: Reliability and energy saving in industrial production line are of great importance. In result, energy efficiency and reliable performance of induction motors, as the dominant loads of energy in industry, have been in the spotlight. Energy saving in motors involves various technical, economical and management actions. Continuous condition monitoring, proper application and good maintenance are amongst different solutions for improving performance of production line and decreasing its down-time. Among those solutions, replacing existing motors with the objective of increasing technical reliability and economical efficiency is promising. With the advent of new generation of high-efficiency motors, decision for replacing low-efficiency motors has become challenging. High cost of new motors and the chance of repairing and reusing of old ones can make it difficult to make a reasonable choice. In fact, energy saving is a function of different variables such as motor cost, efficiency rate, operation hours, down-time cost, payback period and availability of government incentives and an analysis for replacement is impossible if those factors are not included. However, the prerequisite of these studies is an accurate index as a base for comparing the performance of motors. tandard tests usually need detaching motor from its service and performing different measurements. In recent years, online estimation of efficiency in induction motors has been subject of widespread researches and numerous methods have been proposed for it. The common objective of these studies is a method for online efficiency estimation of motors with high accuracy, practical effectiveness and low intrusion. The estimated efficiency can subsequently be employed for comparison of motors and replacement decision. In this thesis, the objective is to develop an algorithm for online estimation of efficiency for in-service induction motor and subsequently decision making about their replacement. For this purpose, two online methods for efficiency estimation based on airgap torque and electrical equivalent circuit have been proposed. The advantage of these methods is to minimize required measurements. Accordingly, by only measuring line voltages and currents, the efficiency of motor is estimated subsequently. In the following, by employing genetic algorithm, the accuracy of estimated parameters of motor is improved. The results also have been verified by a standard test. Finally, the methods are introduced for decision making about replacing low-efficiency, oversized and repaired induction motors with new high-efficiency ones and its different scenarios have been studied. Keywords: efficiency estimation, induction motor, optimization, parameter estimation, energy saving
