Bidding Strategy Research For Hydropower Plant In Electric Power Market

The produce goal of the hydropower plant is changing from original pursuing the most generate electricity quantity into maximal income under the electric power market. The hydropower plants have to be concerned about both electric energy price and power generation amount. How to get the most profits by using the politic bidding is becoming a new working focus point for hydropower plant.The key of realizing politic bidding is forecasting next-day system marginal price (SMP) for hydropower plants. To forecast next-day SMP, a mathematical model based on analogue searching and artificial neural networks (ANN) is developed. For achieving a fine result, forecast next-day SMP mathematical model based on analogue searching of data mining and 3-layers momentum back propagation (MOBP) ANN is designed, whose input is made up of the data of the similar historical days and the adjacent days, and whose output is the SMP of next-day. MOBP is made use of training the model. It is realized using the neural network toolbar in Matlab6.5. The forecast test shows that the method is effective and feasible for forecasting the working day's SMP.In order to improve the precision of the weekend SMP forecast, we establish the weekend SMP forecasting model from analyzing the weekend SMP. Contrast to the former model, the interferential factors is canceled and the training matrix is made up of the load similar days' data. The test shows that the method is effective.SMP forecasting establishes the foundation for realizing the politic bidding in hydropower plants. Taking one hydropower plant of Sichuan Province as an example, we optimize structure of its power generation amount to the network in 2002 by using the SMP forecasting results. Research results have been testified they can increase the incomes of hydropower plants, and show some values in theory and practice.