Efficient and Robust Time-Optimal Trajectory Planning and Control for Agile Quadrotor Flight

Ziyu Zhou, Gang Wang, Jian Sun, Jikai Wang, Jie Chen

Abstract

Agile quadrotor flight relies on rapidly planning and accurately tracking time-optimal trajectories, a technology critical to their application in the wild. However, the computational burden of computing time-optimal trajectories based on the full quadrotor dynamics (typically on the order of minutes or even hours) can hinder its ability to respond quickly to changing scenarios. Additionally, modeling errors and external disturbances can lead to deviations from the desired trajectory during tracking in real time. This letter proposes a novel approach to computing time-optimal trajectories, by fixing the nodes with waypoint constraints and adopting separate sampling intervals for trajectories between waypoints, which significantly accelerates trajectory planning. Furthermore, the planned paths are tracked via a time-adaptive model predictive control scheme whose allocated tracking time can be adaptively adjusted on-the-fly, therefore enhancing the tracking accuracy and robustness. We evaluate our approach through simulations and experimentally validate its performance in dynamic waypoint scenarios for time-optimal trajectory replanning and trajectory tracking.

Introduction

Agile flight, such as drone racing, requires quadrotors to operate at the edge of their physical capabilities. This necessitates the use of time-optimal trajectories—paths that minimize travel time while respecting complex, underactuated dynamics. Historically, this has been a computationally intensive task, often requiring minutes or hours of offline calculation, which precludes real-time adaptation to dynamic environments. Furthermore, high-speed tracking is inherently brittle; small modeling errors or external disturbances can cause standard controllers to fail as they attempt to adhere to a rigid, pre-calculated schedule.

Methodology

This work introduces a two-layer framework designed for both computational efficiency and operational robustness.

Segmented Time-Optimal Planning

The primary innovation in the planning layer is a "divide and conquer" strategy. Rather than treating the time of waypoint passage as a free variable—a common source of complexity in existing methods—the proposed approach fixes the waypoint nodes. The trajectory is divided into segments between these waypoints, with a set number of discrete points assigned to each. The sampling interval ($dt_i$) for each segment is then treated as an optimization variable. This formulation allows the problem to be solved in seconds using standard interior-point methods while maintaining near-optimal performance.

The segmented planning strategy fixes waypoint nodes and optimizes sampling intervals ($dt_i$) for each segment.

Time-adaptive Model Predictive Control (tMPC)

To address tracking fragility, the authors developed a time-adaptive Model Predictive Control (tMPC) scheme. Standard MPC attempts to minimize the distance to a reference point at a fixed time. In contrast, tMPC includes the starting time of the reference trajectory ($t_0$) as an optimization variable in each control loop. This temporal flexibility allows the quadrotor to effectively "adjust its clock," slowing down or speeding up the reference progress based on real-time performance and disturbances, thereby significantly increasing robustness.

Experimental Results

The framework was validated through extensive simulations and real-world experiments using a quadrotor equipped with an Intel NUC11 and tracked by an OptiTrack system.

Planning Efficiency

The segmented planning approach achieved a massive speedup. For an 8-waypoint course, it solved the planning problem in approximately 1.7 seconds, compared to 2,350 seconds for state-of-the-art Complementary Progress Constraints (CPC) methods. In dynamic scenarios, the system successfully replanned trajectories in just 0.12 seconds when waypoints were moved mid-flight.

High-Speed Tracking and Robustness

In physical trials, the quadrotor reached a maximum speed of 10.6 m/s with a tracking error of less than 0.22m. Simulation results under model mismatch conditions (e.g., altered mass or drag) demonstrated that tMPC maintains stability and accuracy where standard MPC configurations fail or deviate significantly.

Comparison of tracking performance: tMPC (orange) maintains the planned path (dashed green) more accurately than standard MPC (blue) when disturbances occur.

Conclusion

The proposed two-layer framework enables agile quadrotor flight that is both fast to plan and robust to track. By decoupling the spatial and temporal constraints in planning and introducing temporal flexibility in control, the system bridges the gap between theoretical optimality and real-world applicability. Future research will explore the inclusion of yaw constraints, transition to onboard vision-based localization, and the integration of general obstacle avoidance.

Summary updated

Abstract

Agile quadrotor flight relies on rapidly planning and accurately tracking time-optimal trajectories, a technology critical to their application in the wild. However, the computational burden of computing time-optimal trajectories based on the full quadrotor dynamics (typically on the order of minutes or even hours) can hinder its ability to respond quickly to changing scenarios. Additionally, modeling errors and external disturbances can lead to deviations from the desired trajectory during tracking in real time. This letter proposes a novel approach to computing time-optimal trajectories, by fixing the nodes with waypoint constraints and adopting separate sampling intervals for trajectories between waypoints, which significantly accelerates trajectory planning. Furthermore, the planned paths are tracked via a time-adaptive model predictive control scheme whose allocated tracking time can be adaptively adjusted on-the-fly, therefore enhancing the tracking accuracy and robustness. We evaluate our approach through simulations and experimentally validate its performance in dynamic waypoint scenarios for time-optimal trajectory replanning and trajectory tracking.
BibTeX
@article{zhou2023efficient,
  author = {Ziyu Zhou and Gang Wang and Jian Sun and Jikai Wang and Jie Chen},
  title = {{Efficient and Robust Time-Optimal Trajectory Planning and Control for Agile Quadrotor Flight}},
  journal = {IEEE Robotics Autom. Lett.},
  year = {2023},
  url = {https://openreview.net/forum?id=73PnizjkEO},
}
Figure