Electrical Fault Detection Using Machine Learning Algorithm For Centrifugal Water Pumps

Abstract

The most essential part of any living being is water. Humans utilize water for various purposes such as cooking, bathing, washing, drinking, cultivating, cleaning, power generation and so on. Water pumps are employed to facilitate easy access near the requirement. Pumps can be classified into many groups according to the method of fluid displacement they inculcate to transport the fluid. Various faults are associated with the working of the water pumps due to the way it is handled, environmental conditions, supply imbalance, poor power quality or due to any other mechanical failures. Hence an effective process to determine the different various faults so as to mitigate the damage is required. This paper analyses the various faults in the centrifugal water pumps driven by induction motors used in agriculture fields and proposes a new algorithm to effectively and efficiently identify the fault and classify it according to its category using machine learning algorithms.