New Energy Access Potential Evaluation in Station Areas Based on Deep Neural Network

Abstract

With the proposal of "carbon peaking and carbon neutrality" goals, the permeability of new energy in the station area is gradually improved. The volatility and randomness of its output greatly challenge the stable operation of the new station area. At present, as the photovoltaic and other new energy access consistently increases, the station area flow direction presents two-way random characteristics and the power converse problem becomes increasingly prominent, which has impeded further access of new energy, had a negative effect on the energy control and electricity safety system of the station area and even the superior distribution network, and affected the normal company marketing, business expansion, etc. In view of the above problems, this paper puts forward a station area new energy access potential assessment method based on deep neural network. Based on station area multi-section historical measurement data, using deep learning method the potential connection between the current station area state and accessible photovoltaic capacity is learned and the data-driven station area new energy access potential assessment model is obtained. This method considers the actual constraints of the station area’s complex environment and analyzes the bottleneck of new energy access, so as to optimize the resource allocation and maximize the new energy access potential in the station area. It is verified by example that this method realizes the accurate and real-time maximum new energy access potential assessment in low-voltage station areas.