Short-term load forecasting is a part of the advanced function of power dispatching automation system. Accurate load forecasting will help to improve power system security and stability, to reduce the cost of power generation. With the development of power industry, short-term load forecasting will play an increasingly important role. Fuzzy system can approximate any function by limited rules, and is robust. And the genetic algorithm is intelligent search algorithm, which is built on biosphere natural selection principle and natural genetic mechanisms. Genetic fuzzy systems is a new hybrid system emergence in recent years, can combine fuzzy system and genetic algorithms to develop their own strengths. For load forecasting problems, this paper takes the fuzzy system with network structure as the basic model, and the improved genetic algorithm is taken to confirm system parameters, and then establish model of prediction system. And comparison and verification through case study. The example shows that the method proposed by this paper can gain a better result than the general fuzzy system.
|