Identification of Shaft Orbit Based on the Grey Wolf Optimizer and Extreme Learning Machine

I341 the search of shaft orbit identification of hydropower generating unit, the parameter and feature subset selection of classifier could not be adaptive, thus the shaft orbit cannot be identified fault fast and accurately. Aiming at the problems above, this paper proposed a novel method to identify the shaft orbit with feature subset and parameter optimized simultaneously namely HGWO-ELM. This method employed grey wolf optimizer to search the global optimal parameter and feature selection and improve it in the aspect of search strategy, food source and update equation. Through the simulation experiment, the methods proposed in this paper shows better performance than other classification methods and have some guidance significance to fault diagnosis of hydropower generating unit.