Attitude Control Method Research Of3-DOF Helicopter Model System

Since the first biplane named "The Aviator" completed the first powered flight of humanhistory, people have never stopped exploring the mystical sky. With the progress anddevelopment of science and technology, people’s demands on the aircraft are no longer limitedto just flying. How to realize the real-time control of the aircraft begins to be scheduled onmore and more people’s agenda.The helicopter model is a favorable carrier for researching the real-time control of theaircraft and has been widely accepted by most researchers in the automatic control field. Thetypicality of the physical model is also very rare in the teaching and research of the controlscience. The research object of this paper is a3-DOF helicopter model experimental system,which is designed by the famous motion control equipment manufacturers, Quanser Company,from Canada. All of our research like mathematical modeling and controller designing isbased on it.Through the analysis of the3-DOF helicopter model experimental system, we can clearlyfind it’s nonlinear, strong coupling, multi-variable, multi-input and output characteristics,which make our research more difficult. Finally we chose the classic control theory PIDmethod and the optimal control theory LQR (Linear Quadratic Regulator, LQR) method todesign controllers. Then we introduced multi-model switching theory and designed LQRmulti-model controller.During the process of designing a controller, firstly, we build a mathematical model ofthe helicopter, which is based on the deep analysis of its hardware system and physicalcharacteristics. And then we modify the model parameters regarding to the different needs ofthe subsequent controller. For example, approximate and linearize the mathematical model forPID controller, transform three groups of differential equation of motion into state spaceequation for LQR controller; establish multi-model group based on the variables of high angleand pitch angle for multi-model LQR controller and so on. Secondly, we do the simulation ofall kinds of controller we designed based on Matlab. In order to measure the quality of allthese controllers, we use them to track the step signal, sine signal and step signal, through theobservation of the tracking accuracy of elevation angle, pitch angle and rotation angle, the size of overshoot and speed of tracking, we can find the control effect of different types ofcontroller directly. After achieving a satisfactory outcome, we apply the best multi-modelLQR controller to the physical model of helicopter by making using of the Real-TimeWorkshop, RTW from Matlab and the control software Wincon from Quanser Company,which is a special experiment platform of motion control. Finally, the simulation resultsdemonstrate a good effect of this controller.