Design, modeling and control of a reconfigurable serial quadrotor for agile aerial manipulation and navigation through narrow gaps

Abstract This paper presents the design, modeling and control of a reconfigurable serial quadrotor (RSQ) with experimental validation. The proposed platform features servo-driven joints between its arms, enabling both ground and in-flight morphological reconfiguration. This capability allows the RSQ to adapt its geometry for navigation in confined environments and for aerial manipulation tasks without additional robotic arms. A comprehensive nonlinear dynamic model is developed to capture the effects of structural reconfiguration, explicitly accounting for time-varying inertia properties, their time derivatives, control allocation variations, and centre of gravity (CoG) shifts. These parameters are derived analytically and validated using high-fidelity computer-aided design (CAD) models across multiple configurations, enabling online model updating for simulation and control purposes. To address the strong dynamic coupling introduced by reconfiguration, a dedicated control architecture based on a proportional integral derivative (PID) controller is proposed. Closed-loop stability is analysed by casting the PID controller into a static output feedback form within a linear parameter varying (LPV) framework. The synthesis conditions are formulated as linear matrix inequalities (LMIs), and the resulting constrained optimisation problem is solved using the whale optimisation algorithm (WOA) to guarantee uniform exponential stability of the closed-loop system. The proposed approach is validated through nonlinear simulations and experimental tests on a custom-built test bench for multiple configurations. The results demonstrate accurate trajectory tracking and stable behaviour during structural changes, highlighting the potential of the RSQ for reconfigurable aerial robotics and manipulation applications.