专家能力的模糊神经网自组织控制器

A New Kind of Self-organization Fuzzy Controller Based on Neural Network with Expert Adjusting Ability

为实现酒精发酵变温控制，提出一种新的、具有专家调整策略的神经网络自组织模糊控制器．用神经网络代替传统模糊控制器的逻辑规则推理合成运算或查表，减少了工作量．采用一种对量化，比例因子进行专家调整的策略，使系统响应的上升时间减少，超调量降低，控制器的鲁棒性，稳定性均良好．

In order to realize the optimal control of temperature in the process ofalcohol ferment, a new kind of self-organization fuzzy controller based on neural networkwith expert adjusting ability is proposed.The system mathematics model is not necessary inthe new controller, which can reduce the influence of system paramenters.The quantizationfactors and the proportion factors are regulated automatically by expert system. Thereasoning method of traditional fuzzy controller is replaced by the neural network methodwhich can improve the real-time dynamic performance. The simulation experiment on thenew controller is carried out. The results show that the new controller has betterrobustness, stability and precision.