Security Evaluation of Thermal Power Plant Based on Grey Clustering and Support Vector Machine Method

As to security evaluation problem of thermal power plant, a kind of security evaluation method based on sample classification is proposed. Feature space that represents security of thermal power plant is presented, and feature vector of security evaluation is obtained by Delphi Method. Feature vector is clustered by using of improved grey clustering method of triangle whitenizated weight function, and sample data of security evaluation of thermal power plant is obtained. Then sample data is trained by using of Support Vector Machine (SVM) to obtain security judgment function, and pattern recognition of security degree of thermal power plant is realized. It can be seen from the results that this method can classify security degree of thermal power plant effectively.