# MetaTune: Adjoint-based Meta-tuning via Robotic Differentiable Dynamics

Source: [MetaTune: Adjoint-based Meta-tuning via Robotic Differentiable Dynamics](<https://www.sota2.com/research/paper/69cb54fb0636e5bf4afc9e48>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-07-04T05:12:22.129Z

Authors: Xiexin Peng, Bingheng Wang, Tao Zhang, Que Dong, Ying Zheng

Published: 2026-03-28T15:37:39.000Z

## Abstract

Disturbance observer-based control has shown promise in robustifying robotic systems against uncertainties. However, tuning such systems remains challenging due to the strong coupling between controller gains and observer parameters. In this work, we propose MetaTune, a unified framework for joint auto-tuning of feedback controllers and disturbance observers through differentiable closed-loop meta-learning. MetaTune integrates a portable neural policy with physics-informed gradients derived from differentiable system dynamics, enabling adaptive gains across tasks and operating conditions. We develop an adjoint method that efficiently computes the meta-gradients with respect to adaptive gains backward in time to directly minimize the cost-to-go. Compared to existing forward methods, our approach reduces the computational complexity to be linear in the data horizon. On quadrotor control tasks, MetaTune achieves competitive or improved tracking performance while reducing gradient computation time by more than 50\\%. In PX4-Gazebo hardware-in-the-loop simulation, the learned policy transfers zero-shot and reduces tracking RMSE by about 15--20\\% in aggressive flight and up to 40\\% under strong disturbances.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Controller and Observer Gain Optimization on Differentiable Simulation JAX-based (train)](<https://www.sota2.com/research/sota/controller-and-observer-gain-optimization-on-differentiable-simulation-jax-based-train.md>): Controller and Observer Gain Optimization; Differentiable Simulation JAX-based (train); RMSE: 0.67.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2603.27313>)

[PDF](<https://arxiv.org/pdf/2603.27313.pdf>)
