The Short-Term Electrical Load Forecast Based On Fuzzy Neural Network

In 2011, the twelfth "Five-Year Plan" formulated by the state proposed to further implement the basic state policy of Energy Conservation and Protect the Environment, and intelligent power grid construction and transformation of power grids are important tasks. The short-term electrical load forecast, which will directly affect the power production and dispatching of power plant, is an important work of intelligent power grid construction and the management of operating and dispatching of electric power system. It also is the important foundation of ensuring the safety and economic and stability of operation for power system, and provides important information security for improving the social and economic benefits of power. So the demands for precision of short-term electrical load forecast becomes higher and higher.At present, there are many kinds of methods to forecast short-term electrical load. One of most widely methods used to forecast is artificial neural network. Because of the characteristics of nonlinear and learning, the artificial neural network can effectively improve the prediction precision, but there still exists some shortcomings such as slowly convergence and easily occurred concussion.Therefore, in this paper, a method of the short-term electrical load forecast based on fuzzy neural network which has the advantages of fuzzy logical control and artificial neural network was presented to improve the disadvantages of artificial neural network and the prediction precision. Firstly, summarize the concept, method and the domestic and foreign current research situation of electrical load forecast. Secondly, briefly introduce the theory of artificial neural network and fuzzy logical control, especially the structure design, computing process, parameter selection and shortcomings improvement of BP neural network and membership function of fuzzy logical control. And then construct a fuzzy neural network model by MATLAB simulation software, and base on the powerful data processing capability of MATLAB, learn and train the electrical load data of a certain power grid. Finally, develop a short-term electrical load forecast software package based on fuzzy neural network by using visual programming software Visual Basic.Using the fuzzy neural network and load-data processing based on the same time of different dates to conduct short-term electrical load forecast in this paper. By comparing and analyzing the prediction precision of the improved BP neural network and fuzzy neural network and other based on fuzzy neural network for short-term electrical load forecast research articles, shows that the proposed method used in this paper improves the prediction precision and operability and has important use value.