Online Sequential Extreme Learning Machine Algorithm for Better Prediction of the Real-time Electricity Price under Dynamic Environmental Changes
2019 IEEE. Spot price forecasting plays a key element in the electricity market. However, such forecasting usually depends on the traditional offline batch learning technologies, which cannot respond to new unexpected changes in the local power system environment. Further, the local price can be affected by dynamic price changes from the connected regions. This paper proposes a novel online learning technology to take into account the above factors to supplement the traditional technologies. It can continuously monitor any potential unexpected events and the price fluctuation from the other connected regions. When the forecasted price exceeds a threshold value due to the unexpected changes, the online learning can be activated and can utilize the recent real-time price data to improve the trained result. The proposed method is validated using a numerical simulation on real market data, and the results show that the proposed method can help in improving the forecast accuracy, especially when unexpected changes occur both locally in the surrounding area.