Midterm Electricity Market Clearing Price Forecasting Using Two-Stage Multiple Support Vector Machine

Table 1

Comparison of previously published midterm forecasting techniques of electricity prices.

Authors	Forecasting objectives and techniques	Strengths	Weaknesses
Torbaghan et al. [1]	Midterm forecasting of the monthly average electricity spot prices using various techniques	Both single and hybrid models are tested	Only focused on the monthly average electricity spot prices (only 12 outputs)
Yan [2], Yan and Chowdhury [3, 4]	Midterm forecasting of the electricity hourly MCP utilizing ANN technique	Extended training data set	Poor performance in forecasting peak prices
Torghaban et al. [5]	Midterm forecasting of the monthly average electricity spot prices utilizing linear forecasting models	Dummy variables are considered in the model indicating seasonality during simulation	Only focused on the monthly average electricity spot prices (only 12 outputs)
Yan and Chowdhury [6]	Midterm forecasting of the electricity hourly MCP utilizing hybrid LSSVM and ARMAX techniques	The proposed hybrid model is capable of adjusting the errors from the previous module	Limited improvements in forecasting peak prices
Pedregal and Trapero [7]	Midterm forecasting of the electricity hourly price utilizing unobserved component models	Utilization of numbers framed short-term forecast results to serve in the midterm forecasting	The proposed model is compared with an ARIMA model which has quite weak performance in midterm forecasting
González et al. [8]	Midterm forecasting of electricity hour price utilizing hybrid approaches based on the analysis between market price and marginal costs	Forecast the real price based on adjusting the prediction values of the marginal price forecasting	Only price data is involved in the proposed study and the forecast of the marginal price is under an ideal electricity market