多模型优化模糊控制算法的应用研究

【作者】 徐前锋；

【导师】 侯志林；

【作者基本信息】 西安理工大学 ， 控制理论与控制工程， 2000， 硕士

【摘要】 模糊控制器是一种仿人智能控制器，它模拟人类思维，采用人类语言信息，所以模糊控制器容易理解、设计简单、维护也方便。模糊控制器基于包含模糊信息的控制规则，所构成的控制系统比常规控制系统稳定性好，鲁棒性高。在改善系统特性时，模糊控制系统不必像常规控制系统那样只能调节参数，还可以通过改变控制规则、隶属函数、推理方法以及决策方法来修正系统特性。遗传算法是一种自适应、启发式、群体型、概率性、迭代式全局收敛算法，利用遗传算法的良好的搜索特性来优化模糊控制器，可以取得很好的控制效果。本文在此基础之上，通过研究众多工业过程在不同的操作条件下的系统结构，将其视为多模型对象（Multi－Model Object),同时对人的决策过程和对事物可信度的判断过程作出分析，提出了多模型优化模糊控制算法，算法的关键是模型匹配程度与控制权重的计算。 在本文中，作者还对与此相关的模糊逻辑系统、模糊辨识器、控制器和遗传算法及其收敛性作了分析。更多还原

【Abstract】 Fuzzy contrOller (FC) is a hod of intelligent controller, whichimitates human thoughts uses the information Of our language, so fuzzy controllercomprehended easily, design simply, maintenance conveniently. Because of fuzzycontroller based on control rules including fuzzy information, stability and robusticityof control systems, which consist of fuzzy controller, are better than those of generalcontroller. During improving system characteristics, fuzzy control systems may modifysystem characteristics by changing control rules, membership function reasoningmethods and decision-making methods, rather adjusting parameters like generalcontrol systems. Genetic algorithm (GA) is a kind of convergent algorithm with self-adaptation, elicitation, colony, probability, and iteration on whole range. OptimizingFC using GA ’s good search characteristics may obtain better control results. Thispaper brings forward Multi-model Fuzzy Optimization Control by researching a lotsof system’s construct of industry process with different operation which is regarded asMulti-model Object and making analysis with people’s decision, whose keys arecalculation of MOdel Matching Degree and Control Weight.In addtion, Author makes analysis to correlative fuzzy logic systems, fuzzycontroller,GA and its astringency in this paper.更多还原