| The middle storage pulverizing system for ChongQing power plant is mainly researched in this paper, and BP neural network,mixed knowledge representation,expert system based on knowledge technology etc are used in this fault diagnostic system. The practice samples are dressed , and the knowledge database of the fault diagnostic is developed , then the fault samples with the signal of zero or one and the model of BP neural network ,fault diagnosis and explaining module,data querying and analyzing, module etc are established . Meanwhile this system is programmed with DELPHI6.0 and the fault simulation tests of the pulverizing system and its concerned boiler fire extinguish is practiced. The fault samples with the signal of zero or one and the model of BP neural network are established, in accordance with nine faults of pulverizing system for ChongQing power plant. During the course of the samples compilation, produce rule ,neural network representation etc are synthesized to organize and express the fault sample of pulverizing system. Before being putted into the database, the sample data is disposed to ensure sample data integrity and no redundancy.The influence to the training circle times with number of hiding layer node,learning rate etc are discussed during the course of samples training. By comprehensive analyses and comparing , the comparatively rational value is adopted to be network's parameters. The result of fault diagnosis simulation tests indicates that the fault diagnosis system could make training error reach to the aim value quickly and efficaciously for all pulverizing system fault samples. At the same time, simulation tests prove that the fault samples with the signal of zero or one of pulverizing system and BP neural network model are correct, and this system faults can be diagnosed exactly and quickly. Obviously, this research is successful and lay the foundation for the development of pulverizing fault diagnostic system.
|
PDF Full Text Request