Integrated Modeling and Robust Control for Full-Envelope Flight of Robotic Helicopters

Abstract

Robotic helicopters have attracted a great deal of interest from the university, the industry, and the military world. They are versatile machines and there is a large number of important missions that they could accomplish. Nonetheless, there are only a handful of documented examples of robotic-helicopter applications in real-world scenarios. This situation is mainly due to the poor flight performance that can be achieved and--more important--guaranteed under automatic control. Given the maturity of control theory, and given the large body of knowledge in helicopter dynamics, it seems that the lack of success in flying high-performance controllers for robotic helicopters, especially by academic groups and by small industries, has nothing to do with helicopters or control theory as such. The problem lies instead in the large amount of time and resources needed to synthesize, test, and implement new control systems with the approach normally followed in the aeronautical industry. This thesis attempts to provide a solution by presenting a modeling and control framework that minimizes the time, cost, and both human and physical resources necessary to design high-performance flight controllers. The work is divided in two main parts. The first consists of the development of a modeling technique that allows the designer to obtain a high-fidelity model adequate for both real-time simulation and controller design, with few flight, ground, and wind-tunnel tests and a modest level of complexity in the dynamic equations. The second consists of the exploitation of the predictive capabilities of the model and of the robust stability and performance guarantees of the [Special characters omitted.] loop-shaping control theory to reduce the number of iterations of the design/simulated-evaluation/flight-test-evaluation procedure. The effectiveness of this strategy is demonstrated by designing and flight testing a wide-envelope high-performance controller for the Carnegie Mellon University robotic helicopter.