Research on remote hydro-generator sets diagnosis system

A type of remote monitoring and diagnosis system is brought forward which based on Matlab Web Server.Firstly,wavelet packet decomposition is introduced to acquire energy features of which reflect hydrogenerator sets performance to be Feature Parameter.Then these Feature Parameters can be adopted as BP Neural Network input variable to realize fault diagnosis.Most of all,it is the first time to adopt Matlab Web Server to hydro-generator sets faults diagnosis field to implement distributed remote monitoring and diagnosis system.Therefore,remote diagnosis application is independent from the OS used on server side.There is no need for software maintenance by clients.And clients can finish remote diagnosis by Web Browser and without installation of Matlab-software.Client users can monitor and diagnose hydro-generator sets by Browser.Finally,further research work is pointed out such as hydro-generator sets fault modeling,accelerating BP Neural Network learning speed and convergence property,improving data transfer speed of Matlab Web Server to meet the needs of real-time diagnosis for hydropower generator sets.