Supervisory Controller Design for an Image-Based Air Surveillance Platform by Using Hybrid Feedback Control Theory

Date

2023-5-31

Author

Akova, Hayrettin Ulaş

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

169
views

112
downloads

Cite This

High precision motion platforms are equipped with electro-optical sensors and image-based trackers to detect and aim accurately at a target in most of the air surveillance applications. Due to the limited field-of-view of electro-optical sensors, some knowledge of target angular position is required to drive the line-of-sight vector of the optical sensors to a close neighborhood of a target unless the target initially appears within the field-of-view of electro-optical sensors. In this thesis, a hybrid supervisory controller robustly uniting the initial target designation and image-based tracking controllers of an air surveillance platform is proposed to enlarge the basin-of-attraction to the entire operational airspace for this problem. A hybrid mathematical model of the supervisory controller is constructed with a logic variable keeping track of the controller being active, and the design of corresponding jump sets (switching logic) of the supervisory controller is presented to ensure the global asymptotic stability of the overall closed loop system. Following this approach, the design of each separate control loop having different feedback sensors, thus open-loop dynamics, and performance requirements is performed separately while exploiting the “divide-and-conquer” benefit provided by hybrid feedback control theory with its intrinsic robustness property. A mathematical model of the electro-optical platform is constructed to design the initial target designation and image-based tracking controllers. A rigid body model of the gimballed structure is obtained by using Newton-Euler equations. The kinematics of the image formation process is modelled by using a pin hole camera model whereas time-varying time-delayed error signals based on lower sampling rate images are represented by a hybrid dynamical model including a sampler and a delay timer, and a memory variable. A simulation model of the overall system is also constructed in MATLAB®/Simulink®. Through the application of the hybrid Lyapunov theorem and the sector nonlinearity technique, the stability requirements for the hybrid image-based tracking loop are derived. The resulting linear matrix inequalities (LMIs) are resolved to obtain a stability plot, and as a result, the controller gains, making the image-based tracking controller globally pre-asymptotically stable. A full state feedback linearizing controller with a time optimal path generator is designed as the target designation controller. Additionally, the hybrid supervisory logic design makes use of the simulation model to guarantee that the maximal solutions are complete. The performance of the overall hybrid closed loop system is demonstrated by experiments on an operational electro-optical platform. By designing some benchmark controllers and performing simulations, it is shown that missing the dominant hybrid nature of the overall system causes some solutions being not complete and undesired switching.

Subject Keywords

Air surveillance, Hybrid feedback control theory, Supervisory control, Sample-and-hold, Time-varying time delay

URI

https://hdl.handle.net/11511/104221

Collections

Graduate School of Natural and Applied Sciences, Thesis