Nonlinear combined forecasting model based on fuzzy adaptive variable weight and its application

Abstract

In order to enhance forecasting precision of problems about nonlinear time series in a complex industry system, a new nonlinear fuzzy adaptive variable weight combined forecasting model was established by using conceptions of the relative error, the change tendency of the forecasted object, gray basic weight and adaptive control coefficient on the basis of the method of fuzzy variable weight. Based on Visual Basic 6.0 platform, a fuzzy adaptive variable weight combined forecasting and management system was developed. The application results reveal that the forecasting precisions from the new nonlinear combined forecasting model are higher than those of other single combined forecasting models and the combined forecasting and management system is very powerful tool for the required decision in complex industry system.

Keywords

nonlinear combined forecasting / nonlinear time series / method of fuzzy adaptive variable weight / relative error / adaptive control coefficient

References

[1] BarryF., AnindyaR.. Forecasting using the trend model with autoregressive errors [J]. International Journal of Forecasting, 2005, 21(2): 291-302

[2] YuanF.-m., WangX.-h., ZhangJ.-m., ZhangLi.. Online forecasting model of tundish nozzle clogging [J]. Journal of University of Science and Technology Beijing: Mineral, Metallurgy, Material, 2006, 13(1): 21-24

[3] MichaelS. L. B., TienC.. Forecasting presidential elections: When to change the model [J]. International Journal of Forecasting, 2008, 242: 227-236

[4] WuX.-l., WangC.-h., ChenX., XiangX.-h., ZhouQuan.. Kalman filtering correction in real-time forecasting with hydrodynamic model [J]. Journal of Hydrodynamics, 2008, 20(3): 391-397

[5] MohammadiK., EslamiH. R., KahawitaR.. Parameter estimation of an ARMA model for river flow forecasting using goal programming [J]. Journal of Hydrology, 2006, 331(1/2): 293-299

[6] ChengC. H., ChenY. S., WuY. L.. Forecasting innovation diffusion of products using trend-weighted fuzzy time-series model [J]. Expert Systems with Applications, 2009, 36(2): 1826-1832

[7] AgamiN., AtiyaA., SalehM., El-shishinyH.. A neural network based dynamic forecasting model for trend impact analysis [J]. Technological Forecasting and Social Change, 2009, 76(7): 952-962

[8] KomorníkJ., KomorníkováM., MesiarR., SzökeováD., SzolgayJ.. Comparison of forecasting performance of nonlinear models of hydrological time series [J]. Physics and Chemistry of the Earth: Parts A/B/C, 2006, 31(18): 1127-1145

[9] LöfM., LyhagenJ.. Forecasting performance of seasonal cointegration models [J]. International Journal of Forecasting, 2002, 18(1): 31-44

[10] TsekourasG. J., DialynasE. N., HatziargyriouN. D., KavatzaS.. A non-linear multivariable regression model for midterm energy forecasting of power systems [J]. Electric Power Systems Research, 2007, 77(12): 1560-1568

[11] TsaurR. C.. The development of an interval grey regression model for limited time series forecasting [J]. Expert Systems with Applications, 2010, 37(2): 1200-1206

[12] LiX.-q., LiC.-sheng.. An improved method of fuzzy variable weight combination forecasting [J]. Journal of Central South University of Technology: natural Science, 2003, 34(6): 708-710

[13] Bar-GeraH., BoyceD.. Solving a non-convex combined travel forecasting model by the method of successive averages with constant step sizes [J]. Transportation Research Part B: Methodological, 2006, 40(5): 351-367

[14] ZouH., YangY.-hong.. Combining time series models for forecasting [J]. International Journal of Forecasting, 2004, 20(1): 69-84

[15] TangX. W., ZhouZ. F., ShiY.. The variable weighted functions of combined forecasting [J]. Computers and Mathematics with Application, 2003, 45(4/5): 723-730