一种模糊控制算法的改进  

Analyzing and improving a fuzzy-control method on the basis of probability theory

作　　者：赵红[1] 邵萍波[1] 谭跃[1] 

机构地区：[1]大连海事大学自动化与电气工程学院,辽宁大连116026

出　　处：《大连海事大学学报》2004年第4期70-73,共4页Journal of Dalian Maritime University

摘　　要：由于目前实际中的模糊控制器通常多采用Mamdani合成模糊推理并结合重心法进行清晰化,此类控制器运算量很大,相应的实用化产品对硬件和软件都有较高的要求,导致其适用性受到一些限制.为此采用概率论中"最佳预测"的概念推导出一种模糊控制算法,并考虑到适用性的需要对该模糊控制算法进行了改进.改进后的算法十分简单,并用当前流行的仿真软件Matlab对其进行了仿真分析.分析结果表明,该算法与普通模糊控制器及文献中介绍的快速智能PI控制器相比,其上升时间和稳态时间均大大减小,具有响应迅速、超调小的特点.Most Fuzzy-controllers generally adopt the Mamdandi-method coupled with the Gravity-method to clarify object. Such controllers require huge works on calculating and a high level of hardware and software to be used in their products, so there are some limitations on their applicability. Considering the applicability, this paper deduces and improves a fuzzy method on the basis of probability. The improved method is very simple. Simulation software Matlab is adopted here to analyze this method. The result indicates that its raising time and stable time significantly decrease compared with normal fuzzy-controller and intelligent PI controller, which is introduced in the reference.

关 键 词：概率论 模糊控制 仿真软件 计算机仿真 

分 类 号：TP273.4[自动化与计算机技术—检测技术与自动化装置]