基于RBFNN的智能车辆转向系统的预设性能控制

RBFNN-based prescribed performance control of intelligent vehicle steering system

针对存在模型非线性和参数不确定性的智能车辆转向系统的预设性能跟踪控制问题,采用径向基函数神经网络对转向系统中的不确定非线性进行在线逼近,结合障碍Lyapunov函数技术为智能车辆的线控转向系统设计预设性能控制器。在控制器设计中,采用动态增益技术补偿控制增益未知对系统控制性能的影响。利用Lyapunov方法分析系统的稳定性,证明在控制器作用下,前轮转角的跟踪误差在预设的时间内收敛至原点预设的邻域;通过数值仿真和整车实验验证了控制方法的合理性。

The prescribed performance tracking control problem of intelligent vehicle steering system with model nonlinearity and parameter uncertainty is studied.The RBFNN(Radial Basis Function Neural Network)is used to approximate the uncertain nonlinearity in the steering system.The prescribed performance controller is designed for the steer-by-wire system of intelligent vehicle based on the barrier Lyapunov function technology.In the design of the controller,the dynamic gain technology is used to compensate the effect of unknown control gain on the system control performance.Finally,the stability of the system is analyzed by Lyapunov method,and it is proved that the tracking error of the front wheel angle can converge to the prescribed neighborhood of the origin within the prescribed time under the action of the proposed controller.The rationality of the control method is verified by numerical simulation and vehicle experiment.