Multiple-global-best guided artificial bee colony algorithm for induction motor parameter estimation

An induction motor is the most commonly used motor in industry today. Motor circuit parameters are essential for designing, evaluating performance, and controlling the applications of the motor. However, it is difficult to measure the electric parameters, e.g., resistance and reactance, of induction motors accurately. Therefore, researchers have noted the parameter estimation of induction motors as an essential optimization problem. The artificial bee colony (ABC) algorithm is an efficient element of bioinspired optimization algorithms and has been successfully applied in numerous engineering applications. However, the ABC algorithm suffers from slow convergence and poor exploitation. Additionally, there are bleak chances of getting a fitter food source for scout bees using the the standard ABC algorithm scheme. Therefore, different solutions have already been proposed to avoid the flaws of the ABC algorithm. Nevertheless, the proposed solutions are either computationally intensive or prone to local optima traps or they require additional control variables to tune. Moreover, there is no systematic way to tune the additional control variables for yielding the optimal performance of the algorithms. Therefore, this research work proposes a novel variant of the ABC algorithm, which capitalizes on multiple global-best food sources rather than a single global-best food source. In addition, this research work proposes a novel scheme for enhancing the performance of the ABC algorithm’s scout bee. Two modifications for the performance enhancement of the ABC algorithm are proposed in this research work. The proposed algorithm is compared with various recently proposed variants of the ABC algorithm and various other available methods for estimating induction motor parameters. The performance of the proposed algorithm is also assessed using the chaotic map initialization technique. The results prove that the proposed algorithm is able to achieve the best convergence among all of the compared algorithms.