二级倒立摆的动态加权模糊控制

针对快速、多变量、强非线性、绝对不稳定特性的二级倒立摆系统,运用基于目标的动态加权模糊控制器对其进行控制,通过模糊推理得到加权系数动态连接各个子目标模糊控制器,使多个子目标模糊控制器在控制过程中既相互竞争又相互协调。解决了一般模糊控制在控制多变量系统时出现的规则组合爆炸问题,大大减少了模糊控制器的规则数,有利于利用专家经验。利用Matlab提供的良好的人机交互界面和功能完备的模糊逻辑设计工具,结合Simulink开放式的、动态的图形化仿真环境,使复杂的模糊控制系统的仿真过程大大简化,缩短了开发周期。实验结果表明该方案控制效果好,鲁棒性强。

Dynamic Weighting Fuzzy Control for Double Inverted Pendulum

An objective-based dynamic weighting fuzzy controller is presented for the double inverted pendulum system which has the characteristics of speediness, multiple variables, serious nonlinear and absolute instability. Sub-fuzzy controllers which were dynamicly joint by weighting factor were designed for each control objective by fuzzy inference, all sub-fuzzy controllers compete and cooperate with each other during controlling process. This method solves the problem of rule number's explosion which occurs when ordinary fuzzy controller was used to control multiple variables system, reduces the number of rules of fuzzy controller effectively and can take better advantage of the experts' experiences. Using Matlab which has good man-computer communication interfaces and fuzzy logic design toolboxes with complete function, combining with opening dynamic graphical simulation condition of Simulink, the simulation progress of complicated fuzzy control system was simplified and the developing period was reduced. The result shows that this method has good control effect and robustness properties.