A fuzzy neural network energy management strategy for parallel hybrid electric vehicle

Hybrid electric vehicle is a complex system to establish the accurate model and to improve the fuel economy. Many researchers aimed to find the optimal energy management strategy in recent years. In order to get the best torque range of the engine and the motor under the standard road condition, improve the fuel economy of the hybrid electric vehicle. In this paper, a parallel electric vehicle is selected to establish the mathematical model which including the vehicle model, powertrain model, and the battery model. And a fuzzy network control strategy was proposed which is based on the Adaptive Neuro Fuzzy Inference System optimization algorithm. After testing the algorithm on the ADVISOR software platform, the results shows that the membership functions have satisfactory influence. Taking the required torque of the clutch and the valve of SOC as the inputs of the fuzzy network controller, under the Urban Dynamometer Driving Schedule drive cycle conditions, the simulation shows that the engine can work in the efficient points of the engine workplace, and the fuel economy greatly improved compared with the traditional vehicle.