PSO BASED TUNING OF PI CONTROLLER FOR A DUAL STAR INDUCTION MACHINE FED BY A FIVE-LEVEL INVERTER

Tuning methods for PID controllers are very important for the process industries. Traditional methods such as Ziegler-Nichols, pole placement, metho²xds often do not provide adequate tuning. In this connection, our proposed method is using Particle Swarm Optimization (PSO) to optimize PI-controller parameters; the PSO algorithm is used off-line to update the PI parameters for controlling the speed of the double star induction machine, the machine is fed by a five-level inverter. Analytical solutions of Pulse Width Modulation (PWM) strategies for multilevel Neutral Point Clamped (NPC) are presented. Digital simulation results demonstrate that the deigned PSO-PI speed controller realize a good dynamic of the DSIM, a perfect speed tracking with no overshoot, give better performance and high robustness.