A data pretreatment technique about power system load modeling

摘要

Short-term load forecasting is one of the most important routine works for power dispatch departments. The accuracy of load forecasting will exert direct effects on the safety, economy and stabilization of the power system running. Portrait and transverse comparability are employed to distinguish and correct bad load data, while wavelet analysis and multiple-time-period analysis used to eliminate long-term increasing weights, thus reducing the impact of the high-speed load increase on the accuracy of load forecasting.