Research Of The Expert System For The Milling System Fault Diagnosis

With the modern power installed capacity more and more big, the automatization, complexity and scale of the system increasing rapidly , the investment is growing. Facing the huge investment complexity system, the reliability, maintainability and effectiveness is become more and more important. In modern coal boiler, pulverizing system and boiler combustion equipment composed the whole important parts of the combustion system. In actual production process, it is especially important to monitor and analysis the coal pulverizing system fault timely.We focused on the certain fuzziness between the powder system fault symptoms and reason to develope a system which is suitable for the difficulty establishing mathematical model of fuzzy expert system. Fuzzy expert system is one of the most successful branch of artificial intelligence. It can use a lot of expert knowledge to reasoning, simulation the experts of decision-making ability, and specializes in solving difficult mathematical model of the system. The biggest advantage is that he can rely on the experience of human expert knowledge, incomplete or uncertain information in the case can still give a reasonable conclusion.According to the milling systems running characteristic and the fault characteristic, this article proposes a general framework of the milling system fault diagnosis system, develop the fuzzy expert system of the running fault diagnosis which is fit for the complex industrial production process and the good man-machine contact surface.The paper's main functions include: milling system fault diagnosis knowledge acquisition and representation; the establishment of the fact libraries; learn from experts in the field relevant information and experience, prepare 11 kinds of signs to strike a sign of the function and the credibility of the membership function; Reasoning mechanism was established to give the exact cause of the malfunction and the appropriate guidance; using VC + + 6.0 and the CLIPS of mixed Programming design the good man-machine interface .