A new method for gyroscope fault diagnosis based on CGA RBFNN and multi-wavelet entropy
Ji Yu, Deyun Zhou, Peng Ju He, Jichuan Huang · 2013
An original method based on CGA (Genetic Algorithm based on Cloud-model) RBFNN (Radial Basis Function Neural Network) is proposed for the online fault diagnosis of gyroscope. Based on the information entropy and wavelet transform theory, wavelet energy entropy (WEE) and wavelet time entropy (WTE) are extracted to be the input of RBFNN. Besides, CGA is used to optimize the parameters of RBFNN. The simulation results show that this new method can reach an efficient search for global convergence, and prevent it from similar problem of partial efficiency in traditional genetic algorithm (TGA). After trained, the RBFNN can accomplish the online fault diagnosis of gyroscope accurately and quickly.