随着电动加载系统的不断发展,对控制精度、动态特性和稳定性提出了更高的要求,常规的小脑模型(CMAC)和PD控制相结合的复合控制策略难以满足加载指标要求。针对无人机舵机电动加载系统的控制需求,提出了一种基于平衡学习、最优权值和自适应学习率的新型小脑模型(BOWA-CMAC)复合控制策略,它在保留小脑模型算法正常学习过程的同时,避免了算法的过学习现象,保证了系统的稳定,同时提高了跟踪精度和动态特性。仿真和实验结果表明,BOWA-CMAC复合控制策略具有很强的鲁棒性,抑制了加载系统的多余力矩,保证了系统的稳定性,有效提高了系统的跟踪精度和动态特性,非常适合于实时控制。
The rapid development of electric loading systems brings out higher demands for control precision, dynamic char-acteristics and stability, which makes it diffficult for compound control strategy with conventional cerebellar model articulation controller (CMAC) and PD to meet the loading requirements. Therefore, on the basis of an electric loading system of unmanned aerial vehicle, a novel hybrid control strategy CMAC is proposed with improvements of the balance learning method, optimal weight and adaptive leaning rate (BOWA-CMAC). It not only retains the normal learning process of the hybrid controller, avoids the excessive self-learning phenomenon and assures the stability, but also improves the tracking precision and dynamic characteristics. The simulation and experimental results demonstrate that the proposed controller has good robustness and can effectively eliminate the surplus torque, assure the stability of the system with high tracking precision and dynamic characteristics in real time control.
[1] Li C G, Jin H T, Jiao Z X. Mechanism and suppression of extraneous torque of motor driver load simulator. Journal of Beijing University of Aeronautics and Astronautics, 2006, 32(2): 204-207. (in Chinese) 李成功, 靳红涛, 焦宗夏. 电动负载模拟器多余力矩产生机理及抑制. 北京航空航天大学学报, 2006, 32(2): 204-207.
[2] Chen K, Huang Y, Sun L. The research of linear rudder square direction electric loading system. Journal of Astronautics, 2008, 29(5): 1515-1520. (in Chinese) 陈康, 黄勇, 孙力. 电动直线舵机方波加载系统研究. 宇航学报, 2008, 29(5): 1515-1520.
[3] Mu X Y, Pei R, Liu Z L, et al. A control system for the electro-hydraulic load-simulator employed in a marine rudder. Control Theory & Applications, 2008, 25(23): 564-568. (in Chinese) 慕香永, 裴润, 刘志林, 等. 用于船舶多级的电液负载模拟器之控制系统. 控制理论与应用, 2008, 25(23): 564-568.
[4] Fu W X, Sun L, Yu Y F, et al. Improving design of control system for DC motor-driven torque control simulator. Journal of Northwestern Polytechnical University, 2008, 26(5): 621-625. (in Chinese) 符文星, 孙力, 于云峰, 等.电动负载模拟器控制系统设计.西北工业大学学报, 2008, 26(5): 621-625.
[5] Qiu H G. The electric steering gear dynamic loading control method based on BP neural network. Xi’an:School of Mechanical-Electronic Engineering, Northwestern Polytechnical University, 2007. (in Chinese) 邱红岗. 基于BP神经网络的电动舵机动态加载控制方法研究. 西安: 西北工业大学机械电子工程学院, 2007.
[6] Shen D K,Hua Q, Wang Z L. Motor driven load system based on neural networks. Acta Aeronoutica et Astronautica Sinica, 2003, 26(3): 525-529. (in Chinese) 沈东凯, 华清, 王占林. 基于神经网络的电动加载系统. 航空学报, 2003, 26(3): 525-529.
[7] Hsu C F, Chung C M, Lin C M, et al. Adaptive CMAC neural control of chaotic systems with a PI-type learning algorithm.Expert Systems with Applications, 2009, 36(9): 11836-11843.
[8] Tsai C H, Yeh M F. Application of CMAC neural network to the control of induction motor drives. Applied Soft Computing, 2009, 9(4): 1187-1196.
[9] Telmltas H. A nonlinear load simulator for robot manipulators. Proceedings of the Annual Conference of the IEEE on Industrial Electronics Society. 2001: 357-362.
[10] Cheng Q C, Wang H Y, Shi L M. Improved fuzzy CMAC neural network. Computer Engineering and Applications, 2009, 45(23): 182-185. (in Chinese) 程起才, 王洪元, 施连敏. 改进的模糊CMAC神经网络. 计算机工程与应用, 2009, 45(23): 182-185.
[11] Liu S R, Zhou G C, Wu Q X. Fuzzy CMAC compensating control for high precision servo system. Control Engineering of China, 2010, 17(6): 836-839. (in Chinese) 刘士荣, 周国成, 吴秋轩. 高精度伺服系统的模糊CMAC补偿控制. 控制工程, 2010, 17(6): 836-839.
[12] Wang Y, Hu S S, Qi J W. Self-organizing fuzzy CMAC neural network and its nonlinear system identification. Acta Aeronoutica et Astronautica Sinica, 2001, 22(6): 556-558. (in Chinese) 王源, 胡寿松, 齐俊伟. 自组织模糊CMAC神经网络及其非线性系统辨识. 航空学报, 2001, 22(6): 556-558.
[13] Cheng H, Qin T, Chen Z H. Simulation of consecutive CMAC-RLS in CSTR system. Computer Simulation, 2008, 25(10): 102-104. (in Chinese) 程辉, 秦廷, 陈宗海. 连续CMAC-RLS在CSTR系统中的仿真研究. 计算机仿真, 2008, 25(10): 102-104.
[14] Jiang Z M, Lin T Q. A new self-learning controller based on CMAC neural network. Acta Automatica Sinica, 2000, 26(4): 542-545. (in Chinese) 蒋志明, 林廷圻. 一种基于CMAC的自学习控制器. 自动化学报, 2000, 26(4): 542-545.
[15] He C, Xu L X, Zhang Y H. Convergence and generalization ability of CMAC. Control and Decision, 2001, 9(5): 523-529. (in Chinese) 何超, 徐立新, 张宇河. CMAC算法收敛性分析及泛化能力研究. 控制与决策, 2001, 9(5): 523-529.
[16] Peng Y F, Lin C M. Adaptive recurrent cerebellar model articulation controller for linear ultrasonic motor with optimal learning rates. Neurocomputing, 2007, 70(16-18): 2626-2637.
[17] Yu W, Moreno-Armendariz M A, Rodriguez F O. Stable adaptive compensation with fuzzy CMAC for an overhead crane. Information Sciences, 2011, 181(21): 4895-4907.
[18] Chen F C, Chang C H. Practical stability issues in CMAC neural network control systems. IEEE Transactions on Control System Technology, 1996, 4(1): 86-91.
[19] Yeh M F, Tsai C H. Standalone CMAC control system with online learning ability. IEEE Transactions on Control System Technology, 2010, 4(1): 43-52.
[20] Zhu D Q, Zhang W. Nonlinear identification algorithm of the improved CMAC based on balanced learning. Control and Decision, 2004, 19(2): 1425-1428. (in Chinese) 朱大奇, 张伟. 基于平衡学习的CMAC神经网络非线性辨识算法. 控制与决策, 2004, 19(2): 1425-1428.
[21] Yang B, Wang J K. Hybrid control based on improved CMAC for motor-driven loading system. Acta Aeronautica et Astronautica Sinica, 2008, 29(5): 1314-1318. (in Chinese) 杨波, 王俊奎. 基于改进的CMAC的电动加载系统复合控制. 航空学报, 2008, 29(5): 1314-1318.
[22] Yeh M F, Lu H C. On-line learning CMAC control system. 12th International Conference on Intelligent Engineering System, 2008: 115-120.
[23] Yang B, Wang Z. Adaptive CMAC hybrid control for rudder electric loading systems. Journal of Beijing University of Aeronautics and Astronautics, 2010, 36(3): 333-337. (in Chinese) 杨波, 王哲. 舵面电动加载系统的自适应CMAC复合控制. 北京航空航天大学学报, 2010, 36(3): 333-337.
CJA