A study of one-dimensional position control of a light iron ball levitated by one electromagnet is proposed in this work. The vertical position of the ball is the target of the proposed control strategy. It is assumed that the position measurement of the ball is inaccessible and thus, estimates of the position are utilized as the feedback signal to the controller. Control of the ball's position is carried out using LQG-based controller with the objective of manipulating the electromagnet current to drive the ball to steady state with minimal error. The integrity of the controller as well as accuracy of the system's model is evaluated via simulation.

The objective of this study is to develop and design fuzzy-based controllers for experimental examination and application to a laboratory scale sun tracking heliostat with dynamic movement about azimuth and elevation axes. The experimental approach accounts for unknown parameters such as, nonlinear static and dynamic frictions, nonlinear and variant effect of gravity on system, magnetic saturation of motors, limitations of power source in supplying rush and steady current and variation in heliostat dynamics due to different spacial and time passing conditions. To meet the objective, a classical PI and PID as well as Fuzzy-PI (F-PI) and Fuzzy-PID (F-PID) controllers are designed and experimentally implemented. The performance of each controller is measured by means of evaluating a cost function that is based on the integral of absolute value of error signal. The results show that for azimuth-axis angle, the cost of F-PI controller for deviation from set point is 67% lower as compared with that of PI controller. Also, it is shown that the application of F-PI controller results in lower cost for elevation-axis angle by 36%, 40%, and 50%, when compared with PI, PID, and F-PID controllers, respectively.

No abstract