BP Neural Network Evaluation Model Combined with Grey - Fuzzy Based on Included Angle Cosine

This paper is aimed to establish a novel evaluation model. It has verified the validity of the model based on the application of thermal power project. First of all, this paper analyzed the theoretical and practical significance of thermal power project. Secondly, it built a specific evaluation system from multi-angle and multi-level. The index system is a great innovation of this article. Thirdly, this paper established a model combining BP neural network with grey-fuzzy evaluation model. Finally, this paper selected a 2×630 MW air cooling thermal power unit in China. Then do the empirical research based on the data of 2010 to 2012 using Matlab. The combined evaluation model can avoid the subjectivity of grey-fuzzy model and the instability of BP neural network model. It has carried on the empirical research of comprehensive benefits, concluded the level. Copyright © 2014 IFSA Publishing, S. L.