Based On Rcm's Hydroelectric Generating Set Condition Based Maintenance And Fault Diagnosis System

Having consulted the domestic and international documents, summarizing the current technological development in the condition-based maintenance of the hydraulic turbine-generator sets ,analyses its present level and trend, According to the fault diagnosis application situation of hydraulic turbine-generator set, It introduces the mix nerve network principle In detail, And carries on the chaos algorithm compared with the BP algorithm , In view of BP slow convergence, the convergence process oscillation seriously, hybrid neural network improves its learning algorithm, namely the right to amend the value, thus easing the above two the seriousness of the problem, Meanwhile details overview of the condition-based maintenance, overhaul the application of condition-based maintenance guidelines, and then establishes a state model, according to the state analysis model, builds RCM condition-based maintenance strategy model and overhaul RCM structure model—Fault Tree Analysis, according to the hydroelectric power plant actual situation ,we formulate the fault tree model of the hydraulic turbine-generator set , unifies the power plant the statistical data to analyze the probability as well as the opposite vertex event which various bottoms event occurred has the probability influence, qualitative analysis the important degree of the various bottoms event in entire fault tree, provides a decision-making for the hydroelectric power plant to make the Condition-based maintenance. And the mixed chaotic neural theory combining with reliability theory each other, utilizes in the fault diagnosis process and increases the rate of fault diagnosis.