Research On Torque Distribution Of Hybrid Electric Vehicle Based On Efficiency Optimal

There are two kinds of power to work at the hybrid electric vehicle. If you want to get the optimal efficiency and the features of reducing emissions,it must make the two power systems, the motor and engine,to work in a relatively efficient area.However, according to different driving cycles, hybrid electric vehicle energy management strategy should be adjusted, in order to make it automatically adjust the torque distribution relationship between the two power sources,following the change of the running condition.Ultimately it can achieve the goal of energy optimization.This article is based on Parallel Hybrid Electric Vehicle. Considered of engine, motor, power battery, transmission efficiency and the vehicle control strategy, in order to improve engine efficiency and reduce the fuel consumption of HEV, a method of vehicle torque distribution based on the engine working point optimization has been built.According to the vehicle dynamics structure of parallel hybrid electric vehicle and combination with the relevant data collected by multi-sensor of the vehicle.Based on the Matlab/Simulink,the simulation model of hybrid bus driver system is built, including the engine,electric motor, battery and forward simulation model of the vehicle. It could obtain efficiency analysis model of vehicle through efficiency analysis which form different work pattern of the HEV and the each part of power train components in the different energy transfer process. In order to achieve the purpose of optimal efficiency, this paper uses the adaptive genetic algorithm to optimize the optimal operating point of the engine, it uses three layer feed forward neural network to distribute the vehicle torque and uses auxiliary torque’ peak load shifting’ provided by electric motor to adjust to the engine working point in real time, it will make engine and electric motor work in near optimal curve. The weight of neural network is adjusted by the temporal difference method, realizing the real-time control. Because the car will be affected by time, environmental change and physical deterioration in the process of operation, so this paper introduced parallel computing and realized Neural network vehicle torque distribution method based on the optimized efficiency in engineering.Finally, construction of the xPC target hardware-in-the-loop system, selected the NEDC conditions and urban comprehensive to simulation. And contrasted the traditional control strategy and put forward in this paper to show that the torque distribution method in this paper has accuracy and practicability. At the same time, hardware-in-the-loop system simulation platform in this paper can effectively reduce the number of the real vehicle test, which has great advantages in cost saving and reduce development time etc.