Answers are grounded in this paper and cite their sources.
The widespread integration of variable energy resources, recent changes in load patterns, and the increase in frequency and intensity of extreme events have heightened the importance of fast and comprehensive grid reliability assessment for planning and decision-making purposes. While the Monte Carlo simulation with DC or AC power flow modeling has been a traditional method for power system reliability evaluation, its substantial computational requirement poses a significant constraint for fast assessment of grid reliability under various decision-making scenarios, such as impacts of component hardening or replacement of conventional generators with variable renewable resources. This paper introduces a Convolutional Neural Network modeled on a graphics processing unit into a non-sequential Monte Carlo Simulation (CNet+NMC) model to enhance computational efficiency without compromising evaluation accuracy for grid reliability assessment. Additionally, the paper investigates the applicability of the proposed model for conducting sensitivity analyses on the impact of power system components on the overall system reliability. It also assesses the proposed model for bus-level reliability evaluation and load prioritization assessment. The proposed method is demonstrated on the IEEE Reliability Test System and is compared to the baseline Monte Carlo simulation approach. Results show that the model provides more than 80% time savings during reliability analysis while maintaining good evaluation accuracy for sensitivity analysis and load prioritization assessments.
No takes yet. Share an insight, caveat, or question.
Olowolaju et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: