神经网络模糊控制器单片机实现

为了实现工业控制过程的智能化，引入智能控制算法，在改善算法与软件仿真基础上，采用单片机控制技术，将模糊控制的规则集存储在神经连接权上，存入单片机的ＲＯＭ中，经过神经网络前向计算就可提取模糊规则，实现神经网络模糊控制器用于实际对象实时控制，详细给出模糊神经控制器的单片机实现方法和设计过程，实验结果表明，它是一性能优良实时控制器。

Microprocessor Realization of Fuzzy Neural Controller

In order to realize the intelligence of industry control process,the algorithm of intelligent function is introduced.Based on the inprovement of algorithm,the technique of microprocessor control is adopted to save the rule set of fuzzy control on the connection weght of neural network ,which is then saved in the ROM of microprocessor.The real time control is realized by the study of feedforward network. An optimized fuzzy neural controller for the automatic constant temperature control system is implemented with an 8098 microprocessor. The real time performance of the controller proposed is studied through experiment.