Aiming at the high short-term prediction accuracy of hourly water consumption in water supply optimal operation,a united time series method was proposed based on horizontal period clustering and vertical residual modification.The horizontal time series were determined as research samples by pattern recognition with high relevancy.After the phase space of horizontal period clustering was restructured,chaotic characteristics of typical period data were analyzed and chaotic prediction model was established.Least square support vector machine was used as a forecasting tool to obtain period flows.Furthermore,to track water consumptions dynamically,vertical residuals were modified by grey model prediction after collecting real-time data.The period historical data from Xiaoshan were supplied in the normal and abnormal case study to forecast the day-ahead hourly water consumption,and prediction accuracy with different methods were compared.Test results show that the method is adept in the short-term forecasting this kind of chaotic time series,which reflects the characters of typical days and real-time water variations.And the new method outperforms other methods obviously,thus it can better satisfy the demands for optimal operation of water distribution system.
ZHANG Qin, Wang-Xiong-Hai, Zhu-Qing-Jian. Short- term prediction of chaotic hourly water consumption based on united time series[J]. Journal of Drainage and Irrigation Machinery Engin, 2011, 29(4): 352-358.
Alvisi S, Franchini M, Marinelli A. A short-term, pattern based model for water-demand forecasting [ J ]. Journal of Hydrology,2007,9 ( 1 ) :39 - 52.
KIDSll O. Stream flow forecasting using different artifi- cial neural network algorithms [ J ]. Journal of Hydrologic Engineering ,2007,532 - 539.
l,i Tianliang,He l.iming,l.i Haipeng. Prediction and a- nalysis of chaotic time series on the basis of support vee-tor[ J ]. Journal of Systems Engineering and Electronics, 2008,19(4) :806 - 811.
Gato S,Jayasuriya N, Roberts P. Forecasting residential water demand: Case study [J]. Journal of Water Re-sources Planning and Management, 2007,133 ( 4 ) : 309 -319.
Towler E,Rajagopalan B,Summers R S,et al. An approach for probabilistic ibrecasting of seasonal turbiditythreshold exceedance [ J ]. Water Resources Research,2010,46 ( 6 ) , W06511 , DOI :10.1029/2009 WR007834.
传真0511--84440033 Copyright 江苏大学杂志社 2010-2015 All