Research on Power Outage Data Mining Pattern using Apriori-based Association Rule Algorithm

Power outage analysis is critical for ensuring grid reliability, optimizing infrastructure, and improving emergency response strategies. Traditional methods for analyzing power outage patterns often face inefficiencies due to the large-scale, real-time nature of outage data and the complex dependencies between events. This research introduces a data mining framework leveraging the Apriori algorithm within Association Rule Mining to extract meaningful patterns from real-time power outage data. The proposed model utilizes the ODIN Real-time Outages dataset, which provides near real-time power outage information at the county level across the China. By applying the Apriori algorithm, the framework efficiently identifies frequent outage patterns, uncovering hidden relationships between occurrences, affected regions, and influencing factors. Experimental results validate the effectiveness of the approach, achieving a Rule Confidence of 92.4% and a Support Threshold of 85.7%, ensuring high-quality pattern extraction. Additionally, a Lift Ratio of 1.8 and Association Rule Coverage of 89.6% confirm the model’s robustness and predictive capability. These findings highlight the potential of Apriori-based rule mining in transforming outage management by providing insights, enhancing predictive maintenance, and supporting intelligent decision-making.