TACO: Trajectory-Aware Controller Optimization for Quadrotors

2025-11-03

AI Summary

Fixed controller parameters in quadcopter trajectory tracking hinder task adaptability and limit tracking accuracy and dynamic feasibility. To address this, we propose TACO (Trajectory-Aware Controller Optimization), a novel framework that jointly integrates a learned trajectory prediction model with a lightweight online optimization mechanism. TACO enables real-time, adaptive tuning of controller gains based on reference trajectory characteristics and the vehicle’s current state, while supporting online trajectory replanning under smoothness and dynamic feasibility constraints to enhance response robustness. Efficient training data generation and model learning are enabled via a parallelized high-fidelity simulator. Experiments demonstrate that TACO significantly reduces tracking error across diverse complex trajectories, outperforming conventional manual tuning approaches. Moreover, its optimization speed is orders of magnitude faster than black-box methods, exhibiting strong potential for real-time deployment.

Technology Category

• Search and Optimization: Learning to Search
• Constraint Satisfaction and Optimization: Constraint Learning and Acquisition
• Intelligent Robots: Learning & Optimization for ROB

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Controller performance in quadrotor trajectory tracking depends heavily on parameter tuning, yet standard approaches often rely on fixed, manually tuned parameters that sacrifice task-specific performance. We present Trajectory-Aware Controller Optimization (TACO), a framework that adapts controller parameters online based on the upcoming reference trajectory and current quadrotor state. TACO employs a learned predictive model and a lightweight optimization scheme to optimize controller gains in real time with respect to a broad class of trajectories, and can also be used to adapt trajectories to improve dynamic feasibility while respecting smoothness constraints. To enable large-scale training, we also introduce a parallelized quadrotor simulator supporting fast data collection on diverse trajectories. Experiments on a variety of trajectory types show that TACO outperforms conventional, static parameter tuning while operating orders of magnitude faster than black-box optimization baselines, enabling practical real-time deployment on a physical quadrotor. Furthermore, we show that adapting trajectories using TACO significantly reduces the tracking error obtained by the quadrotor.

Problem

• Optimizing quadrotor controller parameters for trajectory tracking performance
• Adapting controller parameters online based on trajectory and quadrotor state
• Improving dynamic feasibility of trajectories while maintaining smoothness constraints

Innovation

• Online adaptive controller parameters optimization
• Learned predictive model with lightweight optimization
• Parallelized simulator for large-scale trajectory training