The Grey Forecast Model Based Artificial Bee Colony Algorithm And Its Application

Grey prediction model is one of the most important contents of grey system theory.as well as an significant model of forecasting theories.It is mainly applied to solving the uncertainty problem of "small sample" and "poor information".Compared with other forecasting methods,there are amounts of advantages in grey prediction model,such as small sample,simple model,strong applicability.Moreover,it has been widely used in agriculture,economy,industry and many other fields in recent years.Therefore,the research on grey prediction model is extremely meaningful.Although the grey prediction model has been applied extensively in a lot of professions,it also needs to be further improved.Therefore,this paper took the grey prediction model as research object,and point out the limitations of GM(1,1)model through the analysis of its modeling mechanism and application scope.Aiming at the problems such as limited applicability and low prediction accuracy,a new optimization method was proposed.Firstly,Artificial Bee Colony(ABC)was proposed to optimize the background value of the model,and ABC-GM model was established.In the MATLAB environment,based on the statistical data of China's electricity consumption in 2003-2014,the prediction accuracy of the optimized model is less affected by the coefficient of development.And the prediction accuracy was improved by about 0.5%compared to the original GM(1,1)model.Secondly,based on the artificial bee colony gray prediction model(ABC-GM),this paper present the concept of optimal input sequence and established a new model(IABC-GM).And we verified the model through carrying out examples in the MATLAB environment.The results showed that the prediction accuracy of IABC-GM model was improved by 1.92%and 1.43%respectively,compared with GM(1,1)model and ABC-GM model.