Plant Management Theory And Method Based On Internet And Condition-monitoring

As the only carrier of production, plant has become the key for the globalizationof enterprise under the new economic circumstances. Applying computers, modernnetwork communicating technology and advanced diagnosis & monitoringtechnology to plant management, especially carrying out dynamic management andaided decision for the whole life of the plant in the respects of economy andtechnology may bring enterprises great economic and social benefits. Thisdissertation focuses on the plant management theory and method based on theinternet and condition monitoring for power plant.On the basis of plant management theory in different period (including theLogistics, Terotechnology and Total Productive Maintenance), the theoretical systemof plant management engineering is proposed in this dissertation. And at the sametime, the total solution of plant management in enterprise is also established. Thesolution is founded on Reliability Center Maintenance policy and combined theInternet and condition monitoring technology, thereby providing the developmentdirection of plant management engineering in enterprise equipments maintenance,and laying a foundation for the plant-integrate management.The theory and method of fault diagnosis for power plant is discussed in thisdissertation. Empirical Mode Decomposition (EMD) for the non-stationary signalsanalysis, Step-changed Stochastic Resonance (SR) for the detection of weak signal instrong noise, and the Singular Value Decomposition (SVD) for fault featureextraction and signal distinguish of mechanical equipment are demonstrated. Theintelligent diagnosis method based on Radial Basis Function Neural Network (RBF)is studied. In order to take full advantage of the parallel multiply and addinstructions, circular addressing mode and repeating mode operations in DSP, themaster/slaver intelligent diagnosis model that combined PC and DSP is proposed toimprove the implementing efficiency and satisfy the real-time requirement ofintelligent diagnosis.Aiming at the different fault ratio and essentiality of power plant, themaintenance policy that combined route inspection management and on-linemonitoring is proposed in this dissertation. The reasonable Fault Diagnosis Report(FDR) form is presented to eliminate the "information island " between the plantmanagement and condition monitoring.Based on the demand of power plant management, a scientific reasonableworkflow model for the plant management is established. The information share andcommunication between these function modules and the maintain work mechanismof power plant are also discussed in the dissertation.Finally, the precept, the system truss and prototype of PMIS are proposed, andthe integration technologies between different modules and different systems arediscussed. A perfect PMIS is built and applied in the Dagang Power Plant.