Reliability estimation of neural networks with human factors under emergency of nuclear power plant

Recently, people do not pay attention to the events arised from human factors until human factors events occur more and more frequent. In this paper, the authors take nuclear power plant as reference background and propose a particular neural network model for emergency with human factors. The model comprises of three assessment parts. For each part, the authors come up with specific function in order to assess human factors error probability of emergency. The proposed methods are tested by experiments. From results of experiments, we can easily see that human factors error probability is precise and reliable in the model, that the proposed method is more accurate than a single weibull function, that which active function is better, and that which tolerance is minimum. The method can be applied to human factors reliability analysis in emergency of nuclear power plant and has a great significance for safety accidents in nuclear power plant.