Induction machine efficiency estimation using population based algorithm

There has been a tremendous pressure to predict the in situ efficiency of induction machine (IM) with limited level of intrusion so as to improve IMs performance. Least research work is carried out to make IM efficiency evaluation methods compatible to unbalanced supply and varying load conditions. This paper proposes a novel approach using cuckoo search algorithm (CSA) to obtain efficiency estimation of an IM operating as a motor working with unbalanced supply having under or over voltage issues. CSA improves the searching ability and has capability to adapt to complex optimization problems. Here, CSA optimizes the IM positive sequence parameters at various loading levels. The parameters optimization is done with the use of positive sequence input currents and electrical powers which have been obtained earlier at various operating loading points. Using the optimized parameters, the negative sequence parameters can be evaluated. So, the efficiency of IM can be estimated at different loading levels. The proposed approach is implemented on the MATLAB platform. The effectiveness of the novel approach is established by comparing the results obtained with genetic algorithm (GA).