نوع مقاله : مقاله پژوهشی
نویسنده
گروه مهندسی برق و کامپیوتر، دانشکده مهندسی برق، دانشگاه صنعتی اصفهان، اصفهان، ایران
کلیدواژهها
عنوان مقاله English
نویسنده English
Introduction: In many unmanned aerial missions, driving the tracking error to zero by an instant fixed in advance is an operational requirement rather than a convenience: the vehicle must be stabilised before it reaches a moving landing platform or leaves a hazard zone. Asymptotic control offers no such guarantee, and conventional predefined-time designs rely on time-varying gains that diverge as the settling instant is approached, which causes actuator saturation, amplification of measurement noise near the end of the interval, and the need to truncate or replace the control law at that instant.
Methods: This work presents a robust control scheme based on bounded periodic delayed feedback that stabilises quadrotor altitude and attitude simultaneously. Because all four controlled subsystems are second-order strict-feedback chains, the law is designed in two backstepping steps with a delay equal to one quarter of the predefined time, and four explicit control laws are obtained. The gain remains bounded throughout, switches itself off at the predefined instant, and requires no external switching. Beyond convergence, boundedness of all closed-loop signals, a pre-flight verifiable attitude-cone condition, an extension to unknown mass and unknown moments of inertia, and a bound on the residual band under sampling and input delay are proved
کلیدواژهها English
