基于事件触发的无人驾驶汽车路径跟随预测控制  

Event-triggered predictive path following control for unmanned autonomous vehicle

作　　者：张皓[1] 张洪铭 王祝萍[1] ZHANG Hao;ZHANG Hong-ming;WANG Zhu-ping(Department of Control Science and Engineering,Tongji University,Shanghai 201804,China)

机构地区：[1]同济大学控制科学与工程系

出　　处：《控制与决策》2019年第11期2421-2427,共7页Control and Decision

基　　金：国家自然科学基金项目(61773289);上海市自然科学基金项目(17ZR1445800);中央高校基本科研业务费专项资金项目

摘　　要：研究存在有界扰动的非线性无人车辆模型的路径跟随问题,提出一种基于事件触发的模型预测控制算法,与现有的基于时间周期的模型预测控制算法相比,可以在保证车辆对参考轨迹跟随准确性的同时减少跟随过程中求解优化问题的计算量,降低在线实时优化的计算负担.最后给出无人车运动学模型的仿真结果,对比所提出的控制算法与传统算法,验证了其有效性.This paper studies the path following problem of nonlinear unmanned vehicle models with bounded disturbances,and proposes an event-triggered model predictive control algorithm.Compared with the existing model predictive control algorithm based on time period,the proposed algorithm can guarantee the accuracy of following the reference trajectory of the vehicle and reduce the computational complexity of solving optimization problems in the following process,then it can reduce the computational burden of on-line real-time optimization.Finally,the simulation results of the unmanned vehicle kinematics model are given.Through the comparision with the traditional algorithm,the effectiveness of the proposed algorithm is verified.

关 键 词：无人驾驶汽车系统 事件触发控制 模型预测控制 路径跟随 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]