复杂环境下智能汽车自动驾驶系统研究
|
摘要
为了提高智汽车的驾驶安全性,研究了智能汽车自动驾驶系统,包括路径规划和跟踪两个方面。对于路径规划,在传统RRT算法基础上,将路径转角引入度量函数中,实现了对距离和平滑性的综合度量;将目标偏向策略和相向扩展策略引入到扩展方法中,将随机扩展改进为启发扩展,从而提出了CE-RRT算法,经仿真验证,CE-RRT算法不仅提高了路径规划成功率,而且缩短了规划时间。对于路径跟踪,建立了车辆路径跟踪的线性时变误差模型,设计了线性时变模型预测控制器,通过仿真验证,控制器可以实现对直线路径的完全跟踪,对圆形路径也有很高的跟踪精度。
In order to improve driving safety of intelligent vehicle,autonomous driving system of intelligent vehicle is studied,including path planning and tracking. For path planning,on basis of traditional RRT algorithm,corner angle is introduced to metric function,which can balance distance and smoothness. Target-tendency and meat each other extension strategies are introduced to extension method,which changes random extension to heuristic extension,so that CE-RRT algorithm is proposed. Clarified by simulation,CE-RRT algorithm not only can improve success rate,but can also shorten planning time.For path tracking,linear time varying error model of path tracking is built,and linear time varying model prediction controller is designed. It can be seen by simulation,the controller can track straight path completely,and can track circular path accurately.
In order to improve driving safety of intelligent vehicle,autonomous driving system of intelligent vehicle is studied,including path planning and tracking. For path planning,on basis of traditional RRT algorithm,corner angle is introduced to metric function,which can balance distance and smoothness. Target-tendency and meat each other extension strategies are introduced to extension method,which changes random extension to heuristic extension,so that CE-RRT algorithm is proposed. Clarified by simulation,CE-RRT algorithm not only can improve success rate,but can also shorten planning time.For path tracking,linear time varying error model of path tracking is built,and linear time varying model prediction controller is designed. It can be seen by simulation,the controller can track straight path completely,and can track circular path accurately.
引文
[1]Kala R,Warwick K.Multi-level planning for semi-autonomous vehicles in traffic scenarios based on separation maximization[J].Journal of Intelligent&Robotic Systems Theory&Applications,2013,72(3-4):559-590.
[2]Nash A,Daniel K,Koenig S.Theta*:Any-Angle Path Planning on Grids[J].Journal of Artificial Intelligence Research,2014,39(1):533-579.
[3]Kushleyev A,Likhachev M.Time-bounded lattice for efficient planning in dynamic environments[C]//IEEE International Conference on Robotics and Automation.IEE,,2009:1662-1668.
[4]陈波芝,陆亮,雷新宇.基于改进快速扩展随机树算法的双机械臂协同避障规划方法[J].中国机械工程,2018,29(10):1220-1226.(Chen Bo-zhi,Lu Liang,Lei Xin-yu.Simultaneous obstacle-avoidance motion planning approach for dual arm robots based on improved RRTalgorithm[J].China Mechanical Engineering,2018,29(10):1220-1226.)
[5]Hundelshausen F V,Himmelsbach M,Hecker F.Driving with tentacles:integral structures for sensing and motion[J].Journal of Field Robotics,2008,25(9):640-673.
[6]Zhang K,Sprinkle J,Sanfelice R G.Computationally Aware Control of Autonomous Vehicles:A Hybrid Model Predictive Control Approach[M].Kluwer Academic Publishers,2015.
[7]Akhtar A,Nielsen C,Waslander S L.Path Following Using Dynamic Transverse Feedback Linearization for Car-Like Robots[J].IEEE Transactions on Robotics,2017,31(2):269-279.
[8]刘存香.汽车恒速下坡滑膜变结构控制的仿真研究[J].机械设计与制造,2011(2):126-128.(Liu Cun-xiang.Simulation research on variable structure control for automobile downhill at constant speed[J].Machinery Design&Manufacture,2011(2):126-128.)
[9]熊征伟.基于B样条和多目标遗传算法的三次元送料机械手轨迹规划[J].机械传动,2018,42(5):125-128.(Xiong Zheng-wei.Trajectory planning of three dimensional feeding manipulator based on B spline and multi-objective genetic algorithm[J].Journal of Mechanical Transmission,2018,42(5):125-128.)
[10]唐贤伦,李洋,李鹏.多智能体粒子群优化的SVR模型预测控制[J].控制与决策,2014(4):593-598.(Tang Xian-lun,Li Yang,Li Peng.Model predictive control based on SVR optimized by multi-agent particle swarm optimization algorithm[J].Control and Decision,2014(4):593-598.)
[11]Barzegari M M,Alizadeh E,Pahnabi A H.Grey-box modeling and model predictive control for cascade-type PEMFC[J].Energy,2017(127):611-622.
[2]Nash A,Daniel K,Koenig S.Theta*:Any-Angle Path Planning on Grids[J].Journal of Artificial Intelligence Research,2014,39(1):533-579.
[3]Kushleyev A,Likhachev M.Time-bounded lattice for efficient planning in dynamic environments[C]//IEEE International Conference on Robotics and Automation.IEE,,2009:1662-1668.
[4]陈波芝,陆亮,雷新宇.基于改进快速扩展随机树算法的双机械臂协同避障规划方法[J].中国机械工程,2018,29(10):1220-1226.(Chen Bo-zhi,Lu Liang,Lei Xin-yu.Simultaneous obstacle-avoidance motion planning approach for dual arm robots based on improved RRTalgorithm[J].China Mechanical Engineering,2018,29(10):1220-1226.)
[5]Hundelshausen F V,Himmelsbach M,Hecker F.Driving with tentacles:integral structures for sensing and motion[J].Journal of Field Robotics,2008,25(9):640-673.
[6]Zhang K,Sprinkle J,Sanfelice R G.Computationally Aware Control of Autonomous Vehicles:A Hybrid Model Predictive Control Approach[M].Kluwer Academic Publishers,2015.
[7]Akhtar A,Nielsen C,Waslander S L.Path Following Using Dynamic Transverse Feedback Linearization for Car-Like Robots[J].IEEE Transactions on Robotics,2017,31(2):269-279.
[8]刘存香.汽车恒速下坡滑膜变结构控制的仿真研究[J].机械设计与制造,2011(2):126-128.(Liu Cun-xiang.Simulation research on variable structure control for automobile downhill at constant speed[J].Machinery Design&Manufacture,2011(2):126-128.)
[9]熊征伟.基于B样条和多目标遗传算法的三次元送料机械手轨迹规划[J].机械传动,2018,42(5):125-128.(Xiong Zheng-wei.Trajectory planning of three dimensional feeding manipulator based on B spline and multi-objective genetic algorithm[J].Journal of Mechanical Transmission,2018,42(5):125-128.)
[10]唐贤伦,李洋,李鹏.多智能体粒子群优化的SVR模型预测控制[J].控制与决策,2014(4):593-598.(Tang Xian-lun,Li Yang,Li Peng.Model predictive control based on SVR optimized by multi-agent particle swarm optimization algorithm[J].Control and Decision,2014(4):593-598.)
[11]Barzegari M M,Alizadeh E,Pahnabi A H.Grey-box modeling and model predictive control for cascade-type PEMFC[J].Energy,2017(127):611-622.