滞后复杂系统的多步预报自学习模糊控制  

The Multi-step Predictive Self-learning Fuzzy Control for the Complex System with Time-delay

作　　者：何洪[1] 周恩涛[1] 孙薇[1] 周士昌[1] 

机构地区：[1]东北大学流体传动及控制系

出　　处：《控制与决策》1996年第3期300-303,299,共4页Control and Decision

基　　金：国家教委博士点基金

摘　　要：针对具有纯时滞及非线性的复杂系统，提出一种基于多步预报的自学习模糊控制算法，使控制效果基本不依赖于初始控制表，而控制表可以在线修正，以满足系统快速性和稳态特性的要求。仿真结果令人满意。For the complex system with time-delay, a new type of self-learning fuzzy control algorithm based on the multi-step prediction is proposed in this paper. The controlling effection does not depend on the initial fuzzy controlling table, the controlling table can be adjusted on line. Different learning rules are applied at different states to meet the different requirement on responding speed and the character near steady state. The simulation results are satisfactory.

关 键 词：模糊控制 自适应控制 复杂系统 自学习算法 

分 类 号：TP273.2[自动化与计算机技术—检测技术与自动化装置]