Electricity price thresholding and classification

With the advent of smart grids, electricity consumers are expected to be enabled to respond to electricity price fluctuations. Thus, forecasting short-term future electricity prices is a key component of optimal demand-side management in an enabled competitive environment. While several price forecasting approaches are available, achieving low forecasting errors is not always possible, especially for markets with high price volatility. On the other hand, for demand-side management applications the exact value of future prices is not primarily required, rather, whether prices hit certain thresholds is the basis of making planning decisions. In this paper, classification of future electricity market prices with respect to pre-specified price thresholds is investigated. Two alternative models based on support vector machines are proposed in a multi-class, multi-step-ahead price classification context. Numerical results are provided for price classification in Ontario´s and Alberta´s markets.