基于经典进化策略的移动小车模型预测控制
Model Predictive Control for Mobile Cart Based on Canonical Evolution Strategies
DOI: 10.12677/CSA.2022.122049, PDF,   
作者: 陈肇江：广东工业大学自动化学院，广东 广州；程良伦：广东工业大学计算机学院，广东 广州
关键词: MPC；Canonical ES；移动小车；避障规划；运动规划；MPC； Canonical ES； Mobile Cart； Obstacle Avoidance Planning； Motion Planning
摘要: 模型预测控制(model predictive control, MPC)已经被证明了在无人驾驶、移动机器人导航、机械臂路径规划、过程控制等诸多领域都拥有着出色的控制性能，然而传统的MPC控制不具有进化能力，在复杂的环境下系统往往很难得到最优的控制效果。因此，提出了一种结合经典进化策略(canonical evolution strategies, Canonical ES)的模型预测控制算法，并用于双轮差速移动小车的避障运动规划。在移动小车目标逼近运动规划的基础上，通过训练策略网络可以使得处于未知环境的移动小车快速有效地学习避障策略，并且对于局部最优陷阱拥有更优异的处理方法。仿真实验结果证明了该方法的可行性与高效性。
Abstract: Model predictive control (MPC) has been identified to have excellent control performance in many fields such as unmanned driving, mobile robots navigation, path planning of manipulators and process control, etc. However, traditional MPC does have no ability to evolve, for which is often difficult for the system to get the best effects in complex environments. Therefore, the MPC algorithm combining canonical evolution strategies is proposed. The algorithm contributes to obstacle avoidance motion planning of a two-wheel differential mobile cart. Based on the goal-approximation motion planning of the mobile cart, the mobile cart is able to learn the obstacle avoidance strategies effectively in unknown environment by training the policy network. Moreover, mobile cart will move better when facing local optimal traps. The feasibility and efficiency of the method are verified by experimental results.
文章引用：陈肇江, 程良伦. 基于经典进化策略的移动小车模型预测控制[J]. 计算机科学与应用, 2022, 12(2): 486-497. https://doi.org/10.12677/CSA.2022.122049

参考文献

[1] Zhong, M., Johnson, M., Tassa, Y., Erez, T. and Todorov, E. (2013) Value Function Approximation and Model Predic-tive Control. 2013 IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning (ADPRL), Sin-gapore, 16-19 April 2013, 100-107. [Google Scholar] [CrossRef] 
[2] 孙浩, 陈泽宇, 吴思凡, 潘峰. 基于策略梯度的智能车辆模型预测运动控制算法[J]. 汽车技术, 2021(12): 10-15. [Google Scholar] [CrossRef] 
[3] Rokonuzzaman, M., Mohajer, N., Nahavandi, S. and Mohamed, S. (2020) Learning-Based Model Predictive Control for Path Tracking Control of Autonomous Vehicle. 2020 IEEE International Conference on Systems, Man, and Cybernetics (SMC), Toronto, 11-14 October 2020, 2913-2918. [Google Scholar] [CrossRef] 
[4] Bristow, D.A., Tharayil, M. and Alleyne, A.G. (2006) A Survey of Iterative Learning Control. IEEE Control Systems, 26, 96-114. [Google Scholar] [CrossRef] 
[5] Wang, Y., Gao, F. and Doyle, F.J. (2009) Survey on Iterative Learning Control, Repetitive Control, and Run-to-Run Control. Journal of Process Control, 19, 1589-1600. [Google Scholar] [CrossRef] 
[6] Rosolia, U. and Borrelli, F. (2018) Learning Model Predictive Control for Iterative Tasks. A Data-Driven Control Framework. IEEE Transactions on Automatic Control, 63, 1883-1896. . [Google Scholar] [CrossRef] 
[7] Ioffe, S. and Szegedy, C. (2015) Batch Normalization: Acceler-ating Deep Network Training by Reducing Internal Covariate Shift. International Conference on International Confer-ence on Machine Learning, Lille, 6-11 July 2015, 448-456.
[8] Salimans, T., Ho, J., Chen, X., Sidor, S. and Sutskever, I. (2017) Evolution Strategies as a Scalable Alternative to Reinforcement Learning.
https://arxiv.org/abs/1703.03864
[9] Chrabaszcz, P., Loshchilov, I. and Hutter, F. (2018) Back to Basics: Benchmarking Canonical Evolution Strategies for Playing Atari. Proceedings of the 27th International Joint Conference on Artificial Intelligence, Stockholm, 13-19 July 2018, 1419-1426. [Google Scholar] [CrossRef] 
[10] Tamar, A., Thomas, G., Zhang, T., Levine, S. and Abbeel, P. (2017) Learning from the Hindsight Plan—Episodic MPC Improvement. 2017 IEEE International Conference on Robotics and Automation (ICRA), Singapore, 29 May-3 June 2017, 336-343. [Google Scholar] [CrossRef] 
[11] 邹凯, 蔡英凤, 陈龙, 等. 基于增量线性模型预测控制的无人车轨迹跟踪方法[J]. 汽车技术, 2019(10): 1-7.
[12] Mnih, V., Kavukcuoglu, K., Silver, D., et al. (2013) Playing Atari with Deep Reinforcement Learning.
https://arxiv.org/abs/1312.5602
[13] Nguyen, H. and La, H. (2019) Review of Deep Reinforcement Learning for Robot Manipulation. 2019 3rd IEEE International Conference on Robotic Computing (IRC), Naples, 25-27 February 2019, 590-595. [Google Scholar] [CrossRef] 
[14] Mülling, K., Kober, J., Kroemer, O. and Peters, J. (2013) Learning to Select and Generalize Striking Movements in Robot Table Tennis. The 80 International Journal of Robotics Research, 32, 263-279. [Google Scholar] [CrossRef] 
[15] Dan, S. (2013) Evolutionary Optimization Algorithms: Biologi-cally Inspired and Population-Based Approaches to Computer Intelligence. Wiley & Sons, Inc., Hoboken.