Data-Driven Security Assessment of Power Grids Based on Machine Learning Approach

Security assessment is a fundamental function for both short-term and long-term power system operations. The data-driven security assessment (DSA) criteria will help determine when it is necessary to trigger a dynamic simulation. The DSA criteria will provide a key indicator of switching the simulation method. It will be the link between the traditional isolated dynamic simulation and scheduling simulation.

This paper investigates a data-driven security assessment of electric power grids based on machine learning. Multivariate random forest regression is used as the machine learning algorithm because of its high robustness to the input data. Three stability issues are analyzed using the proposed machine learning tool: transient stability, frequency stability, and small signal stability. The estimation values from the machine learning tool are compared with those from dynamic simulations. Results show that the proposed machine learning tool can effectively predict the stability margins for the three stability metrics, i.e. transient stability, frequency stability and small-signal stability.