基于MFAPC的智能车辆跟随式自动驾驶纵向加速度控制算法

Longitudinal Acceleration Control of Intelligent Vehicle Following Autonomous Driving Based on MFAPC

针对智能驾驶车辆纵向动力学存在非线性、外界干扰等问题,提出一种基于紧格式的无模型自适应预测控制(MFAPC)车辆纵向加速度控制算法,并进行了仿真验证。前馈控制可以提高响应,并补偿系统的非线性特性;MFAPC算法不依赖被控系统的精确模型,仅利用了被控系统的输入/输出数据。前馈控制与MFAPC反馈控制协同作用,实现了对期望加速度快速、精确地跟踪。最后,通过与比例-积分-微分(PID)控制算法进行比较,所提出的控制方法能够在有外界干扰的情况下,仍具有较好的性能表现。

Aiming at the problems of nonlinearity and external interference in the longitudinal dynamics of intelligent driving vehicles,a model-free adaptive predictive control(MFAPC)vehicle longitudinal acceleration control algorithm based on compact format is proposed,and simulation is carried out.Feed-forward control improves response and compensates for the nonlinear charac-teristics of the system;The MFAPC algorithm does not rely on the exact model of the controlled system,but only utilizes the input/output data of the controlled system.Feed-forward control works in synergy with MFAPC feedback control to enable fast and accuratetracking ofdesired acceleration.Finally,by comparing with the proportion integration differentiation(PID)control algorithm,the proposed control method can still have good performance in the case of external interference.