摘要
为探究车辆路径跟踪算法性能差异及适用性规律,分别构建基于预瞄的纯跟踪(PP)算法、前轮反馈控制算法和模型预测控制(MPC)算法的车辆模型,在圆环、蛇行和低、中、高速等工况下进行Simulink和CarSim联合仿真并分析横向控制效果。结果表明:PP算法预瞄距离越短,横向控制精度越高,但稳定性越弱,低速鲁棒性较好;前轮反馈控制算法在高速下具有更小的横向控制误差,且增益系数k影响控制精度;相较于前两者,MPC算法在不同速度下均具有良好的横向跟踪性能。最后,开展实车试验对比验证3种算法的跟踪性能,结果表明,在同一工况下,MPC算法具有更优的跟踪性能。
To explore the performance differences of vehicle path tracking algorithms and the applicability rules of each algorithm,vehicle models based on previewed Pure Pursuit(PP)algorithm,front wheel feedback algorithm,Model Predictive Control(MPC)algorithm are established respectively.Furthermore,Simulink&CarSim simulations are conducted under different conditions including circular ring,S-shaped&low,medium and high speed driving and later control effects are analyzed.The results indicate that,the shorter the preview distance in PP algorithm,the higher lateral control accuracy is,but the stability is weaker,and robustness is good under low-speed condition.Front wheel feedback algorithm has lower lateral control error under high-speed operation,moreover,gain coefficient k has effects on control accuracy.Compared with the other 2 algorithms,MPC algorithm has better lateral tracking performance at different speeds.Eventually,vehicle tests are verified and tracking performance of 3 algorithms are compared.The results indicate that MPC algorithm has better tracking performance under the same condition.
作者
赵颖
俞庭
张琪
杨洪涛
宋胜
Zhao Ying;Yu Ting;Zhang Qi;Yang Hongtao;Song Sheng(Southwest University,Chongqing 400715)
出处
《汽车技术》
CSCD
北大核心
2022年第7期15-24,共10页
Automobile Technology
基金
贵州省科技计划项目(黔科合支撑[2021]一般171)。
关键词
路径跟踪
预瞄理论
纯跟踪算法
前轮反馈控制算法
模型预测控制算法
Path tracking
Preview theory
Pure Pursuit(PP)algorithm
Front wheel feedbacks algorithm
Model Predictive Control(MPC)
