Voltage Sag Risk Assessment Based on Deep Neural Network

With the ongoing advancement of high and emerging technologies, precision power electronic equipment that is highly sensitive to voltage sags has become increasingly prevalent in industrial production. Due to the randomness and severity of voltage sag events, a large number of industrial economic losses and user complaints have been caused. Voltage sag assessment integrates information from both the grid side and the sensitive equipment on the user side to evaluate the risk of voltage sag at specific nodes. This approach allows power supply companies and users to better understand the potential risks associated with voltage sags. It also serves as a theoretical foundation for users to choose optimal connection points based on their specific conditions and to address voltage sag issues effectively. This paper analyzes the main reasons why users are affected by voltage sag, uses the deep neural network model to explore the potential relationship between the actual influencing factors and the user's voltage sag risk, and uses the comprehensive voltage sag index to evaluate the user's voltage sag risk. Finally, the feasibility of the proposed method is verified through real calculation examples.