Model predictive control-based trajectory generation for agile landing of unmanned aerial vehicle on a moving boat

TL;DR

This work presents an MPC-based trajectory generator for agile UAV landing on a moving USV deck in harsh sea states, leveraging predicted USV states and a dynamically weighted objective to synchronize UAV attitude during touchdown. By modeling the USV with a 6 DOF framework and incorporating Gerstner-wave dynamics, the method accounts for deck tilt and wave-driven motion, while FAA modulates the attitude emphasis as the UAV approaches the deck. The approach demonstrates substantial improvements over the state of the art in landing precision and speed, achieving near-100% success across moderate to rough seas in simulations and confirming robustness with real-world experiments on artificial waves. Overall, the framework enables reliable, fast, autonomous UAV landings on dynamic maritime platforms using onboard computation and phase-aware trajectory planning.

Abstract

This paper proposes a novel trajectory generation method based on Model Predictive Control (MPC) for agile landing of an Unmanned Aerial Vehicle (UAV) onto an Unmanned Surface Vehicle (USV)'s deck in harsh conditions. The trajectory generation exploits the state predictions of the USV to create periodically updated trajectories for a multirotor UAV to precisely land on the deck of a moving USV even in cases where the deck's inclination is continuously changing. We use an MPC-based scheme to create trajectories that consider both the UAV dynamics and the predicted states of the USV up to the first derivative of position and orientation. Compared to existing approaches, our method dynamically modifies the penalization matrices to precisely follow the corresponding states with respect to the flight phase. Especially during the landing maneuver, the UAV synchronizes attitude with the USV's, allowing for fast landing on a tilted deck. Simulations show the method's reliability in various sea conditions up to Rough sea (wave height 4 m), outperforming state-of-the-art methods in landing speed and accuracy, with twice the precision on average. Finally, real-world experiments validate the simulation results, demonstrating robust landings on a moving USV, while all computations are performed in real-time onboard the UAV.

Figures (15)

• Figure 1: Motivation (a) manual landing of the UAV in the real world DailyPicksandFlicks, (b) autonomous landing of the UAV in simulation in the high wave environment, and (c) photo of the UAV together with the USV research platform deployed in the real world.
• Figure 2: Illustration of the UAV's and USV's coordinate systems, where $\mathcal{W}$ represents local ENU frame, $\mathcal{C}$ denotes the multirotor's body-fixed frame, and $\mathcal{B}$ symbolizes the USV's body-fixed frame. The rotation matrix from the world frame to the USV’s body-fixed frame is represented as $\mathbf{R}_{w}^{b}$, and the translation vector is denoted as $\mathbf{t}_{w}^{b}$. The rotation matrix $\mathbf{R}_{w}^{c}$ and translation vector $\mathbf{t}_{w}^{c}$ denotes rotation and translation from the world frame to the UAV’s body-fixed frame. The angular velocity of the UAV and USV is denoted with $(p, q, r)$ with the appropriate subscripts.
• Figure 3: Illustration of the developed state machine composed of three phases ensures decision-making throughout the whole process.
• Figure 4: Visualization of the landing maneuver in Rviz together with the trajectory generator pipeline diagram.
• Figure 5: Full system pipeline diagram including USV's estimation/prediction, mission planning, trajectory planning, and the Multi-robot Systems multirotor control system baca2021mrs. Mission and navigation software supplies the flight mode ($m$) to the Trajectory generator where the reference trajectory composed of headings and positions ($\mathbf{r}_d$, $h_d$) is generated. Trajectory tracking controller ensures the trajectory is precisely tracked by producing the desired thrust and angular velocities ($T_d$, $\bm{\omega}_d$) for the Pixhawk embedded flight controller. The State estimator fuses data from Onboard sensors and Odometry & localization methods to create an estimate of the UAV translation and rotation ($\mathbf{x}$, $\mathbf{R}$). Estimated states ($\mathbf{e}$) and a full-states prediction ($\mathbf{p}_d$) of the USV are provided by the State estimator and States prediction software, which exploits the Onboard sensors of the UAV and USV.