| In Power Plants, air feeder system plays a very important role as one of the assistant systems. One hand, it has direct relationship with the combustion of the boiler. On the other hand, its high effciency run can improve the utilization of coal. Therefore the foundation of air feeder Condition-Based Maintenance has economical and secure practical sense for Power Plants. And at the same time, the tradition maintenance system, which is Time-Based Maintenance for the main part, has gradually been replaced by CBM, because it can't satisfy the need of modern Power Plants.This paper designs a CBM system based on ANN about the air feeder of Shizuishan Power Plant, this system can realize data collection, state monitoring, state forcast, trend analysis, fault diagnosis, fault analysis and maintenance decision-making of the air feeder.This thesis particularly expatiates the theory, frame and functional methods of air feeder CBM system. The system collects the main performance parameters of air feeder by the PI real-time database network, and has real-time communication with the PC according to the PI data interface for on-line state monitoring. Synchronously after data pretreatment and feature extraction , the collection data get across the state forecast model and fault diagnosis model based on ANN for the air feeder's current state, the results of the forecast and the diagnosis feedback to the on-line state monitoring after analysis and decision-making, to instruct the repairmen. The system has well reliability for the application of advanced technique of state monitoring, signal processing and fault forecast and diagnosis.The air feeder CBM system has a guarantee for its well running, contemporarily ensures the boiler's safe combustion, the secure and economical running of the Power Plant. It doesn't only reduce the maintenance charge, prolong the maintenance distant, but also extend the life-span of air feeder, increase the economic benefits of the Power Plants.
|
PDF Full Text Request