Automation of multi-fault diagnosing of centrifugal pumps using multi-class support vector machine with vibration and motor current signals in frequency domain

摘要: Reliable detection and isolation of centrifugal pump (CP) faults is a challenging important task in the modern industries. Hence, this paper proposes an artificial intelligence-based multi-fault CPs driven by induction motor. The intelligent fault methodology developed based on multi-class support vector machine (MSVM). mechanical hydraulic are mutually dependent therefore may exist concurrently. present research, assortment various flow instabilities like suction re-circulation, discharge pseudo-re-circulation dry runs considered coexisting with faults, impeller cracks pitted cover plate faults. power spectrum CP vibration motor line-current data used for monitoring condition. best statistical feature combination selected wrapper model. Gaussian radial basis function (RBF) kernel mapping. In addition, RBF parameter (width) MSVM parameters optimally using fivefold cross-validation technique. Also, variation operating speed drastically changes system level owing to change manifestations; hence, work that independent operation proposed tested. Thereafter, it observed remarkably robust successfully classifies multiple individual as well at all tested speeds.