摘要
以国内某电动车型为例,利用Matlab/Simulink建立整车仿真模型,对电动轮汽车典型工况各驱动轮转矩变化规律进行了研究分析,结果表明,在各种典型工况下,电动轮汽车同轴左右侧车轮的理想驱动转矩存在一定差异和规律。在此研究基础上,建立了基于BP神经网络的各轮理想驱动转矩控制系统,仿真结果表明,神经网络控制系统可以实现电动轮汽车典型工况下各轮理想需求驱动转矩的控制,具有较强的泛化能力和计算速度。该控制方法为实现电动轮汽车驱动轮转矩的合理控制提供了新的途径。
The vehicle simulation model of one domestic electric car is established in the Matlab/Simulink. After analyzing the rule of each electric wheel＇s rotary torque, the ideal driving torque shows diversity and rule between left and right wheel on the same axis in many typical working conditions. Based on this research, the driving torque control system of ideal driving torque on each wheel is established based on the BP neural network. The simulation results show that neural network control system is able to drive precisely the electric wheel in every typical working conditions with better generalization ability and computing speed, it become a new research way of electric wheel torque controlling methods.
出处
《机械设计与制造》
北大核心
2016年第12期31-34,38,共5页
Machinery Design & Manufacture
关键词
神经网络
电动轮
驱动转矩
仿真
Neural Network
Electric Wheel
Rotary Torque
Simulation
