神经网络自学习模糊控制及其在合成氨生产中的应用

Self-learning Neural Network Fuzzy Control and Application in the Synthetic Ammonia Production

提出一类基于神经网络的模糊控制设计方案。控制系统中包括两个神经网络，一是利用神经网络进行模糊推理，实现控制规则的推理过程；二是采用另一个神经网络对系统的动态进行跟踪，以实现前向通道中语言变量的模糊区间的优化，从而使控制效果更加理想。

A fuzzy control scheme based on the neural networks(NN) structure is presented. The control scheme includes two NNs. One NN realizes the fuzzy inference, and another optimizes the fuzzy input scalars so that the system dynamic can be tracked. The system performance can be improved by using the scheme. The example of the synthetic ammonia shows the effectiveness of the control scheme.