Nonlinear Control Of Quadrotor Based On Backstepping

The quad-rotor, as a branch of rotor-typed unmanned aerial vehicle, is highlyfavored in both fields of military and civilian utilizations with its characteristics ofvertical take-off and landing, insensibility to environment varying, high mobilityand stability, and easy operation modes. The quad-rotor system is highly-coupled,under-actuated and non-linear inherently, which poses a great challenge to themicro-processor, the sensor, the mechanism, and the navigation and controlalgorithm in terms of system stability. Therefore, the research on the quad-rotor issurely meaningful and it provides the development of the fore-mentioned fields withan efficient platform.The accurate6-DOF dynamical equations are first established taking intoconsideration the disturbances such as the aerodynamic disturbance, gyro torquedisturbance, and gyro misalignment angle, based on which the quad-rotor’s traits ofcoupling and under-actuating are then analyzed. The analysis, together with theactual signal flow in application, helps to construct the dual-loop structure andfour-channel structure of the system. In the dual-loop structure, the attitudes arecontrolled in the inner loop while the positions the outer one; in the four-channelstructure, the channels are height channel, yaw channel, pitch channel, and rollchannel, respectively.Then, a detailed discussion is carried out on the basic idea, derivation process,scope of application, and controller’s structure and feature of the backsteppingmethod. This method is applied on the6-DOF dynamical equations to work out theposition and attitude controllers on the foundation of the dual-loop structure.Simulations under circumstances of spot hover, constant velocity tracking, andtrajectory tracking demonstrate the obtained controllers’ ability to guarantee thesystem stability. However, the controllers fail to restrain the perturbation nicely.Furthermore, to overcome the fore-mentioned backstepping controllers’drawback in constant perturbation restraint, they are optimized considering theresponse properties of the system. Through a thorough simulation, the optimizedcontrollers are proven to be reasonable and effective.Finally, an adaptive quad-rotor control system is designed under the premise ofmass uncertainty and disturbance uncertainty by integrating backstepping methodwith adaptive law. The simulation results show that the obtained controller andadaptive law can not only guaranty system stability but also estimate the accuratevalues of mass and disturbance, hence the effectiveness of the design procedure.