New identification of induction machine parameters with a meta-heuristic algorithm based on least squares method

Abstract

Purpose

The great value of offline identification of machine parameters is when the machine manufacturer does not provide its parameters. Most machine control strategies require parameter values, and some circumstances in the industrial sector only require offline identification. This paper aims to present a new offline method for estimating induction motor parameters based on least squares and a salp swarm algorithm (SSA).

Design/methodology/approach

The central concept is to use the classic least squares (LS) method to acquire the majority of induction machine (IM) constant parameters, followed by the SSA method to obtain all parameters and minimize errors.

Findings

The obtained results showed that the LS method gives good results in simulation based on the assumption that the measurements are noise-free. However, unlike in simulations, the LS method is unable to accurately identify the machine’s parameters during the experimental test. On the contrary, the SSA method proves higher efficiency and more precision for IM parameter estimation in both simulations and experimental tests.

Originality/value

After performing a primary identification using the technique of least squares, the initial intention of this study was to apply the SSA for the purpose of identifying all of the machine’s parameters and minimizing errors. These two approaches use the same measurement from a simple running test of an IM, and they offer a quick processing time. Therefore, this combined offline strategy provides a reliable model based on the identified parameters.

Citation

Zorig, A., Belkheiri, A., Bendjedia, B., Kouzi, K. and Belkheiri, M. (2023), "New identification of induction machine parameters with a meta-heuristic algorithm based on least squares method", COMPEL - The international journal for computation and mathematics in electrical and electronic engineering, Vol. 42 No. 6, pp. 1852-1866. https://doi.org/10.1108/COMPEL-01-2023-0051