Application Of The Artificial Fish School Algorithm Optimized Grey Model In Power Load Forecasting

Power load forecasting is an essential foundation work in power system planning andoperation. Accurately forecast electric load is an important feature to ensure that providereliable power supply for the various departments of the national economy and people’s lives,at the same time, it is the first condition to ensure the power industry to stable development. Itgoes without saying its importance.Until now, domestic and foreign scholars and experts continue to explore loadforecasting, forming a series of effective measures. However, various methods have their ownscope and varying degrees of deficiency, therefore, can be combined with the actual situationin a reasonable choice of prediction methods and possible methods to improve thedeficiencies of the existing work load forecasting is the key.This paper first describes the significance of the topic and purpose, based on theforecasting methods, summarizing the development of load forecasting; focuses on the greyforecasting model which on the basic of grey system method, and pointing out this subject ismainly research the GM (1,1) model. Although the grey forecasting model has been widelyused, but it also has shortcomings. By analyzing the mechanism and its modeling applicationconditions, pointing out limitations of GM (1,1) model, the author propose to use artificialfish swarm algorithm to optimize the model. With the development of technology, multi-agentalgorithm that based on animal behaviors played an increasingly significant role inoptimization problems. Zhejiang University, Li Xiaolei, who proposed the artificial fishschool algorithm in2002is a group of intelligent multi-point parallel optimization algorithm,the biggest advantage of this algorithm is it able to overcome local optima, and achievingglobal extreme.On the basis of traditional grey forecasting model, this paper firstly apply artificial fishschool algorithm to optimization the parameters of the grey prediction model, andestablishing artificial fish grey model(AFSA-GM). In the environment of MATLAB,finished fish algorithm the performance verification as well as the basic grey prediction modeland the improved model simulation. Results showed that fish algorithm has good globalsearch ability, and is suitable for grey model parameter optimization problems; the improvedmodel has high precision and less prediction errors, that AFSA-GM model has higheraccuracy of load forecasting, and has wider application.