Preprint
Article

This version is not peer-reviewed.

A Practical Youla Gain-Tuning Framework for H Controller Realization

Submitted:

10 June 2026

Posted:

12 June 2026

You are already at the latest version

Abstract
Standard H controller synthesis produces robust controllers with well-shaped sensitivity and complementary sensitivity transfer functions, S and T. However, at times H does not enforce strict requirements on sensitivity, in particular the desired requirement that T has unity gain at DC frequency. This results in typically negligible steady-state tracking error, as the H optimization produces T(0)≈1. In drive cycle applications where reference velocity profiles contain extended ramp segments, this negligible deviation is integrated over time into a growing, non-negligible bias. The conventional remedy is to augment the plant with an integrator prior to synthesis, but this increases the order of the plant model and can be inconvenient when the control designer’s modeling has already been completed. This paper presents a post-synthesis gain adjustment method using Youla parameterization that corrects the DC tracking deficiency without modifying the plant or repeating H synthesis. The poles and zeros corresponding to the H controller’s Youla transfer function Y are preserved, with a free parameter K replacing the gain of Y. Re-calculating the controller after solving for the value of K that enforces T(0)=1 results in a hybrid controller that retains the robustness of the original but with improved performance in ramp-input scenarios with minimal effort for the control designer. Simulation results on a vehicle speed tracking problem confirm elimination of accumulating bias while preserving robustness margins from the original H design.
Keywords: 
;  ;  ;  
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
Prerpints.org logo

Preprints.org is a free preprint server supported by MDPI in Basel, Switzerland.

Subscribe

Disclaimer

Terms of Use

Privacy Policy

Privacy Settings

© 2026 MDPI (Basel, Switzerland) unless otherwise stated