Data-Driven Iterative Learning LQG Control of Axial-Gap 5-DOF Self-Bearing Motor

This paper studies the displacement control of the tilt-controlling axial-gap self-bearing motor with single stator. We first work out with the state space description for the motor. To design the control strategy, we adopt Linear-Quadratic-Gaussian (LQG) control to design the balance controller. It is known to us that the selection of matrices Q and R are always determined autonomously, which in terms results to less than perfect control performance in real applications. To overcome this situation, we propose an off-line data-based optimizer to replace the matrices selection process by using fruit fly optimization algorithm (FOA). Finally, the effectiveness of our results are verified through the comparison with LQG by simulation experiments.