A Simulation Evaluation Suite for Robust Adaptive Quadcopter Control

TL;DR

This work tackles the challenge of reproducibly evaluating robust adaptive quadcopter controllers under disturbances and model uncertainty by introducing an open-source, RotorPy-based simulation testbed. It offers a modular control library with both adaptive and non-adaptive controllers, a configurable disturbance model suite, and task-specific metrics for tracking and robustness. The framework supports systematic benchmarking across wind, payload, rotor faults, latency, and uncertainties, including automated extreme-stress tests to identify failure points. This enables fair comparisons and accelerated development of robust quadcopter control methods with practical relevance for real-world applications.

Abstract

Robust adaptive control methods are essential for maintaining quadcopter performance under external disturbances and model uncertainties. However, fragmented evaluations across tasks, simulators, and implementations hinder systematic comparison of these methods. This paper introduces an easy-to-deploy, modular simulation testbed for quadcopter control, built on RotorPy, that enables evaluation under a wide range of disturbances such as wind, payload shifts, rotor faults, and control latency. The framework includes a library of representative adaptive and non-adaptive controllers and provides task-relevant metrics to assess tracking accuracy and robustness. The unified modular environment enables reproducible evaluation across control methods and eliminates redundant reimplementation of components such as disturbance models, trajectory generators, and analysis tools. We illustrate the testbed's versatility through examples spanning multiple disturbance scenarios and trajectory types, including automated stress testing, to demonstrate its utility for systematic analysis. Code is available at https://github.com/Dz298/AdaptiveQuadBench.

Figures (9)

• Figure 1: Overview of our simulation testbed. The framework consists of a dynamics simulator based on RotorPyfolk2023rotorpy, a modular control interface supporting both adaptive and non-adaptive controllers, and evaluation tools for trajectory generation, disturbance injection, performance metrics, and visualization. Users interact with the system through a unified configuration interface that specifies controller selection, disturbance type, trajectory, and evaluation mode.
• Figure 2: Illustration of the quadcopter's $x$-configuration motor layout in the body frame. The directions of rotation for each rotor ($\omega_1$ to $\omega_4$) and the geometric angle $\beta$ between the rotor arms and the body axis $\bm{x}_B$ are indicated.
• Figure 3: Position tracking of l1geo and geo-a along a circular trajectory in the presence of gusty wind, generated using the Dryden turbulence model.
• Figure 4: Position tracking of geo, l1geo, and xadap under a circular trajectory with an off-center payload randomly attached mid-flight. The payload introduces coupled effects on center-of-mass, inertia, and external torque. The deviation from the desired trajectory before payload attachment arises from the transient controller response as the vehicle transitions from hovering to trajectory tracking.
• Figure 5: Position tracking and motor speed response of the l1mpc controller on a randomized trajectory under rotor effectiveness variations. Each rotor’s effectiveness factor is perturbed, simulating actuator-level faults.