Dynamic Identification of Load Model Parameters Based on Data-driven Approach and Its Application in Transient Stability Preventive Control

This article proposes a power system transient stability preventive control method that considers data-driven identification of the distribution network dynamic load model. Firstly, the distribution network and its internal loads connected to each node of the large power grid are equivalently transformed into a dynamic synthetic load model (SLM), and the dominant parameters in the model are extracted. Based on a deep belief network (DBN), the dominant parameters are dynamically identified; Secondly, integrate the identified model into the large power grid and generate data-driven transient stability constraints based on DBN; Establish a transient stability preventive control optimal power flow model with the objective function of comprehensively adjusting generator output and minimizing the total cost of load, considering multiple constraints such as data-driven transient stability margin, and solve it using the NSGA-II algorithm; Finally, the effectiveness of the proposed method was verified through simulation analysis of an improved IEEE 39 nodes system.