Model Predictive Controlled Novel Supermaneuverable Tricopter Design

This paper presents the design, modeling, and control of a novel supermaneuverable tricopter (SMT) featuring dual-axis thrust vectoring on each of its three rotors. Unlike conventional tricopters, which face under-actuation and yaw-pitch coupling, the SMT offers over-actuation with nine independent control inputs for six degrees of freedom. This allows complete decoupling of motion and enables agile maneuvers in complex environments, which is essential for modern UAV applications. The control system is based on a Model Predictive Controller (MPC), formulated to exploit the SMT's over-actuation by optimizing control inputs over a finite prediction horizon while satisfying state and input constraints. The controller uses a linear discrete model to forecast behavior and generate real-time optimal trajectories. Simulation results demonstrate that the MPC effectively tracks complex reference trajectories with high precision, without violating actuator or dynamic constraints. The proposed implicit MPC implementation demands only modest computational resources, making it deployable on embedded systems without the need for conversion to explicit MPC. The combination of minimal processing overhead, trajectory tracking accuracy, and constraint adherence confirms the practicality of the approach. Keywords: Decoupled, MPC, Novel, Over-actuation, Supermaneuverable, Tricopter.