Dynamics Modeling and Pitching Parameters Identification of a Novel Hybrid UAV

摘要

Unmanned Aerial Vehicle (UAV) is an appealing topic for aeronauticalresearchers due to its tremendous application in the world. To make controllerdesign possible for a novel UAV design, dynamics modeling – a process ofderiving a set of differential equations governing the motion of the aircraft -is essential. In this paper, we present a conceptual approach to obtain theparameterized dynamics equation based on application of Newton’s secondlaw and approximations of aerodynamics effects on the UAV. Fromthere, twoidentification methods are introduced, one bases on maximum likelihoodwhile the other employs linear regression to estimate the aircraft’s dynamicparameters such as aerodynamic and control and stability derivatives throughan example involving our new hybrid UAV developed from fixed wingaircraft and a tricopter. Wind tunnel tests for a one-third scaled model arecarried out to receive outputs of the model, such as Euler angles and rotationrates, from prescribed input signals, which are rotors’ speeds, then pitchingparameters are identified. Estimated model would then be validated toanother set of experiment to show the fitness, hence remarks regarding theaccuracy of the dynamics model and the parameters themselves can be made.The articles show that for the derived mathematical model, the estimationresults were well fitted, and cross-validation also indicates that the model wasfine enough. The methods have strong implications about its generality thatis applicable to other novel vehicle designs.