is the fuel consumption per 100 kilometers of the hybrid vehicle ( ), is the fuel density ( ), is the mileage of vehicle operation ( ), , is the vehicle speed at time is the engine torque at time is the engine speed at time and the fuel consumption rate under the engine torque , with the unit of , which is obtained by looking up the table according to the engine speed and torque.

[0096] Furthermore, the multi-objective optimization data model in the includes an additional cost function and a fuel saving rate function; the additional cost function is:

[0097] ;

[0098] In the formula: is the maximum power of the motor, is the maximum power of the engine, is the cost function, is the motor cost, is the battery cost, is the cost other than the motor and the battery; the cost is mainly related to the maximum power of the motor, the maximum power of the engine, and the battery capacity, and the cost function is mainly obtained through the agreement pricing standard between the vehicle manufacturer and the parts manufacturer. is the additional cost (yuan), which is the increased cost of the hybrid vehicle compared with the original diesel vehicle, mainly considering the new costs of the motor, the battery, and other controllers, as well as the cost saved by matching a small-displacement engine.

[0099] The fuel saving rate function is:

[0100] ;

[0101] In the formula: is the market statistical fuel consumption data, and the fuel saving rate is the proportion of the fuel consumption reduction of the hybrid vehicle compared with the original diesel vehicle.

[0102] For each typical working condition characteristic, with Taking [[ID=]] as the input, under a set of parameters configured as motor, engine, battery capacity, axle speed ratio, transmission model, and tire model, through the wheel-side power distribution function based on the optimal energy management strategy, the simulated fuel consumption is obtained, the fuel saving rate is calculated, and the additional cost is obtained. For different configuration parameters, through the backward simulation model, different additional costs and fuel saving rates can be obtained. In order to select the best configuration among multiple configurations, from the economic perspective, select the configuration with low cost and high fuel saving rate. Therefore, establish a multi-objective function. Only pursuing low cost, that is, small motor and battery power, and small battery capacity, the fuel saving rate is low at this time. Low cost and high benefits are contradictory and cannot be directly solved. Therefore, establish a dual-objective model, design and improve the use of the fast non-dominated sorting genetic algorithm to obtain the Pareto optimal solution set

[0103] Apply NSGA-2 (Non-dominated Sorting Genetic Algorithm) to solve this multi-objective problem:

[0104] NSGA-II is a commonly used algorithm for solving multi-objective optimization models at present. Compared with the traditional genetic algorithm, it introduces the elitist strategy and combines the immune genetic improvement algorithm, which can not only better retain the excellent individuals in the parent population, but also inhibit the individuals with too high concentration to prevent the algorithm from falling into the local optimal solution. In this model, the motor, engine, battery capacity, axle speed ratio, transmission model, and tire model are variables, are two objective functions, and set the NSGA-II algorithm parameters as follows: crossover probability , mutation probability , the population size is 100, and the maximum number of iterations is 400.

[0105] The method of the present invention uses the wheel-side power and vehicle speed time-series working conditions of big data clustering as the input for hybrid powertrain parameter matching. The wheel-side power includes the power requirements under braking and driving conditions, which truly reflects the information characteristics of the total mass, slope, resistance, road and traffic conditions of the original working conditions; based on the vehicle dynamics balance formula and the principle of the optimal energy management strategy, a wheel-side power distribution model and a backward simulation model of the hybrid system are constructed to solve the problems of the hybrid powertrain motion chain and power flow decomposition, and simulate and obtain the fuel consumption values under different parameter combinations; establish a multi-objective function with low additional cost and high fuel saving rate, and apply the NSGA-II algorithm to optimize and solve the powertrain parameters. The solved powertrain parameters include the motor, engine, battery capacity, axle speed ratio, transmission model, and tire model to achieve the optimal power matching decision.

[0106] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A parameter optimization method for a hybrid powertrain based on backward simulation of wheel-end conditions, characterized in that It includes the following steps: S1: Preprocess the sample data, obtain the wheel-end power and vehicle speed data of the sample, and calculate to obtain the wheel-end data; S2: Extract the dynamic feature matrix and static feature matrix of the wheel-end data and perform feature clustering to obtain the wheel-end data D of the clustering center samples; S3: Build a backward simulation model of the hybrid power system and calculate the simulation fuel consumption; The building of the backward simulation model of the hybrid power system includes a wheel-end power distribution function based on the optimal energy management strategy. Input the wheel-end data D and the configuration parameters of the motor, engine, battery capacity, axle ratio, transmission model, and tire model into the wheel-end power distribution function based on the optimal energy management strategy to calculate the simulation fuel consumption. The wheel-end power distribution function based on the optimal energy management strategy is as follows: Under driving condition, the expression for power distribution is as follows: Where: P + (t) is the wheel-side drive power of the hybrid vehicle at time t, P eng (t) is the drive power allocated to the engine at time t, η1 is the transmission efficiency of the engine power system, P ele+ (t) is the drive power allocated to the motor at time t, η2 is the transmission efficiency of the motor power system, P ele_max+ (t) is the maximum power generated by the motor at time t, P ele峰 (t) is the peak power of the motor at time t, P ele额 (t) is the rated power of the motor at time t, SOC(t) is the battery SOC value at time t, SOC low is the minimum SOC value allowed for the normal operation of the battery; n min_mf and T min_mf are respectively the engine speed and engine torque corresponding to the operating condition point with the lowest instantaneous fuel consumption rate of the engine under the conditions of simultaneously satisfying the motor and battery constraints and the gearbox gear feasible region at time t; Under braking condition, the expression for power distribution is as follows: Where: P - (t) is the wheel-side braking power of the hybrid vehicle at time t, P ele- (t) is the braking power allocated to the motor at time t, P 机械制动 (t) is the power that needs to be provided by mechanical braking at time t when it exceeds the motor battery recovery capacity, P ele_max- (t) is the maximum power recovered by the motor at time t, SOC max is the highest SOC value during normal battery operation; Battery remaining power calculation function: Where: η ele (t) is the motor efficiency at time t, η bat is the efficiency of the battery, W ess represents the battery capacity; The motor speed and torque are calculated based on the vehicle speed v(t) and the motor power P at the current moment ele (t): where: n ele (t) is the motor speed at time t, i g_ele (t) is the gearbox gear at time t, i 0_ele (t) is the drive axle ratio at time t, and r is the tire radius; The simulation function for calculating the simulation fuel consumption is: where: f fuel is the fuel consumption per 100 kilometers of the hybrid vehicle, ρ is the fuel density, S is the mileage of the vehicle operation, T eng (t) is the engine torque at time t, m f (n eng (t), T eng (t)) is the engine speed n eng (t) at time t and the fuel consumption rate under the engine torque T eng (t); S4: Establish a multi-objective optimization data model based on the additional cost and fuel-saving rate; The multi-objective optimization data model includes an additional cost function and a fuel-saving rate function; the additional cost function is: COST=f1(P mor , P eng , W ess ) = c pmor + c wess + c other Where: P mor is the maximum power of the motor, P eng is the maximum power of the engine, f1 is the cost function, c pmor is the cost of the motor, c wess is the cost of the battery, c other is the cost other than the motor and the battery; The fuel-saving rate function is: Fs = (fuel mark - f fuel ) / fuel mark × 100% where: fuel mark is the fuel consumption data statistically obtained from the market; S5: Use the NSGA-II algorithm to solve the established multi-objective optimization data model.

2. The parameter optimization method for a hybrid powertrain based on rearward simulation of wheel-end conditions according to claim 1, wherein The specific method of S1 is: Based on the time of the rotational speed signal, use the linear interpolation method to obtain the corresponding values of the vehicle speed, torque, load, and slope signals, align the vehicle speed, rotational speed, torque, load, and slope signals of the sample data, identify the driving condition and braking condition, and calculate the corresponding wheel-end power: P 制动 (t) represents the wheel-end power under the braking condition at time t, satisfying the vehicle speed v(t) > 0 and the braking signal break(t) > 0; m is the vehicle mass; g is the acceleration due to gravity; θ is the longitudinal slope, calculated from the altitudes at times t + 1 and t - 1; ρ air is the air density; C D is the air resistance coefficient; A is the frontal area; δ is the conversion coefficient of rotating masses; a is the driving acceleration; P 驱动 (t) represents the wheel-end power of the driving condition at time t, satisfying the vehicle speed v(t) > 0 and the torque T(t) > 0; P 发动机净输出 P(t) represents the net output power of the engine under the driving condition at time t. 风扇 P(t) represents the fan power under the driving condition at time t. 打气泵 P(t) represents the air pump power under the driving condition at time t, and η is the comprehensive efficiency of the vehicle transmission system at the current moment. Obtain the wheel-end data of a single sample through the wheel-end power.

3. A method for optimizing the parameters of a hybrid powertrain based on backward simulation of wheel-end conditions according to claim 2, characterized in that The described P 发动机净输出 (t) is calculated through the engine speed n eng (t) and the engine torque T eng (t) at time t, and is obtained as follows:

4. A method for optimizing parameters of a hybrid powertrain based on backward simulation of wheel-end conditions according to claim 2, characterized in that The static feature matrix in S2 is a two-dimensional distribution matrix of wheel-end power - vehicle speed. The vehicle speed range is 0 - 120 km / h, with an interval of 5 km / h, and the wheel-end power range is -600 - 400 kw, with an interval of 50 kw; the dynamic feature matrix in S2 is a state transition probability matrix of wheel-end features. Specifically: Segment the operation segments according to continuous driving conditions and continuous braking conditions, extract the average power, duration, and power standard deviation as features for each operation segment. The average power range is -600 - 400 kw, with an interval of 100 kw, the duration range is 0 - 300 s, with an interval of 60 s, and the power standard deviation range is 0 - 200, with an interval of 50, to obtain 200 states and form a state transition probability matrix.

5. A method for optimizing the parameters of a hybrid powertrain based on rearward simulation of wheel-end conditions according to claim 4, characterized in that In S2, the method for feature clustering is: Use the K-means method to cluster the static feature matrix of the sample, and use the silhouette coefficient evaluation method to select the value of K to obtain K1 types of typical working conditions. Then use the K-means method to cluster the dynamic feature matrix of each type of sample in K1 types. After two clusters, obtain K2 types of typical working conditions; Use the wheel-end data D of the clustering center samples in each class to represent the characteristics of K2 types of typical working conditions.

6. The parameter optimization method for a hybrid powertrain based on rearward simulation of wheel-end conditions according to claim 1, characterized in that The feasible region condition of the transmission gear is: Where: n eng (t) is the engine speed at time t, i g_eng (t) = i g_ele (t), i 0_eng (t) = i 0_ele (t), n min and n max are the lower limit and upper limit of the speed respectively.

Images