Study On The Methodology Of Multi-objective Optimization And The Scheme Verification For Nuclear Power Plant

The pressurized water reactor(PWR)occupies the major share in marine nuclear power plants,because,compared with other types of reactors,the technique of PWR is more developed,its safety is more reliable and the operating performance is more agreeable.Keeping ship propulsive power,if the weight or size of the PWR could be reduced,it would make contribution to the sailor living condition and ship payload.Hence,it not only has theoretical significance,but also creates real value by optimizing the weight,volume and other designing targets of a PWR in its design phase on the premise of meeting the design requirements of related components and systems.In addition,because of the complexity of nuclear power system,the strong coupling of designing parameters and the non-linear relationships between different designing targets,single-objective optimization of a PWR may result in deteriorations of some designing focus when one target gets to the optimum,this would even threat to the overall performance of the nuclear power plant.Hence,it is necessary to conduct the research on the multi-objective method for the plant.Moreover,to examine the availability of the optimized scheme,it has to study its transient behaviors under design basic accidents.To solve above problems,in this thesis,the concept of "non-dominated solution" is introduced to provide a new method for the multi-objective optimization design of a PWR;and the RELAP5,which is the internationally-recognized best-estimated code for light water reactor,is utilized to study the transient behaviors of an optimized scheme under some representative design basic accidents.The main contents are as follows:1?Based on the hyper-mutation strategy of immune clone algorithm and the depth-search strategy of complex algorithm,the traditional non-dominated sorting genetic algorithm-II is modified on the purpose of improving its optimization performance.Through four sets of testing functions,the non-dominated solutions,obtained by the modified non-dominated sorting genetic algorithm-?,the non-dominated sorting genetic algorithm-?,the distance method and the immune memory clone constrained multi-objective optimization algorithm,are compared.The results indicate the modified algorithm outperforms the other three on the convergence,continuity,uniformity and solutions extensity.This algorithm could be employed to solve constraint multi-objective optimization problem in a more effective way.2.Based on the developed nuclear components mathematical models,and with the help of the modified algorithm,the design of the steam generator is optimized to reduce its weight and primary coolant flow-rate.The results manifest that the reactor inlet coolant temperature is a critical variable that influence the distribution of its non-dominated solutions;the weight and flow-rate could be at most reduced by 16.5%and 17.0%,respectively.The primary system is optimized for reducing its weight and volume;the results indicate that the reactor inlet coolant temperature influences the distribution of its non-dominated solutions;the weight and volume could be at most reduced by 9.8%and 11.4%,respectively.The whole system is optimized for reducing its weight and volume,the secondary pressure of the steam generator influences the distribution of its non-dominated solutions,the weight and volume could be at most reduced by 10.4%and 23.63%,respectively.3.The RELAP 5 models for the primary systems of the prototype PWR and one of its optimized schemes are established,respectively.Steady states are also obtained for both models.The simulation results from the RELAP5 model are compared with corresponding data of the prototype.The results indicate the relative error of the RELAP5 model is within 1.21%.4?Based on the steady states of the prototype and the optimized scheme,their transient behaviors are studied under loss of feed water accident,station blackout accident and small-break loss of coolant accident.The influences to the process of the accidents,caused by the variations of some operational parameters,are also analyzed.The results show that during these three accidents,the variations of main parameters in the optimized scheme have analogous tendencies with that in the prototype PWR,the reactor safety of the optimized scheme could be ensured in the studied accidents;during the loss of feed water and station blackout accidents,the flow-rate of auxiliary water per steam generator should be greater than 10.5kg/s;in the small-break accident,with 100mm break at the cold leg,the reactor core safety of the optimized scheme benefits from the quicker actions of reactor scram and high-pressure injection.