Liang Xu
Home
Biography
Publication

Publication

Selected publications are listed below, grouped by research topic. For a complete list, please see my Google Scholar and ORCID pages.

Networked Control Systems

How does imperfect communication, such as channel fading, packet loss, delay and power constraints, limit what a networked control system can achieve? We characterized the channel capacity required for mean-square stabilization of linear systems over power-constrained fading and finite-state Markov channels; showed that packet drops make the number of components in the posterior of the optimal estimator grow exponentially, and designed tractable event-triggered estimators; and derived explicit consensusability conditions that couple agent dynamics, network topology and channel statistics. We also studied consensusability and distributed control of physically interconnected systems, and recently extended these results to output consensus of heterogeneous agents under Markov packet loss. Many of these results are collected in our Wiley-IEEE Press monograph.

  • Z. Zhang, L. Xu, X. Ren and X. Wang, “Distributed Output Consensus for Heterogeneous Multiagent Systems With Markov Packet Loss”, IEEE Transactions on Cybernetics, vol. 56, no. 7, pp. 4068-4081, 2026. [IEEE Xplore]

  • J. Zheng, L. Xu, Q. Hu and L. Xie, “Control over Communication Networks: Modeling, Analysis, and Design of Networked Control Systems and Multi-Agent Systems over Imperfect Communication Channels”, IEEE Press Series on Control Systems Theory and Applications, Wiley-IEEE Press, 2023. [Wiley]

  • M. Turan, L. Xu and G. Ferrari-Trecate, “On Consensusability of Linear Interconnected Multi-Agent Systems and Simultaneous Stabilization”, IEEE Transactions on Control of Network Systems, vol. 9, no. 1, pp. 403-414, 2022. [IEEE Xplore] [arXiv]

  • L. Xu, Y. Mo and L. Xie, “Remote State Estimation with Stochastic Event-triggered Sensor Schedule and Packet Drops”, IEEE Transactions on Automatic Control, vol. 65, no. 11, pp. 4981-4988, 2020. [IEEE Xplore]

  • L. Xu, B. Guo, C. Galimberti, M. Farina, R. Carli and G. Ferrari-Trecate, “Suboptimal Distributed LQR Design for Physically Coupled Systems”, the 21st IFAC World Congress, IFAC-PapersOnLine, vol. 53, no. 2, pp. 11032-11037, 2020. [ScienceDirect]

  • L. Xu, Y. Mo and L. Xie, “Distributed Consensus over Markovian Packet Loss Channels”, IEEE Transactions on Automatic Control, vol. 65, no. 1, pp. 279-286, 2020. [IEEE Xplore] [arXiv]

  • J. Xu, L. Xu, L. Xie and H. Zhang, “Decentralized Control for Linear Systems with Multiple Input Channels”, Science China Information Sciences, vol. 62, no. 5, 2019. [Springer]

  • J. Zheng, L. Xu, L. Xie and K. You, “Consensusability of Discrete-Time Multi-Agent Systems with Communication Delay and Packet Dropouts”, IEEE Transactions on Automatic Control, vol. 64, no. 3, pp. 1185-1192, 2019. [IEEE Xplore]

  • L. Xu, L. Xie and N. Xiao, “Mean Square Stabilization over Gaussian Finite-State Markov Channels”, IEEE Transactions on Control of Network Systems, vol. 5, no. 4, pp. 1830-1840, 2018. [IEEE Xplore]

  • L. Xu, J. Zheng, N. Xiao and L. Xie, “Mean Square Consensus of Multi-Agent Systems over Fading Networks with Directed Graphs”, Automatica, vol. 95, pp. 503-510, 2018. [ScienceDirect]

  • L. Xu, Y. Mo, L. Xie and N. Xiao, “Mean Square Stabilization of Linear Discrete-Time Systems over Power Constrained Fading Channels”, IEEE Transactions on Automatic Control, vol. 62, no. 12, pp. 6505-6512, 2017. [IEEE Xplore]

  • L. Xu, N. Xiao and L. Xie, “Consensusability of Discrete-Time Linear Multi-Agent Systems over Analog Fading Networks”, Automatica, vol. 71, pp. 292-299, 2016. [ScienceDirect]

  • L. Xu, Y. Mo and L. Xie, “Mean Square Stabilization of Vector LTI Systems over Power Constrained Lossy Channels”, the 2016 American Control Conference, pp. 7129-7134, 2016. [IEEE Xplore] [arXiv]

  • L. Xu, L. Xie and N. Xiao, “Mean Square Capacity of Power Constrained Fading Channels with Causal Encoders and Decoders”, the 54th IEEE Conference on Decision and Control, pp. 2465-2470, 2015. [IEEE Xplore] [arXiv]

Learning and Control

How can neural networks and data enter the control loop without giving up guarantees, and how can control theory make neural networks more reliable? Neural networks for control: structured neural controllers that are stable for any trainable parameters (for port-Hamiltonian, Lagrangian and multi-agent systems), learned neural contraction metrics and Lyapunov functions, and fast neural MPC with control barrier functions. Control for neural networks: Hamiltonian deep networks with provably non-vanishing gradients, contractive neural ODEs that are robust to adversarial inputs, and efficient verification of neural control barrier functions. Data-driven control and finite-sample analysis: robust controllers designed directly from input-output data, finite-sample guarantees for system identification and LQR design over lossy channels, global convergence of policy gradient under packet loss, and safety filters with provable performance bounds.

  • T. Tian, Z. Zhang and L. Xu, “Finite-Sample System Identification of Linear Systems Under Random Packet Loss”, IEEE Transactions on Control of Network Systems, vol. 13, no. 3, pp. 1284-1295, 2026. [IEEE Xplore]

  • Z. Zhang, T. Tian, X. Yi and L. Xu, “Global Convergence Analysis of Gauss-Newton Policy Gradient for LQR with Random Packet Loss”, the 20th IEEE International Conference on Control and Automation (ICCA), pp. 1918-1923, 2026. [IEEE Xplore]

  • X. Ren, Y. Ji, J. Qian, L. Xu and X. Wang, “KNODEOB-MPC: A Knowledge-Based Neural ODE With Online Observer MPC Framework for Quadrotor”, International Journal of Robust and Nonlinear Control, vol. 36, no. 8, pp. 4689-4704, 2026. [Wiley]

  • J. Zhang, H. Zhang, C. Liu, X. Wang and L. Xu, “Efficient Verification of Neural Control Barrier Functions with Smooth Nonlinear Activations”, arXiv preprint arXiv:2605.07757, 2026. [arXiv]

  • Z. Lin and L. Xu, “Curvature-Guided Safety Filters: State-Dependent Hessian-Weighted Projection with Provable Performance Bounds”, IEEE Control Systems Letters, vol. 10, pp. 139-144, 2026. [IEEE Xplore] [arXiv]

  • M. Zakwan, L. Xu and G. Ferrari-Trecate, “Robust Convolution Neural ODEs via Contractivity-promoting Regularization”, the 64th IEEE Conference on Decision and Control, pp. 8028-8033, 2025. [IEEE Xplore] [arXiv]

  • K. Wang, T. Li, L. Xu, Q. Hu and K. You, “Safety Meets Speed: Accelerated Neural MPC With Safety Guarantees and No Retraining”, IEEE Robotics and Automation Letters, vol. 10, no. 11, pp. 11411-11418, 2025. [IEEE Xplore] [arXiv]

  • J. Qian, L. Xu, X. Ren and X. Wang, “Distributed Neural Network-Based Control for Multi-Agent Lagrangian Systems with Stability Guarantees”, the 19th IEEE International Conference on Control and Automation (ICCA), pp. 391-396, 2025. [IEEE Xplore]

  • J. Qian, L. Xu, X. Ren and X. Wang, “Structured Deep Neural Network-Based Backstepping Trajectory Tracking Control for Lagrangian Systems”, IEEE Transactions on Neural Networks and Learning Systems, vol. 36, no. 6, pp. 11650-11656, 2025. [IEEE Xplore] [arXiv]

  • Z. Zhang, L. Xu, Y. Mo and X. Wang, “Learning Control for LQR with Unknown Packet Loss Rate Using Finite Channel Samples”, arXiv preprint arXiv:2501.02899, 2025. [arXiv]

  • M. Zakwan, L. Xu and G. Ferrari-Trecate, “Neural Exponential Stabilization of Control-affine Nonlinear Systems”, the 63rd IEEE Conference on Decision and Control, pp. 8602-8607, 2024. [IEEE Xplore] [arXiv]

  • H. Jiang, J. Zhang, L. Xu, X. Ren and X. Wang, “Neural Lyapunov Control for Interconnected Systems with Unknown Dynamics”, the 43rd Chinese Control Conference, pp. 5363-5368, 2024. [IEEE Xplore]

  • C. Galimberti, L. Furieri, L. Xu and G. Ferrari-Trecate, “Hamiltonian Deep Neural Networks Guaranteeing Non-vanishing Gradients by Design”, IEEE Transactions on Automatic Control, vol. 68, no. 5, pp. 3155-3162, 2023. [IEEE Xplore] [arXiv]

  • M. Zakwan, L. Xu and G. Ferrari-Trecate, “Robust Classification using Contractive Hamiltonian Neural ODEs”, IEEE Control Systems Letters, vol. 7, pp. 145-150, 2023. [IEEE Xplore] [arXiv]

  • L. Xu, M. Zakwan and G. Ferrari-Trecate, “Neural Energy Casimir Control for Port-Hamiltonian Systems”, the 61st IEEE Conference on Decision and Control, pp. 4053-4058, 2022. [IEEE Xplore] [arXiv]

  • L. Xu, B. Guo and G. Ferrari-Trecate, “Finite-sample-based Spectral Radius Estimation and Stabilizability Test for Networked Control Systems”, the European Control Conference, pp. 2087-2092, 2022. [IEEE Xplore] [arXiv]

  • L. Xu, M. Turan, B. Guo and G. Ferrari-Trecate, “Non-conservative Design of Robust Tracking Controllers Based on Input-output Data”, the 3rd Annual Learning for Dynamics and Control Conference, PMLR vol. 144, pp. 138-149, 2021. [PMLR] [arXiv]

  • C. Galimberti, L. Xu and G. Ferrari-Trecate, “A Unified Framework for Hamiltonian Deep Neural Networks”, the 3rd Annual Learning for Dynamics and Control Conference, PMLR vol. 144, pp. 275-286, 2021. [PMLR] [arXiv]

Robot Navigation and Manipulation

Our early work designed adaptive, output-feedback and robust controllers for robot manipulators and underactuated mechanical systems under kinematic, dynamic and actuator uncertainties. More recently we work towards embodied intelligence: robust visual odometry in dynamic and low-light scenes (BRIGHT-VO improves pose accuracy by about 25% in low light), and G-DRAGON, which grounds natural-language commands in OpenStreetMap with an on-board language model and completed person-search missions of up to 500 m on a real ground vehicle in unseen urban areas. For vision-language navigation, HAVANA is a dual-system video-action model in which a slow system produces a language-conditioned latent plan from the observation history and a fast system turns the latest plan and the current observation into local trajectories; the two run asynchronously, so no future video is generated at inference time, and the model achieves state-of-the-art results on OmniNavBench and generalizes to unseen real-world environments across different robots. For safe real-time motion planning, DS-MPPI splits the horizon of sampling-based model predictive control into a proximal segment refined by the cross-entropy method and a distal segment explored by MPPI soft weighting, and integrates control barrier functions for safety; in hardware-in-the-loop quadruped obstacle avoidance on flat, irregular and stair terrains it achieves 95% average task success and 96% average operational safety.

  • L. Xu, K. Wang, X. Fang, Z. Mao, Q. Hu and K. You, “DS-MPPI: Horizon-Segmented Safe Sampling for Real-Time Model Predictive Control”, IEEE Robotics and Automation Letters, early access, 2026. [IEEE Xplore]

  • D. Wang, Y. Du, C. Wang, X. Li and L. Xu, “HAVANA: A Hierarchical Video-Action Model for Vision-Language Navigation”, arXiv preprint arXiv:2610.24122, 2026. [arXiv]

  • D. Wang, Y. Du, J. Sun, Y. Xue, Y. Zhang, B. Xiao, C. Wang and L. Xu, “G-DRAGON: Geospatial Reasoning and Dynamic Planning for Retrieval-Augmented Outdoor Navigation”, IEEE Robotics and Automation Letters, vol. 11, no. 7, pp. 8768-8775, 2026. [IEEE Xplore] [arXiv]

  • D. Wang, Y. Yang, X. Chen and L. Xu, “BRIGHT-VO: Brightness-Guided Hybrid Transformer for Visual Odometry with Multi-modality Refinement Module”, the 34th International Joint Conference on Artificial Intelligence (IJCAI), pp. 1936-1944, 2025. [IJCAI] [arXiv]

  • H. Wen, X. Chen, L. Xu, X. Ren and X. Wang, “Semantic-Aware Monocular Point-Line Visual Odometry in Dynamic Environments”, the 2024 China Automation Congress, pp. 4329-4334, 2024. [IEEE Xplore]

  • D. Ling, L. Xu, X. Ren and X. Wang, “Dynamic Object Suppression in Visual Odometry via Adaptive Masked Flow Refinement”, Chinese Conference on Pattern Recognition and Computer Vision (PRCV), Lecture Notes in Computer Science, pp. 222-235, 2024. [Springer]

  • L. Xu, Q. Hu and Y. Zhang, “L2 Performance Control of Robot Manipulators with Kinematics, Dynamics and Actuator Uncertainties”, International Journal of Robust and Nonlinear Control, vol. 27, no. 6, pp. 875-893, 2017. [Wiley]

  • L. Xu and Q. Hu, “Output-feedback Stabilisation Control for a Class of Under-actuated Mechanical Systems”, IET Control Theory & Applications, vol. 7, no. 7, pp. 985-996, 2013. [IET]

  • Q. Hu, L. Xu and A. Zhang, “Adaptive Backstepping Trajectory Tracking Control of Robot Manipulator”, Journal of the Franklin Institute, vol. 349, no. 3, pp. 1087-1105, 2012. [ScienceDirect]

Ph.D. Thesis

  • L. Xu, “Control over Networks with Fading Communication Channels”, Doctoral thesis, Nanyang Technological University, Singapore, 2018. [DR-NTU] [pdf] [slides]