Load Forecasting Model Based on Amendment of Mamdani Fuzzy System

When neural networks are used to forecast short-term power load, it can learn the experience by training and generate mapping rules, but these rules are not directly understood in the network. By using the method of integrating neural networks and fuzzy logic, neural networks only settle historical load information. Moreover, fuzzy logic considers the factors which have great effect to load varying, such as air temperature and holidays, etc. According to the own characteristics of short-term load, the membership function are constructed, and the modifying of basic load heft is realized, which can enhance the load forecasting results veracity to a certain extent.