Design and implementation of an optimal multiple-model adaptive controller for line-of-sight stabilization

Abstract

Vibration reduces the ability of a camera to detect far-away targets that lead to the use of stabilized platforms. Despite much research on stabilized platforms, the need for more accurate systems still exists as a consequence of sensor resolution improvements. In this article, a multiple-model adaptive controller is designed for a single-axis system. There are two important problems associated with multiple-model controllers: number of models and their locations in an operating space. After describing the problem, the stability of the control scheme is verified in theory. Next, the closed-loop stability condition is used to find the minimum number of models. After that, models are located in a space of an unknown parameter by defining a cost function for line-of-sight stabilization. The performance of the introduced multiple-model controller is evaluated in simulation and real-time implementation. Then it is compared with a proportional-plus-integral controller and an H∞-optimal controller. As a result, the stability accuracy of the test platform improves in both simulation and real-time implementation.