Design And Application Of Revising Fuzzy Control Based On On-line Inference

Because complex industry object has non-linearity, uncertainty and time-change, this paper puts forward revising fuzzy controlling arithmetic based on on-line inference to satisfy demand of computer process control system. Because fuzzy control do not need to know accurate mathematic model of object, but it is based on human thinking and producing experience; and because it applies language rule to describe controlling process and bases rule to modify the controlling arithmetic and parameter , at the same time, this mathematic has self-adaptability, it will realize computer process control.Fuzzy control rules are generated by a simple equation by way of on-line inference , on the basis of real time controlling. It avoid complicated reasoning. What's more, adjusting fuzzy control rules on line with the weighting modificatory gene α reflects different weight between deviation and its rate of change, which conforms to the nature of a control process. In order to improve arithmetic robustness, a method for optimizing the weighting modificatory gene a is adopted in the applied rule-adjustable fuzzy controller. The system can quickly achieve the optimization point via self-optimization online. Eventually, it is added in configuration software. Through DDE channels , it realize communication with configuration software and directly realize control of computer process control system. The result shows that revising fuzzy controlling arithmetic based on on-line inference have good quality and will have good developing foreground and be widely used in industrial control.