A Data-driven Technique Based on Power System Reliability Assessment

Power system reliability assessment is critical to ensure the safety of power systems. This paper proposed a method to quantitatively assess the risk of power systems using Bayesian networks. Based on the fuzzy comprehensive evaluation (FCE) method, indicators and importance weights for power system reliability assessment are established, indicators that may pose an impact on power system reliability are measured quantitatively, and finally, power system reliability is rated according to the evaluation matrix and indicator weights. The risk of the whole power system was evaluated by using the historical data of power system reliability. The results of the study can provide a theoretical basis for power system design, prevention and reduction of system accidents, and a reference for improving the reliability of power systems.