一种神经元控制模型参考自适应伺服系统  

Model Reference Adaptive Servo System Based on Neuron

作　　者：冯国楠[1] 张力[1] 金万煌 孙亮[1] 

机构地区：[1]北京工业大学自动化系,100022

出　　处：《控制与决策》1997年第4期312-316,共5页Control and Decision

摘　　要：针对存在于摩擦非线性和负载转动惯量时变的伺服系统,提出一种神经元模型参考自适应控制设计方法。导出了控制对象离散数学模型,介绍了伺服系统结构和控制器设计,并用波波夫超稳定性理论证明该系统是渐近超稳定的。仿真结果表明系统具有良好的动态性能、稳态精度和鲁棒性。A method of designing neuron based DC servo control system with model reference adaptive con-trol scheme is described, which overcomes the bad effects of dry friction nonlinearity and time variable inertia load. The discrete time mathematical model of the controlled plant is derived. The structure of the servo con-trol system and the design method of the controller are given. The servo control system is proved to be asymptotically hyperstable by means of Popov hyperstability theory. The simulation results show that the dy-namic performance, stationary precision and the robustness of the servo control system are all good.

关 键 词：神经元 控制模型 自适应控制 超稳定 伺服系统 

分 类 号：TP273.2[自动化与计算机技术—检测技术与自动化装置] TM921.541[自动化与计算机技术—控制科学与工程]