Study On Power Generation Enterprise' Bidding Strategies Based On Price Forecasting

At present,the electricity industry in the world-wide is undergoing profound changes.The reform of the power industry is to improve the efficiency of electricity production,rationalize the electricity price formation mechanism,provide high-qulity,safe power products,and promote the power industry itself healthy development and finally,to reform the entire society for the better economic and social efficiency.The marketization of power industry is a development trend of the current power industry in the world.In the condition of electricity market,the economic benefits of power plants are derived from its generation capacity.However, the income of the power plants depends largely on the on-grid prices for electricity.The electricity price is the fulcrum of the electricity market,which affects the vital interests of each power plant. Therefore, electricity price forecasting for all participants is of great significance.It has become an important part of electricity market.The paper studies the bidding strategies based on predicting the market clearing price of the power generation companies in the new electricity market environment. At first, by analyzing the features of BP neural network and giving an account of the problem of traditional BP neural network that the criterion-Mean Square Error cost function can't consider the higher-order statistical statistisc of the systems, the neural network model based on Minimum Entropy Error cost function is proposed to improve the preciseness of electricity price prediction,while the batch-sequential algorithm is utilized to improved the forecasting velocity. Secondly,according to the characteristics of randomness and fuzziness in electricity price sequence,cloud theory is combined with neural networks. The logical reasoning is integrated into the cloud neural network,which is a form of evolution of fuzzy neural network. Finally, the model is used in a real case to forecast the next-day electricity price, which has a satisfactory prediction results.