Research On Maintenance Decision Of Pumped Storage Unit Based On Reliability

At present,pumped storage units equipment maintenance mainly adopts Time Based Maintenance mode.The maintenance period is mainly based on the national standard,industry standard and enterprise operation and maintenance experience,so it is easy to lead to over-maintenance and waste of resources.Condition Based Maintenance mode is in the pilot implementation of the pumped storage industry.The equipment online monitoring system only monitors the real-time status and can't predict the fault,thus can't give advice on the timing of the maintenance.In this thesis,some problems in the equipment maintenance decision of the pumped storage unit are studied.For the subsystem and parts pumped storage units equipment,it mainly solve problems about equipment importance assessment,how to choose the maintenance method and how to determine the maintenance timing.In order to achieve the goal of improving equipment reliability and reducing equipment maintenance costs by optimizing maintenance decisions.In this paper,the qualitative analysis method,the Analytic Hierarchy Process and the Monte Carlo simulation method are used respectively to evaluate the importance of the energy storage unit equipment.Among them,the Analytic Hierarchy Process is consistent with the Monte Carlo simulation method,butthe results of Monte Carlo simulation are more objective and reliable.Pumped storage unit equipment maintenance methods are mainly divided into four types:Condition Based Maintenance,Time Based Maintenance,Hidden danger detection and After-the-fact Maintenance.By means of RCM analysis,decision tree analysis in which equipment are classified according to importance degree,and fuzzy comprehensive evaluation method,maintenance mode decision are made respectively.The results of RCM method analysis are greatly affected by the level of professional skill workers.The decision tree method is simple to perform and is directly related to device importance classification.The fuzzy comprehensive evaluation method is also subjective,but it is easy to implement by computer.For the equipment that implements the time based maintenance mode,the calculation method of the optimal maintenance period is given for the reliability,availability and economy as the decision-making objective under the perfect maintenance mode and the assumed device reliability distribution function.The case study shows the impact of the reliability function and decision objectives on the decision results.For the equipment of the condition based maintenance mode,the variable weight fuzzy comprehensive evaluationmethod is adopted to evaluate the equipment status accurately.The operating equipment operation experience judgment and the prediction of residual working life of components based on BP neural network are used to determine the timing of condition based maintenance.Both methods are suitable for short-term decision making and have certain practical value.