Real-time electricity price prediction method based on improved Adam algorithm optimization deep neural network

Abstract

The invention relates to a real-time electricity price prediction method based on an improved Adam algorithm optimization deep neural network. The method comprises the following steps: (1) obtaining related data of electricity price and influence factors thereof in a power system as sample data; (2) carrying out normalization preprocessing on the related data of the electricity price and the influence factors thereof in the power system; (3) determining an input mode, an output mode, a hidden layer number, a hidden layer neuron number, a hidden layer transfer function, an output layer transfer function and a loss function of the neural network, improving the traditional Adam algorithm, and using the improved Adam algorithm to optimize the deep neural network model; (4) taking the influence factor which is highly correlated with the actual electricity price as an input quantity, taking the predicted electricity price as an output quantity, training a deep neural network model based on an improved Adam algorithm, and optimizing parameters of the deep neural network model; and (5) processing electricity price influence factor data of different nodes in the power system by using the finally optimized deep neural network model, and predicting real-time electricity prices of different nodes. According to the method, the data utilization sufficiency can be improved, the convergence speed of training is accelerated, and the accuracy of electricity price prediction is improved.

Description

Technical field [0001] The present invention relates to the field of electricity market, in particular, to a real-time electricity price prediction method based on improved ADAM algorithm optimization depth neural network. Background technique [0002] Real-time electricity prices refers to the marginal cost of electricity to provide electrical energy in consideration of the power system operation and basic investment. It directly responds to the market price and the relationship between market prices and current market purchase costs. Is one of the most ideal electricity price mechanisms. Accurate prediction of real-time electricity prices, one hand can provide a reliable value basis for purchasing power users, thereby formulating scientific power policies; on the other hand, it can provide an important reference for the power market regulatory authorities, which in turn has established a reasonable market rules to promote The health, stable and orderly development of the electr...

AI Technical Summary

Problems solved by technology

[0004] The limitation of the forecasting method based on real-time electricity prices is that: with the integration of new energy and new equipment into power grids at all levels, the time series of electricity prices presents more complex nonlinear characteristics, which makes it difficult for time series forecasting methods to select appropriate input variables number; and the real-time electricity price prediction method using artificial neural network, it is easy to cause the prediction model to overfit and affect the prediction performance of the model; although the prediction method based on support vector machine overcomes the existing problems in the artificial neural network prediction method. It has disadvantages such as poor generalization ability and slow convergence, but large-scale training sample data will lead to a significant decrease in its calculation timeliness

Therefore, it is difficult to achieve the desired effect by using the existing real-time electricity price forecasting method

Images