Abstract
The estimated parameters accuracy of poly-phase induction motors is crucial for effective performance prediction and/or control in various manufacturing applications. This study investigates hybrid algorithm between particle swarm optimization and Jaya optimization algorithms for finding the optimal parameters estimation of poly-phase induction motors. It is carried out using the manufacturer's operation characteristics on two poly-phase induction motors. Numerical results show the capability of the proposed hybrid optimization algorithm. The proposed algorithm has competitive performance compared with conventional algorithms as well as with differential evolution and genetic algorithms. Experimental verifications are carried out on three-phase and six-phase induction motors. Also, it emulates the closeness between experimental and estimated parameters with fast convergence compared to other algorithms. Also, the results reflect the high robustness of the proposed algorithm compared with other algorithms for varied iteration numbers, population size and convergence.
1. Introduction
The Poly-phase Induction Motors (PIMs) are the most used electrical machines [1]. They contribute around 60% of electric power converted to mechanical energy [2]. PIM are favored due to their ruggedness and simplicity in the industry section as 90% of industrial motors are IMs [3]. Examples of induction motors applications involve motor tools equipped with induction motors, adjustable speed motors and pumps [4]. To achieve the target performance of induction machines, the accurate modelling is considered as crucial issue for PIMs [5]. These issues involve the transient and steady-state behavior. The model expresses the stator and rotor windings voltage balance, flux linkages and currents, the air-gap power, and the electromagnetic torques. Therefore, finding the unknown parameters of these machines is a complicated nonlinear non-smooth optimization problem [6]. It aims at achieving the highest closeness degree between the estimated parameters and those of the actual ones. Therefore, the objective function of the considered parameter estimation problem is the minimum deviation between estimated and actual parameters with preserving these parameters within their permissible operating boundaries. To satisfy the parameter identification process, several optimization algorithms have been developed to guarantee the accurate PIM models. In this regard, this paper proposes the hybrid algorithm between particle swarm optimization and Jaya optimization algorithms HPJOA form finding the optimal unknown PIM parameters.
