Real-time Koopman data-driven model predictive control on field-programmable gate arrays: a deterministic control framework for unknown robotic manipulator systems

This paper presents an field-programmable gate array-based implementation of Koopman data-driven model predictive control. The proposed method is applied to robotic manipulators with unknown or partially known nonlinear dynamics. Specifically, a high-dimensional control model for the nonlinear system is established using the extended dynamic mode decomposition method based on input–output data. To implement this, the lifting computation and model predictive control solving are implemented using hardware description languages on Xilinx Kintex-7 XC7K325T operating at 200 MHz, achieving 18.12% look-up table utilization and 3.33% digital signal processor deployment with robust timing closure (4.091 ns critical path, 0.383 ns worst-case negative slack). Finally, end-effector trajectory-tracking experiments were performed on the UR5e robotic manipulator, validating the effectiveness of the controller in real-time robotic manipulator control with deterministic performance.

Authors

Institutions

Publication Details

Journal
Robotica
Published
2026-11-19
DOI
https://doi.org/10.1017/s026357472610397x
Primary Topic
Advanced Control Systems Optimization
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Real-time Koopman data-driven model predictive control on field-programmable gate arrays: a deterministic control framework for unknown robotic manipulator systems

Zongsheng Geng, Xingwen Zhou, Shi Yan, Dongdong Zhao et al.
Robotica
Advanced Control Systems Optimization
article

Real-time Koopman data-driven model predictive control on field-programmable gate arrays: a deterministic control framework for unknown robotic manipulator systems

Zongsheng Geng, Xingwen Zhou, Shi Yan, Dongdong Zhao, Mingming Tang
article en

Abstract

This paper presents an field-programmable gate array-based implementation of Koopman data-driven model predictive control. The proposed method is applied to robotic manipulators with unknown or partially known nonlinear dynamics. Specifically, a high-dimensional control model for the nonlinear system is established using the extended dynamic mode decomposition method based on input–output data. To implement this, the lifting computation and model predictive control solving are implemented using hardware description languages on Xilinx Kintex-7 XC7K325T operating at 200 MHz, achieving 18.12% look-up table utilization and 3.33% digital signal processor deployment with robust timing closure (4.091 ns critical path, 0.383 ns worst-case negative slack). Finally, end-effector trajectory-tracking experiments were performed on the UR5e robotic manipulator, validating the effectiveness of the controller in real-time robotic manipulator control with deterministic performance.

Robotica
Lanzhou University (CN)
Advanced Control Systems Optimization
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon