Research on Fault Localization of Intelligent Building Mechanical and Electrical Equipment Based on Improved Grey Wolf Algorithm

The Grey Wolf Optimization (GWO) algorithm, which mimics the social structure and hunting habits of grey wolves and views the location of prey as the best option, is introduced in this article. An enhanced grey wolf optimization algorithm is suggested, which employs Chebyshev mapping to create the initial population and enhances the nonlinear strategy of control parameters and position update formulas in order to address the grey wolf algorithm's limitations with regard to population diversity and the harmony between global search and local optimization. The classifier's recognition accuracy is directly correlated with the fitness function's value. The experimental verification section collected signals from two types of household electrical loads (linear component incandescent lamps and nonlinear component microwave ovens) under normal and fault conditions, totaling 400 sets of data. SVM parameters were optimized using the GWO technique, and the results demonstrated that the classification performance was at its best when the SVM's kernel parameter (g) and penalty parameter (C) reached particular levels. The classification effect of linear load incandescent lamps is better than that of nonlinear load microwave ovens.