由於電力衍生性商品的崛起，對電力價格建模於能源市場中已成為一個熱門議題。2003年 Huisman等人引進了狀態轉換模型，並已在許多研究中，藉由比較過去的電價模型確立了狀態轉換模型 (Regime-switching model) 在捕捉電力價格動態上的突出表現。而在狀態轉換模型之下，價格的預測由於未來的不確定狀態而變得十分困難。在本篇論文中，我們引進了與價格相對應的需求過程，並在不同的模型假設下改進狀態轉換模型於電價上的預測能力。我們也於 PJM 電力市場上收集了一年期的資料並比較在不同狀態假設下的預測表現。Modeling the electricity price has become a popular issue in energy market due to the rise of electricity derivatives. Regime-switching model have been introduced to electricity market in Huisman and Mahieu (2003) , And has been confirmed its outperformance by comparing to other models in many studies. Under the regime-switching model, prices become hard to forecast due to the unknown current states in the future. In this paper we involve the demand process to improve the forecasting performance under different model assumptions. We collect one-year day-ahead prices in PJM market and compare the forecasting result with two different regime assumption models.
