Evaluation System For Nuclear Accident Based On Fuzzy Neural Network

We want to be first time to analyze the dose received by people around theNuclear power plant and make the emergency action plan,so that it make the lossless.This paper mainly introduce the process,functional design and key technologiesabout the evaluation system for nuclear accident.Based on the J2EE and GeographicInformation System.Compare to other evaluation system for nuclear accident which make Emergencyplan only by estimating the value to exceed the limen, This paper adopted the expertevaluation system which is more intelligent fuzzy neural network structure to giverecommendations for action. According to the observation data of input or accordingto the Operating condition work out the source term and the value to compare thelimen, we could take the numerical values as the input of the expert evaluationsystem, then we make them fuzzy by puting them into the neural network. Adoptthree floor ahead BP neural network structure, use BP arithmetic training networkparameter,then give recommendations for action.This paper firstly introduces the origin of OIL and intervention measures of thecorresponding various OIL levels. In this paper, we take the new operationalintervention levels as the judging standard,and bring the distance and meteorologicalfactors which could influence the decision in my system. Then introduced the fuzzyneural network,it is picked up through self-study trait of neural network. Moreover,Fuzzy system can also make up the shortcoming of slow operational rate of the neuralnetwork. So fuzzy neural network in the evaluation system will have lowermisinformation rate,higher security and stronger environmental adaptive capability.The thesis introduces the construction of fuzzy neural network for the evaluation system for nuclear accident,expounds the design course and the arithmetic. We havetrained the fuzzy neural network,and got good results. It is proved the thought thatbringing the fuzzy neural network in emergency decision is correct.