Fault Diagnosis of Gyroscope Based on HAFSA-SVM

Aiming at the difficulty in extracting fault features of gyroscope output signals. A fault diagnosis method based on improved artificial fish swarm algorithm (AFSA) and support vector machine (SVM) was proposed. Firstly, the gyroscope signal is decomposed by three-layer wavelet packet, and the energy spectrum is extracted as the feature vector. Secondly, by combining artificial fish swarm algorithm and genetic algorithm (GA), a hybrid artificial fish swarm algorithm (HAFSA) is proposed to optimize the kernel function parameters and penalty factor of support vector machine. Finally, HAFSA-SVM model is established to identify and diagnose gyroscope faults. Results show that the HAFSA-SVM method has higher classification accuracy in the fault diagnosis of gyroscopes. Compared with AFSA-SVM, CV-SVM and GA-SVM, its diagnostic accuracy has been improved by nearly 2.5%, 4.2% and 5.9%, respectively, which provides a new idea for the fault diagnosis of gyroscopes.