Gain-scheduling Control For Unmanned Aerial Vehicle

Unmanned aerial vehicle (UAV) is developing to faster, it’s aerodynamic characteristics is more complex, and the performance is more superior. But, the system of UAV’s maneuver flight in the large airspace is a time-varying system with strong coupling and nonlinear. A controller with a constant parameter is difficult to guarantee the global system to obtain the desired performance; therefore, the gain-scheduling control law is the ideal solution for the problem. With the requirement improving for controllers, how to design a reasonable and effective control system is a problem which needs people to explore new ideas in engineering application and deepen the research in theory.Based on traditional gain-scheduling control thought, this paper summarizes the characteristics of gain-scheduling and the process that how to implement it. With the UAV’s longitudinal movement as the control object, two gain-scheduling controllers are studied.In the first place, the six degrees motion equations of the plane in space are built, under assumptions, the equations are linearized and decoupled to longitudinal and lateral motion equations.In the second place, taking airplane pitching angle control system an object, based on the traditional gain-scheduling control law, PID controllers are designed respectively at21feature points cover the global flight envelope using the classical control theory. Then the global controller in whole flight envelope is obtained by interpolation. The performance of the controller is verified by date simulation;And then, aim at the shortcoming in the design of the traditional gain-scheduling controller, the parameter dependent Lyapunov function theory is studied. The knowledge of parameter dependent model and polytope are used to transform the control system, then the aircraft longitudinal short-period gain-scheduling state feedback controller is solved by LMI. The theory is validated rationality effectively;Last, according to research results and existing problems, this paper puts forward some future possible research directions for gain-scheduling.