基于神经网络三自由度非线性四轮转向汽车控制仿真

摘要

为了改善四轮转向汽车的操纵稳定性,建立包含轮胎的非线性三自由度车辆模型,通过神经网络训练得到后轮控制器,利用神经网络控制器联合PID控制,分别与前轮转向、比例转向控制、横摆角速度反馈控制进行时域仿真对比.仿真结果表明:神经网络控制器联合模糊PID控制可以有效的控制车辆的质心侧偏角,减少横摆角速度的瞬态响应增益,缩短稳定时间,从而提高了车辆低速时的机动性和高速转向的稳定性,提高了运行车辆的安全性、平稳性.%To improve the stability of the four-wheel-steering vehicle, three degrees of freedom nonlinear four-wheel-steering vehicle dynamic model was established containing the tire mode,and the neural network controller of 4WS vehicle was trained Using the neural network controller united with fuzzy PID controller compared during the front wheel string, four wheel steering vehicle with proportional control, cross feedback the speed control with simulation. Simulation results show that the nerve network control united with PID controller can be effective control of yaw rate and reducing the stability of time, improving the safety and stability,which increases the mobility in low speed and the stability of the vehicle when steering in high speed.