An Artificial Neural Network Approach for Short-Term Electric Prices Forecasting
In this paper, a forecasting system of electric price is proposed to predict the short-term electric prices for avoiding the risk due to the electricity price volatility. Based on the Back-propagation Neural Network(BPN) and Orthogonal Experimental Design(OED), a New Artificial Neural Network Approach(NANNA) is constructed in the searching process. The data cluster, including Locational Marginal Price(LMP), system load, temperature, line-flow, are first collected and embedded in the Excel Database. In order to get a better solution, the OED is used to automatically regulate the parameters during the NANNA training process. Linking the NANNA and Excel database, the NANNA retrieved the input data from Excel Database to perform and analyze the efficiency and accuracy of the predicting system until the forecasting system is convergent. Simulation results will provide the participants to obtain the maximal profits and raise its ability of market’s competition in a price volatility environment.
M. T. Tsai and C. H. Chen, "An Artificial Neural Network Approach for Short-Term Electric Prices Forecasting", Advanced Materials Research, Vol. 267, pp. 985-990, 2011