Predictive Control Method of Autonomous Vehicle Based on Tracking-error Model
LI Peixin,JIANG Xiaoyan,WEI Yanding,ZHOU Xiaojun
DOI: https://doi.org/10.6041/j.issn.1000-1298.2017.10.045
2017-01-01
Abstract:For the trajectory tracking problem of autonomous vehicle,on the basis of analysis of kinematic model of vehicle,a model based predictive control method for autonomous vehicle trajectory tracking was designed.Firstly,a linear error model of vehicle kinematics was obtained by using a successive linearization approach,and it was used to predict the future behavior of the vehicle.Secondly,based on this model,it was possible to get a sequence of optimal control by using the linear MPC method and minimizing the objective function,and the first element of this sequence was applied to the system.Lastly,three typical test trajectories (lane change course,figure eight course and road course) were designed and the tracking controller was tested in the virtual simulation platform.The platform was set up on real-time multi-body dynamics software Vortex and visual rendering software Vega Prime.In order to meet the real-time requirements of the platform,two computers were used for dynamic resolving and visual rendering respectively,and the high level architecture (HLA) was adopted to realize the synchronization and data interaction between Vortex and Vega Prime.Simulation results showed that this controller can track the reference trajectory quickly and stably,the distance error and heading error were in a reasonable range.The refresh rate of Vortex and Vega Prime was stabilized at about 30 Hz,the error was within ± 0.05 Hz,indicating that the controller can meet the real-time requirements of the system.