Tianhua Gao

A Sliced Learning Framework for Online Disturbance Identification in Quadrotor SO(3) Attitude Control

Fri, 06 Feb 2026 00:00:00 +0000

This work introduces a dimension-decomposed geometric learning framework called Sliced Learning for disturbance identification in quadrotor geometric attitude control. Instead of conventional learning-from-states, this framework adopts a learning-from-error strategy by using the Lie-algebraic error representation as the input feature, enabling axis-wise space decomposition (“slicing”) while preserving the SO(3) structure. This is highly consistent with the geometric mechanism of cognitive control observed in neuroscience, where neural systems organize adaptive representations within structured subspaces to enable cognitive flexibility and efficiency. Based on this framework, we develop a lightweight and structurally interpretable Sliced Adaptive-Neuro Mapping (SANM) module. The high-dimensional mapping for online identification is axially “sliced” into multiple low-dimensional submappings ( “slices”), implemented by shallow neural networks and adaptive laws. These neural networks and adaptive laws are updated online via Lyapunov-based adaptation within their respective shared subspaces. To enhance interpretability, we prove exponential convergence despite time-varying disturbances and inertia uncertainties. To our knowledge, Sliced Learning is among the first frameworks to demonstrate lightweight online neural adaptation at 400 Hz on resource-constrained microcontroller units (MCUs), such as STM32, with real-world experimental validation.

Introduction Video:

For testbed experiment video:

SANM

Sun, 01 Feb 2026 00:00:00 +0000

SANM-augmented Geometric Control Work in Progress. This project is currently under active development.

Online Identification using Adaptive Laws and Neural Networks for Multi-Quadrotor Centralized Transportation System

Thu, 13 Nov 2025 00:00:00 +0000

This paper presents an online adaptive–neuro identification framework to enhance the robustness of centralized multi-quadrotor cable-suspended payload transportation systems. The proposed method addresses the challenges arising from highly coupled nonlinear dynamics, unknown payload parameters, and time-varying disturbances by introducing a dimension-decomposed learning strategy.

The core idea is to decompose the high-dimensional payload error space into multiple low-dimensional subspaces, referred to as “slices.” Within each slice, shallow neural networks and adaptive laws operate in parallel to identify model uncertainties and disturbance dynamics online. This structure, termed Sliced Adaptive-Neuro Mapping (SANM), enables efficient real-time learning without requiring persistent excitation or offline training, while preserving theoretical transparency.

The SANM module is integrated as a feedforward compensator into a payload-centric geometric control architecture. By learning directly from Lie-algebra–based error representations, the proposed framework avoids explicit modeling of complex cable–quadrotor couplings and improves disturbance rejection for both translational and rotational payload dynamics. Stability of the overall closed-loop system is rigorously analyzed using Lyapunov theory, and uniform ultimate boundedness of tracking and estimation errors is established under unknown payload parameters and time-varying disturbances.

Numerical simulations demonstrate that the proposed method significantly enhances tracking performance and robustness compared with conventional centralized control approaches. The results indicate that dimension-decomposed adaptive–neuro learning provides an effective, lightweight, and theoretically grounded solution for robust aerial cooperative transportation.

Experience

Sat, 25 Oct 2025 00:00:00 +0000

Robustness Enhancement for Multi-Quadrotor Centralized Transportation System via Online Tuning and Learning

Fri, 22 Aug 2025 00:00:00 +0000

This work presents an adaptive–neuro geometric control framework for a centralized multi-quadrotor cooperative transportation system carrying a cable-suspended payload. The main objective is to enhance system robustness against both parametric uncertainties (e.g., unknown payload mass and inertia) and external disturbances (e.g., wind forces and moments), which are common challenges in real-world aerial transportation tasks.

Building upon an existing centralized geometric control architecture, the proposed approach augments the first-level payload control signals with online tuning and learning mechanisms. Specifically, multiple radial basis function (RBF) neural networks are employed to approximate unknown disturbance dynamics in both translational and rotational subsystems, while adaptive laws are simultaneously introduced to estimate uncertain payload parameters. These two mechanisms interact cooperatively: adaptive parameter tuning globally scales the controller behavior, while neural networks locally compensate for time-varying disturbances.

A key feature of the proposed method is that no pre-training or persistent excitation (PE) condition is required. All neural network weights and model parameters are updated online using Lyapunov-based adaptation laws, ensuring bounded estimation errors. The authors provide a rigorous Lyapunov stability analysis, showing that the closed-loop system achieves semi-global practical stability under bounded disturbances and modeling uncertainties. When reference model parameters are matched and static attitude tracking is considered, stronger stability guarantees can be obtained.

The effectiveness of the proposed control strategy is validated through numerical simulations of a three-quadrotor transportation system. Comparative studies demonstrate that, relative to conventional geometric control, the proposed adaptive–neuro approach significantly improves robustness in scenarios involving large payload mass mismatch, inertia uncertainty, and strong external disturbance forces or moments.

Overall, this work contributes a lightweight, theoretically grounded, and practically implementable solution for robust centralized aerial transportation, with clear potential for extension to real-world experimental validation and more comprehensive stability analysis in future work.

American Control Conference (ACC) 2025

Tue, 08 Jul 2025 13:00:00 +0000

Numerical Simulation of a Novel Cargo Handling Strategy: Using a Centralized Cable-Linked Dual-Multirotor System

Fri, 23 Aug 2024 00:00:00 +0000

This paper investigates a novel cargo handling strategy for aerial transportation using a centralized cable-linked dual-multirotor system. The work focuses on enabling autonomous loading, transportation, and unloading of cargo without relying on onboard mechanical manipulators, which are often inefficient and complex for medium- and long-distance UAV logistics.

To achieve this goal, the authors propose a Tug-of-War (ToW) method, in which two multirotors pull against each other through a shared cable to establish internal tension and maintain cable rigidity during loading and unloading. By equipping the cargo with two passive hooks, the system can autonomously dock, lift, and transport payloads of arbitrary shape and weight while preserving a safe distance between the multirotors and mitigating cable sagging effects.

From a modeling perspective, the system is formulated using a rigid cable-linked representation on nonlinear configuration manifolds, where the horizontal cable segment is treated as the centralized controlled object, and the hooks, suspension segments, and attached cargo are modeled as disturbances. Building on this formulation, a hybrid control strategy is developed by integrating geometric control with a model reference adaptive nonlinear model predictive control (MRA-NMPC). This combination enables adaptive regulation of cable geometry (gamma angles) and robust control performance under unknown payload parameters and model mismatches.

Numerical simulations of both loading and transportation processes demonstrate the feasibility of the proposed ToW method and validate the adaptability and robustness of the hybrid controller in the presence of disturbances and uncertainty. The results indicate that the proposed strategy effectively bridges the gap between centralized and decentralized cable-suspended transportation methods, providing a flexible and scalable solution for autonomous aerial cargo handling.