Research On Maintenance Decision Making Of Nuclear Power Plant Based On Multi-objective Optimization

After the Fukushima nuclear accident,public opinion have more requirements to the safety of nuclear power plants and the requirements about maintenance are also paid more attention.At the same time,on one hand maintenance costs have been accounted for life cycle management about 30% and economic efficiency affects nuclear powers' competitiveness compared to other energy sources.On the other hand,there exists a severe problem of excessive maintenance in nuclear plants,resulting in unnecessary waste of maintenance costs.Therefore,how to make the plants safer and the cost lower is of great significance to the development of nuclear powers' maintenance.In this thesis,a basic strategy of maintenance optimization was proposed by studying disciplines and basic theories of maintenance optimization,as well related researches and practices in maintenance optimization.Based on the strategy,the framework of maintenance decision model of nuclear power plant system based on multi-objective optimization was established and multi-objective functions was demonstrated as one of important tasks by analyzing it.Then the thesis brought forward a method to optimize the maintenance model and gave an example.First of all,this thesis summarized academic research and practical results,evaluated the trend,applications,and effectiveness of academic research.Secondly,maintenance and reliability theory were introduced and analyzed.Necessary hypotheses about preventive maintenance in nuclear power was made to construct nuclear power preventive maintenance model.This concept was used to deduce the effective age before and after preventive maintenance.Then the failure rate and reliability in each maintenance node can be obtained.And preventive replacement was introduced in the maintenance process to deduce the function of reliability parameters with time during the operation period.Furthermore,this theses constructed the multi-objective incomplete preventive maintenance optimization model at component level which include reliability,availability and cost.On the basis of single device maintenance optimization,considering the influence of the connection among devices in a system on the system reliability,rose the component level up to system level.And constructed the multi-objective incomplete preventive maintenance optimization model at system level which include system reliability and system total cost.Finally,a program called MDO(maintenance decision optimization)was made include the models at component level and system level for decision making.And taking the water supply system of nuclear power plant as an example,the whole maintenance optimization process was verified.Moreover,optimize the multi-objective to get the Pareto optimal solution set.Then a selection criteria was given to decider to get the ideal maintenance scheme and gave a guidance for preventive maintenance of nuclear power plants.