Non-Contact condition monitoring system for outdoor ceramic insulators based on their radio frequency signatures
The research centers around the feasibility study of a method to detect faulty outdoor insulators and to classify the type of defects within the insulator. Every different type of defect results in a specific electromagnetic Signature that can be acquired using an antenna and oscilloscope. Using these Signatures some specific features (eg statistical or spectral features) can be extracted which are used to train computer learning algorithms so that they can classify subsequent signatures into their respective classes. At kinectrics, field tests will be performed to obtain this data and to study how this technique could be applied in a real world scenario. This would also allow the study of the various different forms of noise that are known to pollute the measurements. Upon successful development this technology will result in a safe to use, cost effective, accurate and convenient way to test outdoor insulators for defects.