Combinating Forecast Of Mid-term Electricity Demand In Guizhou Province

Network Planning determines whether the power supply can meet the needs of re gional economic development,and power load forecast is the basis of network planning. After simulating and analyzed historical data(2003-2013) of Guizhou province through several models,we found that using partial least squares regression (PLS) can overco me the dependency issues between explanatory variables,and the prediction accuracy ar rived at 96.95%.At the same time, the accuracy of GM(1,1) reached 96.53%.In additio n,the combined forecasting method can integrate the advantages of single forecasting model and reduce the risks of relying a single prediction model.At the end,the paper combined PLS with GM(1,1) to simulate power consumption of Guizhou province.Fina lly,the paper obtained power consumption of Guizhou province from 2015 to 2O2O.Ac cording to the power grid construction strategy and the general trend of economic dev elopment in Guizhou Province,the paper proposed the outlook of network planning in years of 2016-2020.