A Generalized Auto-Regressive Conditional Heteroskedasticity Model for System Marginal Price Forecasting Based on Weighted Double Gaussian Distribution

The regularity of conditional variance variation and the statistical distribution characteristic of residual error of system marginal price(SMP) in electricity market are researched.On this basis,by means of leading in the generalized auto-regressive conditional heteroskedasticity(GARCH) model a GARCH model based on the assumption of weighted double Gaussian(WDG) distribution is proposed to research the variation regularity of SMP.Taking the actual data from PJM market in USA and NSW market in Australia as samples,both GARCH model and the proposed GARCH-WDG model are tested,the testing results show that the GARCH model can offer good estimation and forecasting results of SMP,and the performance of GARCH model can be further improved by the proposed GARCH-WDG model.