摘要
针对伺服系统二次型最优控制存在的问题,提出了基于模糊神经网络补偿的二次型最优控制方法,该控制方法利用模糊神经网络的实时学习能力,能够及时补偿被控对象建模不准确、参数摄动和外界干扰等非线性因素对控制系统性能的影响,增强控制系统的自适应能力,有效提高控制系统的跟踪性能和抗干扰鲁棒性能。仿真试验结果验证了该控制方法的有效性。
A novel quadratic optimal control method based on fuzzy neural networks compensation is proposed for the inherent problem of servo system quadratic optimal control. The real-time learning capabilities of fuzzy neural networks are used to compensate the effect result from imprecise models, parameter uncertainty and extern disturbance of the control system, enhance the adaptive capability of the whole control, and effectively improve the tracking performance and robust performance. Simulation results show the effectiveness and applicability of the proposed method.
出处
《控制工程》
CSCD
北大核心
2009年第3期264-267,共4页
Control Engineering of China
关键词
二次型最优控制
模糊神经网络
补偿
伺服系统
quadratic optimal control
fuzzy neural networks
compensation
servo system