An integrated fuzzy regression–analysis of variance algorithm for improvement of electricity consumption estimation in uncertain environments
- 142 Downloads
This study presents an integrated fuzzy regression–analysis of variance (ANOVA) algorithm to estimate and predict electricity consumption in uncertain environment. The proposed algorithm is composed of 16 fuzzy regression models. This is because there is no clear cut as to which of the recent fuzzy regression model is suitable for a given set of actual data with respect to electricity consumption. Furthermore, it is difficult to model uncertain behavior of electricity consumption with conventional time series and proper fuzzy regression could be an ideal substitute for such cases. The algorithm selects the best model by mean absolute percentage error (MAPE), index of confidence (IC), distance measure, and ANOVA for electricity estimation and prediction. Monthly electricity consumption of Iran from 1992 to 2004 is considered to show the applicability and superiority of the proposed algorithm. The unique features of this study are threefold. The proposed algorithm selects the best fuzzy regression model for a given set of uncertain data by standard and proven methods. The selection process is based on MAPE, IC, distance to ideal point, and ANOVA. In contrast to previous studies, this study presents an integrated approach because it considers the most important fuzzy regression approaches, MAPE, IC, distance measure, and ANOVA for selection of the preferred model for the given data. Moreover, it always guarantees the preferred solution through its integrated mechanism.
KeywordsFuzzy regression Fuzzy mathematical programming Electricity consumption Analysis of variance Uncertainty Mean absolute percentage error Index of confidence
Unable to display preview. Download preview PDF.
- 5.Azadeh A, Ghaderi SF, Anvari M, Saberi M (2006a) Measuring performance electric power generations using artificial neural networks and fuzzy clustering. In: Capolino GA, Franquelo LG (eds) Proceedings of the 32nd annual conference of the IEEE Industrial Electronics Society, IECON, Paris, 2006Google Scholar
- 7.Azadeh A, Ghaderi SF, Tarverdian S, Saberi M (2006c) Integration of artificial neural networks and GA to predict electrical energy consumption. In: Capolino GA, Franquelo LG (eds) Proceedings of the 32nd annual conference of the IEEE Industrial Electronics Society–IECON’06, Conservatoire National des Arts & Metiers, Paris, 2006Google Scholar
- 24.Modarres M, Nasrabadi MM, Nasrabadi E, Mohtashmi GR (2003) Evaluation of fuzzy linear regression models: a mathematical programming approach. In: Proceedings of 4th seminar on fuzzy sets and its applications, Babolsar, pp 129–135.Google Scholar
- 53.Körner R, Näther W (1998) Linear regression with random fuzzy variables: extended classical estimates, best linear estimates, least-squares estimates. Inform Sci, 109, pp 95–118.Google Scholar
- 54.Özelkan EC (1997) Multi-objective Fuzzy Regression Applied to the Calibration of Conceptual Rainfall–Runoff Models, Unpublished PhD Dissertation, Department of Systems and Industrial Engineering, The University of Arizona.Google Scholar
- 55.Özelkan EC, Duckstein L (2000) Multiobjective fuzzy regression: a general framework. Computers and Operation Research 27, pp 635–640.Google Scholar
- 56.Weron R, Misiorek A (2008) Forecasting spot electricity prices: A comparison of parametric and semiparametric time series models. International Journal of Forecasting 24(4):744–763Google Scholar