A Predictive Dependent Power Storage Control Model Based on GA-BP Algorithm

Energy forecasting can assist the dispatching of power grid, but the uncertainty of energy forecasting troubles power grid to a large extent. With the development of energy storage technology, the combination of energy prediction and energy storage reduces the uncertainty of energy prediction to a certain extent. At present, most studies on the combination of energy prediction and energy storage pay too much attention to the accuracy of prediction, while in the general trading market, economic factors such as price and cost have a more important impact. In this paper, a predictive dependent power storage control model is proposed. The model uses genetic algorithm to find the energy storage power that minimizes the cost of energy storage construction and energy prediction. This model uses genetic algorithm to find the energy storage power that minimizes the cost of energy storage construction and energy prediction. By comparing with BP model trained with accuracy as index, we prove that we can greatly reduce energy prediction cost and storage cost under the condition of sacrificing certain accuracy.