Intelligent Building Electrical Fault Diagnosis Based on Relevant Vector Machine Optimized by Intelligent Algorithm

Smart buildings are developing rapidly, and electrical fault diagnosis is becoming increasingly important for the stable operation of smart buildings. To improve the accuracy and efficiency of electrical fault prediction, this study constructs a model based on an improved gray wolf optimization algorithm to optimize the relevant vector machine. A multi‐core correlation vector machine model is proposed by optimizing the parameters of the correlation vector machine through an improved gray wolf optimization algorithm. The results demonstrated that the accuracy of the proposed model reached 96.8%, which was the highest improvement of 17.2% compared to the comparative model. On an imbalanced dataset, the accuracy of the research model was 96.4%, the recall was 86.4%, and the F1 value was 0.91, verifying that the model performed well in classification tasks and had strong generalization ability. At the same time, the false alarm rate of the system in practical applications was 1.3%, and the false negative rate was 2.6%, indicating that the system had strong real‐time monitoring and identification capabilities for electrical faults. This study provides a theoretical basis for electrical fault diagnosis in smart buildings and reliable technical support for the stable operation of smart buildings. © 2025 Institute of Electrical Engineers of Japan and Wiley Periodicals LLC.