基于神经网络的4WD电动汽车AFS/DYC解耦控制研究

Research on AFS/DYC Decoupling Control of 4WD EV Based on Neutral Network

基于二自由度线性化模型,分析了质心侧偏角和横摆角速度之间的耦合现象;在七自由度非线性汽车模型基础上,利用BP神经网络构造了自适应解耦控制器。设计了2个车辆姿态参数控制器,以状态参数理想模型的输出作为控制目标,分别控制汽车的质心侧偏角和横摆角速度,实现车辆主动安全控制。仿真结果表明,基于BP神经网络的解耦控制能够跟踪状态参数的理想值,提高汽车在极限工况下的行驶安全性和操纵稳定性。

Based on a 2-freedom linear model, the coupling phenomenon between sideslip angle and yaw rate was analyzed. According to a 7-freedom nonlinear vehicle model, an adaptive decoupling controller based on BP neutral network was designed. Two vehicle attitude parameter controllers were designed which aims to converge the sideslip angle and the yaw rate to the outputs of the desire model respectively in order to realize the active safety control. Simulation results show that the decoupling control based on BP neutral network can track the desire values, so that it enhances the driving safety and the handling stability of the vehicle under extreme conditions.