基于BP神经网络的电动汽车电子差速器设计  

Design of electronic differential of electric vehicle based on BP neural network

作　　者：李全民 Li Quanmin(College of Intelligent Manufacturing and Electrical Engineering,Guangzhou Institute of Science Technology,Guangzhou 510540)

机构地区：[1]广州理工学院智能制造与电气工程学院,广州510540

出　　处：《现代计算机》2023年第3期102-107,共6页Modern Computer

基　　金：广州理工学院2021年科研重点项目(2021KZ001)。

摘　　要：在对汽车转向Ackerman模型分析基础上,设计了能用于电子差速器的BP神经网络架构。在Matlab环境中利用样本数据训练得到了最优权值并进行了测试,结果表明设计的神经网络差速器真实地反映了内外驱动轮转速与车体速度和转角之间的实际动态关系,能有效实现差速。最后提出了一种用于分布式驱动电动汽车的BP神经网络差速器设计方案。Based on the analysis of Ackerman model of vehicle steering,a BP neural network architecture for electronic dif⁃ferential is designed.In the Matlabenvironment,the optimal weights are obtained by training the sample data and tested.The re⁃sults show that the neural network differential truly reflects the nonlinear relationship between the speed of the inner and outer driving wheels and the speed and angle of the car body,and can effectively achieve differential speed.Finally,a differential design scheme based on BP neural network for distributed drive electric vehicles is proposed.

关 键 词：电动汽车 电子差速器 BP神经网络 

分 类 号：U469.72[机械工程—车辆工程]