Real-Time Perception Approach of Power Supply Capacity of Active Distribution Network Driven by Hybrid Data-Model

Real-time and accurate perception of active distribution network (ADN) power supply capacity (PSC) has important guiding significance for active control and real-time scheduling of ADN. In order to achieve this goal, we propose a hybrid data-model driven approach. Firstly, the uncertain factors in the system are modeled, and generation of multi-run scenarios for perception the PSC of the system by Monte Carlo Based Simulation Approach. Secondly, identifying the index to measure the PSC of the system, and the model-driven approach calculates the PSC of the system in different scenarios to obtain the training samples for data-driven approach. The PSC index calculated by the trained data-driven model are transmitted as the initial value to the model-driven algorithm to further analyze the PSC of the system. The data-model hybrid driven approach can effectively combine the advantages of the two approaches to achieve real-time accurate perception of the of the system's PSC, so as to guide the active control and real-time scheduling of the ADN. We finally use a modified IEEE33 node example to test the correctness and efficiency of this approach.