Observed Control - Linearly Scalable Nonlinear Model Predictive Control with Adaptive Horizons

Abstract:This work highlights the duality between state estimation and model predictive control. A model predictive controller, observed control, is presented that uses this duality to efficiently compute control actions with linear prediction horizon length scalability. The proposed algorithms provide exceptional computational efficiency, adaptive time horizon lengths, and early optimization termination criteria. The use of Kalman smoothers as the backend optimization framework provides for a familiar implementation supported by strong theoretical guarantees. Additionally, a formulation is presented that separates linear model predictive control into purely reactive and anticipatory components, enabling any-time any-horizon observed control while ensuring controller stability for short time horizons. Finally, the method is extended to nonlinear systems and non-quadratic cost functions to obtain locally-optimal control of complex systems while maintaining linear prediction horizon scalability and adaptive-horizon capabilities.

16 pages, 8 figures. Submitted to IEEE Transactions on Automatic Control on 8/17/2025. Date of latest revision 7/29/2026

Optimization and Control (math.OC); Robotics (cs.RO); Systems and Control (eess.SY)

arXiv:2508.13339 [math.OC]

(or arXiv:2508.13339v2 [math.OC] for this version)

https://doi.org/10.48550/arXiv.2508.13339 Focus to learn more arXiv-issued DOI via DataCite

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.