Parameter Identification of DC Motor Drive Systems using Particle Swarm Optimization

Abstract

Accurate parameter estimation is of increasing importance in system modeling and control of highperformance motor drive systems. The inaccurate calculation or assumption of the system modal parameters may cause instability and/or bias control performance. This work presents the mathematical modeling, simulation, and experimental study to estimate the mechanical parameters of a two-mass-model system in DC motor drives using Artificial Swarm Intelligence. The electrical parameters such as resistance and inductance are usually provided by the DC motor manufacturer. However, the mechanical parameters such as the moment of inertia, and viscous friction will vary when the motor is connected to a mechanical load for a given application. Computer modeling and parameter estimation of the system were carried out using MATLAB/Simulink software. The parameters estimation of the mechanical side of a DC Motor drive system will be converted into an optimization problem using an objective function. The standard PSO algorithm is compared to the proposed modified PSO algorithms in the literature.