Application of AI Techniques for Condition Monitoring of AC Drives For Prognostic Diagnosis

Abstract

Electrical motors consume around two-thirds of total industrial power; amongst them, newlineinduction motors constitute upto seventy five percent of total motor loads. Induction newlinemotors are extremely reliable, robust and efficient with low maintenance.They newlineare preferred in high power industrial drives.Downtime of these motors affects the newlineproductivity of the plant or process that uses these motors. The financial loss due to newlinefailure of electric drives is huge and a major concern in mining and manufacturing newlinesectors due to loss of production. Modern drives involve use of high power electronic newlinecircuits and the control is automated. Further digital technologies are used newlineextensively for control of the power electronics and also for automation. This has resulted newlinein more possible failure modes in the industrial drives, thus causing a threat newlineto reliability. A simple failure of the control system or any device in the power electronic newlineconverter leads to shut down of the entire plant, resulting in huge losses. Often newlinethe faults occurring in the drive system go unnoticed and advance over Time, as the newlinemotor remains operational but under reduced efficiency. Incipient faults are those newlinethat are not severe enough to activate the primary protection such as relay. Normally, newlinethese faults do not disrupt the operation of the drive and often the drive continues to newlinerun . Hence they go unnoticed, but are potential hazards, that may over time create newlinedisruption and possibly result in down time of the entire system. Hence continuous newlinemonitoring of the drive system is essential for early detection of incipient faults and newlineprevention of a cascading effect. newline