Short-term electricity price forecasting has always been one of the important issues in the electricity market. An accurate and applicable electricity price forecasting method can effectively avoid the risk of the trading market to maximize the economic effect. With the development of deep neural networks, machine learning methods have been widely used in the research direction of electricity price forecasting. This paper proposes a hybrid model of short-term electricity price forecasting based on BiLSTM-Attention. The model first uses the similar day method to select the similar days of the day to be predicted, and then introduces the BiLSTM model to predict the short-term electricity price. The results show that the proposed model can obtain high-precision short-term electricity price forecasting results in the ideal electricity market.
Aiming at the complexity and poor accuracy of short term electricity price prediction in electricity market, this paper proposes a medium and long term electricity price prediction model based on attention mechanism convolutional neural network and long term memory network. The model takes the historical electricity price data as input after data preprocessing. First, the convolutional neural network is used to extract the features of the historical electricity price series. Secondly, the extracted feature vector is constructed into a time series and input into the long and short term memory network, and the internal change law of the feature is fully mined for training. Then, the attention mechanism is introduced to reduce the error of model prediction, highlight the influence of important information and assign weight. The attention mechanism is weighted to sum the output of each time step in the long and short term memory network. Finally, the final predicted value is calculated by the function. Finally, a numerical example is given to verify the accuracy of the proposed model.
In response to the problem of low enthusiasm for energy efficiency transformation of high energy consuming enterprises under the traditional fixed electricity pricing model, a differential pricing mechanism is adopted in the process of power reform to promote energy conservation and emission reduction. Virtual power plants aggregate multiple distributed resources, providing energy-saving equipment transformation for high energy consuming enterprises while also providing power guarantee. This not only alleviates the power supply shortage problem of power supply companies, but also reduces their operating income, This has led to a cooperative and competitive relationship between virtual power plants and power supply companies. This article constructs a two-layer optimization model for this scenario, simulates and analyzes the decision-making game between the virtual power plant and the power supply company, obtains the final optimized results, and compares the user benefits and social comprehensive benefits before and after the user transformation.