A New Approach to Tune PID for A Propeller Driven Pendulum Using PSO

Abstract

This project presents a method to tune the PID controller using Particle Swarm Optimization (PSO) for a propeller pendulum system. The propeller driven pendulum consists of a motorized fan connected to end of long plastic stick, the other end of the pendulum is pivoted and free to swing. The objective is to balance the pendulum at a given angle. The built system is interfaced to the PC using NI-ELVIS DAQ assistance and Labview system. The system transfer function is considered to be that of a standard 2nd order system and is obtained using the step response method. The problem of tuning the PID controller is formulated as an optimization problem and is solved using MATLAB simulation. Four time-based objective functions were constructed. Each objective function was optimized separately by employing different PSO to obtain the parameters of the four PID controllers for four time-based objective functions. The performance of four PID controllers was compared to each other in terms of peak overshoot, settling time, etc. The parameters of the simulated PID controllers were tested in real time for built propeller system.

Date of Award	Jan 2013
Original language	American English
Supervisor	Youssef Abdel-Magid (Supervisor)