Deregulated Electric Energy Price Forecasting in NordPool Market using Regression Techniques

Deregulated electricity market day-ahead electrical energy price forecasting is important. It is influenced by external parameters and it is a complicated function. In this work two neighboring regions in the NordPool market are analyzed to provide day-ahead electrical price forecasting using regression techniques. The characteristics of the NordPool market trading behavior leads to unanticipated price peaks at daily, weekly and annual level. The considered two Nordic regions have different energy generation sources (e.g Norway has controllable hydro power, Denmark has non-controllable wind-power) therefore day-ahead electrical energy price forecasting in deregulated market for these two neighboring countries can impact the energy generation dynamics. Due to the dynamics of the electricity markets and region divisions this work proposes electric energy price forecasting using regression tools based in the k-Nearest Neighbor regressor, to capture the small increments in changing price behavior, and an autoregressor on top to capture the finite gradient of the occasional spikes in the price cycles. The electric energy price forecasting analysis of these two regions have shown that there is much more accurate forecasting in the Norwegian market compared to Danish market. To improve the forecasting in Danish market impact of external parameters (e.g meteorological parameters) should be considered for improving the forecasting.