科研之友
微信
新浪微博
Facebook
分享链接
摘要
Wind power projects face an uncertain external environment, they are complex projects in themselves and the capabilities of the designers, erectors and operators are limited. All this makes the identification of investment risks for wind power projects extremely complicated. In this paper, we propose a method for identifying the investment risk scientifically and accurately using a back propagation (BP) neural network. Specifically, we propose a hybrid wavelet transform fuzzy BP neural network (WT-FBPNN) optimisation model based on the construction of a risk evaluating index system. This improved model can not only exploit the time frequency localisation characteristic of wavelet transforms (WT), but also enhance the fit precision and algorithm convergence speed. The simulation results show that this model is reliable, and that this method of identifying the investment risk of wind power projects is feasible.
- 出版日期2014-10
- 单位华北电力大学（保定）; 中国农业科学院农业经济与发展研究所; 华北电力大学
