ABSTRACT
Power industry restructuring is gradually evolving in China. A potential reform in the near future is to open the retail markets [1-2]. By then, the Load Serving Entity (LSE) will no longer be a monopoly, but an independent market participant instead. From the trading point of view, the risk faced by a LSE is mainly from electricity price volatility of the generation side and the uncertainty of load demand. Future load can be estimated using load forecasting methods [3-4], and electricity price volatility can be forecasted by methods such as the wavelet transform and artificial neural networks [5-6], but forecasting error is inevitable. The LSE should properly and dynamically allocate the amount of electricity to be purchased among various markets so as to maximize the profits and minimize the risks.
