Stochastic Redesign of Small UAV over Its Control Surfaces and Autonomous Flight System
摘要
The main purpose of this article is to enhance autonomous flight performance of a small unmanned aerial vehicle (SUAV) by stochastically redesigning its control surfaces (i.e., elevator and aileron) and autonomous flight system (AFS). An SUAV is manufactured in Erciyes University Drone Laboratory (ERUDL) and named as Erciyes-Qtar-SUAV. Its control surfaces can be varied before flight. AFS parameters (gains of relevant PID controllers) and control surfaces’ parameters are stochastically redesigned to minimize autonomous flight performance index by using stochastic optimization approaches. The obtained results are used in a simulation environment of SUAV autonomous flight. Using stochastic and simultaneous design approach for an SUAV, autonomous flight performance is maximized. For flight tests of SUAVs, authorization of the Directorate General of Civil Aviation in Republic of Turkey is required. For improvement of autonomous flight performance of an SUAV, composing a creative solution and also a creative algorithm for use of stochastic redesign on SUAV’s control surfaces and its AFS is the first essential contribution of this article. In addition, application of simultaneous perturbation stochastic approximation (i.e., SPSA) for the mentioned aim is another significant contribution. These important contributions reason as well much less energy consumption and green environment.