Study On Safety Evaluation Method Of Thermal Power Plant

With the rapid development of electric power industry, the safety of power generation was obviously prominent. The method of safety evaluation based on classification of sample data was proposed for thermal power plant.The feature space, which represented safety characteristics of thermal power plant, was put forward. The concrete evaluation contents of feature space were provided. The feature value to evaluate safety of thermal power plant was got by Delphi's method. Then improved grey clustering based on triangle whitenization function was used to cluster feature vectors. And sample data of safety evaluation were got.Afterward, sample data were trained according to two methods which were support vector machine and radial basis function network. The model parameters which are safety criterion function and radial basis function were obtained. Based on the two models, test data were examined. The results show that the method can effectively classify the safety level and provide a way of evaluating safety of thermal power plant.