The Research Of Fuzzy Neural Network PID Controller Based On Knowledge

Due to the complication of modern industrial process, and the increase of nonlinearity, uncertainty and complexity of the practical production processes, the conventional PID controller can no longer meet our requirement. In recent years some advanced intelligent control methods have been applied in the PID control field.In order to solve the problem that conventional PID controller itself cannot adjust parameters online, this paper presents a compound control strategy which uses both the fuzzy neural network control and PID control, and conducts a thorough research on a PID controller based on Takagi-Sugeno (T-S) fuzzy neural network. Combining the advantages of neural network and fuzzy logic, the fuzzy reasoning is realized through the use of neural network.One of the key problems in the establishment of neutral network model is the increase of the input dimension of neural network, which results in an exponential increase of fuzzy rule number. This exponential increase leads to an outsize of FNN scale, which causes a dimension disaster. To deal with the problem, we employ Rough set theory; attribute reduction based on Rough set, and the classification of information system to find the minimal set of decision rules. The fuzzy neural network is characterized with the advantages of neural network and fuzzy reasoning.This paper proposed a designing method of fuzzy neural network PID controller based on rough set. The method combines the rough set and fuzzy neural network, which can derive control rules from input-output data effectively and find the minimal set of rules. Not only can this method solve the"rule explosion"problem and express qualitative fuzzy knowledge easily, but it also has preferable self-study ability. Results obtained from a simulation using...