RCM Based Condition Maintenance System Study For The Blower Fan Of Power Plant

As the machine is more and more bigger and complex, machine maintenance becomes more and more complex and important. Demand for machine maintenance boost maintenance technology increasingly perfect. Maintenance mode have passed through five primary phases: Break-down Maintenance(BM), Time-based Maintenance(TBM), Condition-based Maintenance(CBM), Corrective Maintenance (CM), Optimization Maintenance(OM). RCM , based on Fault Mode ,Effect and Criticality Analysis(FMECA), is well-developed one of OM , which combined BM> TBM, CBM and CM ,and then choose the best maintenance mode for special part and component.CBM would be the intending mainstream in the field of machine maintenance, which aim at prolonging MTBF(mean time before failure) , cutting down maintenance items and reducing maintenance costs . CBM is such method using multi-technique, mastering the equipment's running information, forecasting the occurrence and development of failure, analyzing techno-economy, adopt the best maintenance mode and its interval for the running condition.In this paper, the concept of RCM has been applied to process of machine maintenance for Fan Browser of power plant. Then introduced the system of CBM and its subsystem: state inspection, signal processing and fault diagnosis. First, we describe the overall configuration of the system of state inspection which adopt the most advanced Virtual Instrument(VI) technology.Then the model of signal processing analyzed the original signal from three different filed: time field and frequency field ,which can provide us lots of fault symptoms from different aspects. The system of fault diagnosis is a multi-layers model , its basic idea is : in order to avoid excessively complex because of too many inputs of the system, we sorted fault symptoms into primary fault symptom ,minor one and more minor one, and got the diagnosis result step by step.In the end, we computerize the system of BCM by use of LabVIEW. In the process, we transferred MATLAB to make use of its strong calculate ability and its function of neural network toolbox.