CS Paper Radar

Daily listing

2026-11-17 · eess.SY

39 papers first submitted this day (UTC) in eess.SY — Systems and Control. arXiv announces a day's submissions the next evening, so the newest day fills up over ~24 h.

Score = weighted mean ofCitations 20%Upvotes 10%Code stars 10%Venue 15%Code/data 10%Novelty 15%Momentum 15%Substance 5%how it works

AI & Machine Learning 5

Learning Regions of Attraction for Transverse Feedback Linearization under Controller Singularities

cs.AI eess.SY ·2026-11-17 ·Abdullah Waqar Malik, Hamza Tariq, Adeel Akhtar
code/data
Abstract · 27 pages (10 main + references + appendix), 1 figure, 8 tables

Transverse feedback linearization (TFL) achieves path following by exactly linearizing the error dynamics, but its controller is undefined on a singular set, such as zero speed for a unicycle or zero thrust for a quadrotor. The constrained region of attraction, the set of initial states from which the closed loop converges to the path without violating state or regularity constraints, determines where the controller can be safely activated. Exploiting the exact linear error dynamics, we show that this region is the largest forward invariant subset of the admissible set. We then decompose its state-constrained Zubov value function into an analytically known quadratic cost-to-go and a safety value function. We learn only the latter, by fitted value iteration without infinite-horizon trajectory labels. The resulting estimate combines an inner layer verified by interval branch and bound with a learned outer layer tested on samples. On a unicycle following a circle and a sinusoid and on a planar quadrotor, it covers on average 72.0%, 83.4% and 19.1% of the simulated region of attraction. In comparison, the quadratic cost-to-go and two physics-informed neural network (PINN) solvers of Zubov's equation cover at most 11.3%, 4.0% and 0.0% on average. Code and animations are available at https://gradslab.github.io/roa_tfl

0.48

When does deep reinforcement learning pay off in HVAC supervisory control? Four small Modelica case studies trained directly on FMUs

cs.LG cs.AIeess.SY ·2026-11-17 ·Bingnan Zhao, Harshala Gammulle, Tharindu Fernando
code/data
Abstract · 16 pages, 4 figures, 2 tables. Code, models and trained agents: https://github.com/Bingnan-Zhao/hvac-drl-lab

Deep reinforcement learning (DRL) is often proposed for the supervisory control of heating, ventilation and air-conditioning (HVAC) systems. The reported gains are, however, difficult to interpret, because the baselines are often weak and the attainable optimum is seldom known. This paper examines four small, reproducible case studies based on an office building in Brisbane, Australia: pre-cooling under a time-of-use tariff; chiller staging with condenser water reset; chilled water supply temperature reset with a wet cooling coil; and chilled water storage under time-of-use prices and a demand charge. The plants were modelled with the Modelica Buildings library, and in three of the cases a soft actor-critic agent was trained directly on the exported functional mock-up unit (FMU). Each agent was compared with tuned rule-based controllers and, where practicable, with a near-optimal reference from hourly exhaustive search or dynamic programming with perfect foresight. In the two cases without storage the agent reached the near-optimal reference, but the margin above a well-chosen rule was small: 1.3% for condenser water reset and 0.18% for chilled water reset. In the latter case two common reset rules consumed more energy than a fixed set point. With a chilled water storage tank the margin was considerably larger. The agent outperformed both rules (-7.8%, compared with -3.9% and -6.3%) and adopted midday charging during the solar off-peak period, yet it obtained only about 45% of the saving achieved by the perfect-foresight schedule and did little to limit the demand peak. In the pre-cooling case a hand-tuned rule remained superior (-12.0% compared with -5.6%). These results suggest that studies of DRL for HVAC control should state the margin between the best rule and a near-optimal reference. Some practical difficulties encountered in training on co-simulation FMUs are also reported.

0.48

Deep Reinforcement Learning for Heat Pump and Underfloor Heating Control: An End-to-End Case Study

cs.AI eess.SY ·2026-11-17 ·Antonio Manjavacas, David Suárez-Caro, Alejandro Campoy-Nieves, Guillermo Rey, Gonzalo Satrústegui +1 more
Abstract · Preprint submited to journal

Efficient control of underfloor heating (UFH) systems coupled with heat pumps (HPs) is challenging due to the high thermal inertia of buildings and the slow dynamic response of UFH systems. Deep Reinforcement Learning (DRL) offers a data-driven control approach capable of simultaneously optimizing energy consumption and thermal comfort. However, the application of DRL-based controllers in occupied buildings under real operating conditions remains limited. This work presents and end-to-end case study (development, evaluation, and real-world deployment) of a DRL-based controller for an integrated UFH-HP system. First, a simulation-based pilot study was conducted to assess different DRL algorithms, control strategies, and configurations. Subsequently, a digital twin of a real residential building was developed and used to train DRL agents under simulated operating conditions. The best-performing agent was then deployed in the physical building without additional online fine-tuning. The results show that the controller successfully maintained indoor temperatures within the prescribed comfort limits while dynamically adjusting the HP supply temperature and UFH valve positions. These findings demonstrate the feasibility of deploying DRL controllers in occupied buildings and highlight their robustness and the practical challenges associated with transferring control policies from digital environments to real heating systems.

0.31

Residual-Guided Lightweight Reinforcement Learning for Risk-Aware Trajectory Selection with a Frozen World Model

cs.AI cs.ETcs.LGcs.RO+1 ·2026-11-17 ·Jiaqi Wu, Yongqi Dong, Yijie Zhong, Maple Shao, Yide Tao +4 more
Abstract · 6 pages, 2 figures

In automated driving, world models provide future scene predictions for evaluating candidate trajectories, but fixed selection rules may not capture scene-dependent trade-offs. We propose a residual-guided reinforcement learning (RL) framework for lightweight trajectory selection with a frozen world model and a fixed candidate bank. The framework uses proximal policy optimization (PPO) to fine-tune a joint policy over candidate selection and bounded trajectory residuals. A context-conditioned ranking network learns candidate preferences from feedback obtained by evaluating local perturbations of the selected trajectories. At inference, the selector combines learned preference and predicted risk ranks using fixed weights and returns the selected candidate trajectory without modification. Fine-tuning updates 16,106 parameters across the actor and critic, while the deployed selector retains only the 11,725-parameter ranking network, with the residual branch and critic used exclusively during training. In open-loop evaluation on the nuScenes validation set, the method reduces average Euclidean trajectory error (L2) by 3.99% relative to the RiskWorld model baseline without RL-based selection. Under matched budgets for feedback accesses and actor updates, residual exploration reduces L2 by 0.83% relative to the categorical PPO variant trained without residuals. The method also maintains lower L2 than RiskWorld when risk features are removed or shuffled.

0.26

GP--RL--MPC: Uncertainty-Aware Reinforcement Learning-Augmented Model Predictive Control with Gaussian Process Postdecision State-Value Functions

cs.LG eess.SY ·2026-11-17 ·Jon Li, Jay H. Lee
Abstract · 19 pages including references and appendices, 3 figures, 2 tables. Submitted to L4DC 2027

We propose Gaussian Process Reinforcement Learning-Augmented Model Predictive Control (GP--RL--MPC), which uses a learned critic as the MPC terminal penalty to account for costs beyond the prediction horizon. Gaussian process regression (GPR) enables data-efficient critic learning from limited, noisy rollouts within optimistic approximate policy iteration (OAPI), with soft updates mitigating policy chattering. A tunable risk coefficient weights GP epistemic uncertainty in the terminal cost, regulating sensitivity to uncertain critic predictions. The postdecision state formulation shifts disturbance averaging offline to account for aleatoric uncertainty from the terminal transition onward. Under matching objectives and admissible actions, a one-step horizon recovers greedy value-space RL. After fitting GP hyperparameters by exact marginal likelihood, we construct a sparse deterministic training conditional (DTC) posterior using inducing points selected by pivoted Cholesky (PC), reducing online evaluation costs as the training dataset grows. In a continuous stirred-tank reactor (CSTR) subject to exogenous disturbances, GP--RL--MPC with an appropriately selected risk coefficient and policy screening reduces the sum of squared errors (SSE) by an average of 80.52% relative to an otherwise matched baseline MPC across five random seeds (sample standard deviation: 3.48%).

0.25

Systems & Networking 1

Manufacturing Differential Sparsity: Planner-Hardware Co-Design for Efficient World Model Planning

cs.AR cs.AIcs.LGcs.RO+1 ·2026-11-17 ·Wenbo Zhu, Paul Yi-Chia Chen, Yuanhan Li, Zishen Wan, Christopher Torng
Abstract · 14 pages, 11 figures

JEPA-style latent world models let embodied agents plan toward new goals with model predictive control, but they are still too slow for real-life embodied control: sampling-based MPC repeats predictor inference tens of thousands of times per plan, and even on high-end GPUs one planning step can take tens of seconds. We observe substantial latent locality in this repeated-inference workload: the predictor maps close actions to close latents, and inside a quantized predictor this closeness takes a form that hardware can exploit, differential bit-level sparsity. Unlike an input stream, whose temporal similarity is given, a planner itself decides which candidates it evaluates together, so the planner can choose its own references and, with reference forest sampling, manufacture more sparsity: it couples the sampled action sequences of each group to a shared root, with competitive planning quality. We build the accelerator on the insights of the rich line of bit-level sparsity accelerators and design it and the planner tightly coupled as a single co-design: the planner increases sparsity and exposes structure for hardware scheduling, and the bit-exact hardware effectively uses the sparsity that the planner creates. On three JEPA world models, our co-design is 1.28-1.80x faster than a dense accelerator and uses 38-52% less energy, and it is 1.22-1.58x faster than prior bit-sparse design and uses 16-36% less energy, with the same area. To our knowledge, our planner is the first planner that manufactures differential bit-level sparsity across its candidates and exposes its structure for hardware scheduling, and our co-design is the first algorithm-hardware co-design for JEPA-style world-model planning.

0.17

Robotics, Control & Signals 33

Magflow: Differentiable Simulation of Industrial-Grade 6D Magnetic Levitation

cs.RO cs.LGeess.SY ·2026-11-17 ·Philipp Hartmann, Jannick Stranghöner, Klaus Neumann
top venuecode/data
Abstract · 8 pages, 5 figures, 1 table. Accepted for publication in IEEE Robotics and Automation Letters (RA-L). Project page: https://sites.google.com…

Magnetic levitation with six-degrees-of-freedom enables frictionless transport and micrometer-scale positioning in principle, but low-level control is inherently challenging due to its complex and unstable dynamics. Industrial systems typically rely on deep hardware-software co-design, where mover and stator geometries are jointly designed to enable a compact, hand-crafted approximation of the electromagnetic force computation, thereby keeping control computationally tractable at kilohertz rates. However, this approach imposes overly conservative constraints on the operating workspace and the range of supported mover topologies. Instead of reducing the underlying physics to a hand-crafted approximation, we compute the physics directly. We introduce Magflow, a fully differentiable, GPU-parallel simulator for 6D magnetic levitation. It is used to train the first neural controller for an industrial-grade 6D magnetic levitation system that is learned purely from simulated physics. We show that the learned neural controller transfers zero-shot from simulation to real hardware while satisfying hard real-time constraints at 4 kHz. The learned controller achieves qualitatively new behaviors and even levitates a mover topology for which no hand-crafted solution exists. These results demonstrate that co-design is not a physical necessity for 6D magnetic levitation control. More broadly, control learning from simulated physics can find solutions beyond the reach of control engineering in demanding industrial domains. Source code and demonstration videos are publicly available at https://sites.google.com/view/magflow.

0.59

COAST: Safe Flight Corridor Planner for Fast and Robust Snap Minimized Trajectories

cs.RO eess.SY ·2026-11-17 ·Alejandro Pelaez, Randal W. Beard
code/data
Abstract · 14 pages, 4 figures, 2 tables. Submitted to the 2027 American Control Conference (ACC 2027)

Current B-Spline trajectory planners face a critical trade-off: they either sacrifice computational speed to maintain strict Safe Flight Corridor (SFC) guarantees, or they abandon hard safety bounds in favor of soft-constraint penalties to achieve real-time performance. We present COAST (Corridor-Optimized Aerially-Safe Trajectories), a B-Spline based, online trajectory planner capable of formulating snap minimized trajectories with SFC guarantees. COAST divides the problem into two stages: an A* based front-end that generates geometrically simple SFCs (represented as a chain of rectangular prisms), and a quadratic programming back-end for control point optimization. To handle the non-convexity of SFC intersections, we leverage the finite support property of B-Splines, to ensure the spline never exits SFC boundaries. The control point optimization incorporates MINVO transforms for kinodynamic bounding, allowing COAST to design non-conservative, dynamically feasible trajectories, while a snap minimization matrix calculation provides flight smoothness. This formulation provides a formal geometric containment guarantee, while ensuring mean trajectory computation times under 5 ms for real-time deployment. We validate our results through high-fidelity simulation experiments. The source code can be found at: https://github.com/byu-magicc/Minimum-Snap-Trajectories

0.48

From Temperature-Limited to Battery-Limited Legged Locomotion

cs.RO cs.LGeess.SY ·2026-11-17 ·Raphaël Ségas, Balázs Opra, Léo Ghafari, Thomas Stewart
hot direction
Abstract · 9 pages, 7 figures, 2 tables. Supplementary video in the ancillary files. Dataset: https://doi.org/10.5281/zenodo.24274994

On legged robots, sustained locomotion often ends in thermal shutdown rather than battery depletion: the heat of walking concentrates in a few joint motors, which overheat long before the battery is empty. We train thermally aware locomotion policies by reinforcement learning without ever observing temperature, and demonstrate this on a Unitree Go2 quadruped. We first identify a coupled 13-node lumped RC thermal network, fitted on two runs of a single robot and calibrated against the measured electrical power, that predicts all twelve motor temperatures with a mean RMSE below 1 $^\circ$C across different robot instances, gaits, payloads, and speeds. Under a sustained gait, the temperatures the motors will settle at follow directly from the gait-averaged squared torques and joint speeds. Our core contribution is a reward that drives down the hottest predicted asymptotes: thermal time constants exceed the credit-assignment horizon of reinforcement learning by orders of magnitude, but predicted asymptotes respond to torque and joint speed within a single gait cycle, so the policy learns to lower and spread heat with no temperature input. On the robot, the policy turns sustained locomotion from temperature-limited to battery-limited: with a 5 kg payload at 1 m/s, a learned baseline reaches thermal shutdown after 19 minutes (1.1 km), whereas our policy runs the battery down after 174 minutes (10.0 km) with its hottest motor 14 $^\circ$C below the shutdown threshold in a 26 $^\circ$C room; on an unloaded factory circuit with stairs and ramps, the baseline overheats after 25 laps (6.2 km), while our policy completes 47 (11.2 km) on one battery with 22 $^\circ$C of margin.

0.45

MARS: Feature-guided physics-informed deep learning for real-time battery health and lifetime prediction

eess.SY ·2026-11-17 ·Innocent Izabayo, Haiquan Zhao
published
Abstract · 14 pages, 10 figures. Published in IEEE Transactions on Transportation Electrification, 2026

Accurate and robust estimation of battery state-of-health (SOH) and remaining useful life (RUL) is essential for safe and cost-effective operation of electrified transportation systems, yet it remains challenging under operating variability, non-stationary aging, and non-Gaussian measurement disturbances. This paper proposes Multi-scale Adaptive Robust State Estimation (MARS), a multi-scale adaptive robust state-estimation framework that couples a dual-branch temporal encoder with an equivalent-circuit-model-constrained differentiable generalized-correntropy adaptive Kalman filter. The encoder disentangles fast intra-cycle voltage-current-temperature dynamics from slow inter-cycle degradation, and the fused degradation context adaptively parameterizes the cycle-level process covariance, measurement covariance, and correntropy bandwidth. Through the unrolled robust filtering recursion, MARS performs physics-consistent state propagation and generalized-correntropy-based innovation correction, while uncertainty decomposition and Monte Carlo dropout provide calibrated prediction intervals for SOH estimation and conditional RUL prediction. Experiments on the NASA and CALCE benchmarks demonstrate improved SOH accuracy, achieving mean absolute error/root mean square error values of 0.72%/0.92% and 0.63%/0.80%, respectively, together with competitive conditional RUL performance against representative data-driven and hybrid baselines. The resulting conditional-RUL prediction intervals achieve coverage probabilities of 100.00% and 99.24% on NASA and CALCE, respectively. Ablation, disturbance-sensitivity, near-end-of-life, and runtime analyses further verify the contribution of each module and the practical inference cost for online battery management.

0.44

FM-Mamba: An Efficient Native 2D Selective State Space Model with Stability-Aware Discretization

eess.SY cs.AIcs.CV ·2026-11-17 ·Shengyuan Hu, Yiyong Sun, Weihang Shan, Yongzhe Chang
code/data
Abstract

When applied to two-dimensional features such as images, Mamba-based methods either flatten the feature map into 1D sequences and fuse multiple scans, which breaks spatial adjacency and multiplies computation, or recurse directly on the 2D grid but rescale the recurrence with a fixed input-independent factor and aggregate only the predecessors on a single anti-diagonal, leaving the discrete step unconstrained by 2D stability theory. We propose \textbf{FM-Mamba}, a native 2D selective state space model built on the general Fornasini--Marchesini (FM) model. Its coupled recurrence evolves along both spatial axes; cross-axis coupling induces a discrete diagonal-predecessor term $\tilde A_0$, and two independent input-dependent step sizes $Δ_1(i,j)$, $Δ_2(i,j)$ govern propagation along the two axes, so that the coefficients adapt per position and per direction. A normalized ratio $r_k=a_{0,k}/(a_{1,k}a_{2,k})$ yields a closed-form piecewise bound on the two step sizes; when a configuration falls outside the stable interval the steps are projected onto this bound, and the criterion is re-verified per position after training, which suppresses systematic local amplification without global damping. The recurrence is executed by a dual-diagonal wavefront scan: each update gathers information from two anti-diagonals, and the input enters the current state through a single projection, so that neither training nor inference flattens the feature map. Code is available at https://github.com/BitSsm/FM-Mamba-Runtime

0.43

AI Agents Executing an RMS and EMT Ride-Through Workflow on a 2,000-Bus Synthetic Texas Grid: A Case Study with Open-Source Tools

eess.SY ·2026-11-17 ·Tam Kemabonta, Brighton Ombuki, Vladimir Abdelnour
code/data
Abstract · 5 pages, 1 figure, 4 tables. Submitted to the 2027 IEEE PES General Meeting

This paper reports a case study of engineer-supervised AI agents executing a predefined RMS and EMT ride-through workflow on the 2,000-bus ACTIVSg2000 synthetic Texas grid with open-source tools. The test case is a 133 MW battery with a co-located 130 MW data-center load, assessed against ride-through criteria fixed before simulation and for loss of synchronism elsewhere in the grid. One agent ran 44 RMS cases and 15 without-project comparisons in ANDES; the other ran a 220-row EMT intake bench and an 84-row system study in DPsim and ParaEMT on a Kron-reduced area. Across the 44 RMS cases, 127 criterion cells fail, including loss of synchronism in all 44. Two findings concern the project plant. The battery keeps absorbing reactive power during sags on the light-load schedule (B3, as low as -19.8 Mvar on S4 T1), and the data-center load returns only at 5.00-5.56 s after long bolted faults (L3), which misses the ERCOT load-return requirement. In 13 of the 15 comparisons, which the authors selected after slips or speed-gate stops, the same machines slip without the project; the other 29 runs have no without-project comparison. RMS and EMT values differ in network, plant model, controls, operating point, and base, and so they are reported as separate, unaligned observations. The paper reports the errors the authors identified in the agent logs, 63 author messages, and 3.8 days from agent start to the integrated study report. The agents executed and documented the predefined ride-through and stability workflow, which covers one part of an interconnection study.

0.42

Stability-Constrained Imitation Learning for Performance-Guaranteed Neural Feedback Frequency Control in Microgrids

eess.SY ·2026-11-17 ·Sajid Kamal, Yichen Zhang, Feng Qiu, Tianqiao Zhao
Abstract · 10 pages

Guaranteeing performance for frequency control is challenging within traditional control paradigms, as it demands precise disturbance estimation and real-time control signal computation within a timeframe of seconds. This paper addresses this challenge by proposing a stability-certified imitation learning framework for frequency regulation in low-inertia microgrids. Based on the frequency response model of the microgrid, an offline numerical optimal control is employed to generate near-optimal pairs of state and control dataset, which are subsequently used to train a deep neural network (DNN) feedback controller. To ensure the closed-loop stability, the DNN controller is trained with a set of Lyapunov-based stability constraints derived using absolute stability theory. Such an imitation learning paradigm can significantly improve the training efficiency, while the resulting DNN controller acts as an end-to-end function that simultaneously handles disturbance estimation and control optimization. The proposed framework is validated on a modified IEEE 33-node diesel-wind microgrid. The DNN controllers are deployed in wind turbine generators (WTGs) to provide decentralized, intelligent supplementary active power support to always maintain frequency deviations within predefined safety bounds under different disturbances while preserving closed-loop stability.

0.40

CRA: Continuous Risk Assessment for Safety-critical Visual Navigation

cs.RO cs.CVeess.SY ·2026-11-17 ·Lihao Zheng, Ying Wu, Chen Jiang, Randy Freeman
code/data
Abstract · Project page: https://isaac-restaurant-cra.cra-demos.workers.dev

Traditional safety-critical control methods often rely on predefined constraints that separate states into safe and unsafe regions. While recent studies have extended geometric safety toward semantic safety, real-world safety is rarely purely binary. The risk associated with an object or human can vary continuously with its physical properties, like motion, and interaction context; for example, a person carrying a heavy object or a hot object moving at high speed may require a different safety response from an ordinary static obstacle. In this paper, we propose a vision-based continuous risk assessment framework for safety-critical visual navigation. Our method uses RGB and LiDAR observations to estimate four risk factors (kinetic, potential, contact-severity, and thermal) and combines them into a continuous risk score for each assessed object or person. These estimates are integrated into a Poisson safety function with a Laplace guidance field as a safety filter, enabling online monitoring and intervention in open-world robot navigation. We demonstrate the proposed method in a simulated restaurant scene.

0.33

Capability Composition in Heterogeneous Teams via Barrier-Guided Diffusion Sampling

cs.RO cs.LGcs.MAeess.SY ·2026-11-17 ·Diana Laura Morales, Ruoyu Lin, Magnus Egerstedt, Fabio Pasqualetti
theory
Abstract

The agents in a heterogeneous team can cooperate to accomplish tasks that exceed their individual capabilities. The team trajectories for a cooperative task can be generated by a generative model, provided that the model has been trained on demonstrations of the agents cooperating. Yet, demonstrations of cooperation are costly and may not exist for tasks that the team has never performed. The conditions under which the agents can combine their capabilities, however, are physical properties of the team and are known before any demonstration is collected. We encode the conditions as constraints on the team trajectories, and we enforce the constraints while sampling from generative models trained on demonstrations of each agent alone. We prove that the proposed method can generate every team trajectory that satisfies the constraints, although the demonstrations contain no cooperation. Hence the constraints determine which team trajectories can be generated, and the demonstrations determine how likely each team trajectory is. Numerical results on two cooperative tasks show the effectiveness of the method, which generates cooperative team trajectories that satisfy the constraints and requires no retraining when agents join or leave the team. Finally, we implement the method on a multi-robot testbed.

0.33

Satisfiability Modulo Theories for Robust Synthesis of Signal Temporal Logic Specifications

eess.SY cs.RO ·2026-11-17 ·Philip Smith, Kevin Leahy
Abstract · Submitted to American Control Conference

Signal Temporal Logic is a popular framework for describing a robotic system's goals. Once described using a temporal logic framework, synthesis turns the goals into actual steps for the robots to follow. Currently, the state of the art for sound and complete synthesis is using Mixed-Integer Linear Programming (MILP), which struggles on long time horizons and large environments, preventing the construction of guaranteed safe plans for long deployments. This paper introduces a new synthesis method using Satisfiability Modulo Theories (SMT) which scales better for long deployments and adapts to environmental changes more quickly. We do this by introducing an encoding method for STL into SMT constraints, and an encoding method for discrete-time dynamical systems modeled as a transition system. We then conduct an empirical comparison of our SMT based methods against the state of the art MILP based methods. Results show that for this class of problems, SMT is an effective encoding strategy that performs the best on long time horizon tasks.

0.33

Sim2CBF: From Forward Simulations to Implicit Control Barrier Function Safety Filtering without Model Derivatives

cs.RO eess.SY ·2026-11-17 ·Luzia Knoedler, Yihan Liu, Shaohang Han, Yifei Dong, Jana Tumova
Abstract · 9 pages, 6 figures

Failure avoidance in complex robot-environment interactions is challenging because the underlying dynamics are difficult to model accurately, and available predictive models may not provide reliable derivatives. We present Sim2CBF, a failure-avoidance framework that solves the Control Barrier Function (CBF)-Safety Filter (SF) optimization problem using only forward-model evaluations, thereby avoiding differentiation of the predictive model. Specifically, we approximate the CBF constraint using finite differences and local linear interpolation over sampled control inputs while preserving failure-avoidance guarantees. The underlying CBF is implicit and constructed from backup-policy rollouts, enabling the use of expressive predictive models such as physics-based simulators and learned world models. Compared with purely sampling-based SFs, our formulation requires fewer control samples and yields less conservative behavior than direct nominal-backup blending. We evaluate Sim2CBF in simulation, hardware box-toppling, and vision-based control tasks, demonstrating its applicability to contact-rich and learned dynamics, as well as its adaptability to changes in the predictive model.

0.33

A Discrete Digital Isolator with Shielded Planar Transformer for Gate Driver Applications

eess.SY ·2026-11-17 ·Thomas A. Conway
Abstract · 5 pages, 11 figures

Galvanically isolated gate drivers are required in many power electronics applications. This paper describes an implementation of a digital isolator using discrete components with a planar coreless PCB signal transformer providing the isolation. The use of an electrostatic shield between the transformer windings is shown to provide sufficient common-mode transient immunity to support IGBT half-bridge drivers, an application subject to a severe dV/dt transient during switching. The design is verified through the experimental implementation of a prototype half bridge using two of the proposed discrete digital isolators with SiC power transistors and demonstrates common-mode transient immunity greater than 15 kV/us with very low implementation complexity.

0.32

Classical heat exchanger fouling-resistance laws as first-moment identities of additive stochastic deposition

eess.SY ·2026-11-17 ·Simone Casolo
unusual
Abstract · Submitted

Heat-exchanger fouling is monitored through the fouling resistance, whose record is routinely fitted with the classical linear (Ebert--Panchal) or asymptotic (Kern--Seaton) law, and the shape of the fit is often taken as evidence of the underlying deposition or removal mechanism. We show that it cannot be. If the deposit grows by non-negative, statistically stationary increments, the mean fouling resistance is exactly linear without removal and exactly the Kern--Seaton asymptote under first-order removal, whatever the microscopic deposition law. Shear-driven erosion, per-element detachment, intermittent sloughing and saturation of the surface all give the same mean curve, so a fit identifies a net deposition rate and a relaxation time, and nothing else. Modelling the deposit as a Levy subordinator, we show where the missing information lies and what it costs to retrieve it from monitoring data. Removal mechanisms separate in the variance: sloughing leaves a plateau scatter that persists even for smooth deposition and can be read from a record that saturates. Deposition laws separate only in higher cumulants, which the measurement noise of a thermally reconstructed resistance attenuates; the record needed to distinguish them grows as the fourth power of the conductance error, which places it beyond reach at realistic measurement accuracy. For the time to reach a cleaning limit, the deposition law matters only near the threshold; farther away, the mean rate and its uncertainty dominate the prediction.

0.31

Connections Between Instrumental Variable, Pseudo-Linear Regression, and Prediction Error Methods in Terms of Their Minimized Criteria

eess.SY ·2026-11-17 ·Bernard Vau, Henri Bourlès
Abstract

This paper explores the connections between the criteria directly or indirectly minimized by Prediction Error (PE), Instrumental Variable (IV), and Pseudo-Linear Regression (PLR) algorithms for the identification of output-error models. By definition, PE methods aim to minimize the variance of the prediction error. In contrast, the correlation functions appearing in the stationarity conditions of IV and PLR algorithms are not the gradients of any underlying loss function involving a signal variance. Nevertheless, if the true system lies in the model set, we show that one IV algorithm minimizes asymptotically the variance of an unmeasurable signal called the equivalent prediction error. This signal was introduced previously to analyze the bias distribution of PLR schemes. It turns out that, regardless of whether the true system lies in the model set, some iterative IV and PLR algorithms aim to minimize the prediction error variance and can therefore be interpreted as indirect PE algorithms.

0.30

Finite sample guarantees of MPPI: suboptimality, sample complexity and closed-loop guarantees

eess.SY math.OC ·2026-11-17 ·Changyi Lei, Dragan Nesic, Ye Pu
theory
Abstract

This paper considers model predictive path integral (MPPI) control for nonlinear finite-horizon optimal control problems with bounded inputs, allowing nonconvex and nonsmooth costs. We prove theoretical guarantees that quantify the suboptimality caused by positive temperature and finite sampling on MPPI and establish their implications for closed-loop guarantees. Specifically, we derive a suboptimality bound for infinite-sample MPPI that characterizes the effects of temperature and prior coverage, followed by a high-probability finite-sample approximation error bound that reveals the dependence on sample size, temperature, and cost variability. By incorporating these bounds into a terminal-constraint-free MPC framework, we further establish semi-global practical recurrence of the closed-loop system. Numerical examples involving a four-link planar manipulator illustrate the temperature-sampling tradeoff and the influence of prior quality on approximation accuracy and closed-loop performance.

0.29

Model Augmentation for thermo-mechanical hysteretic dynamics

eess.SY ·2026-11-17 ·Samarth Toolhally, Ozan Celik, Daniel Deenen, Maarten Schoukens, Siep Weiland
Abstract

Accurate prediction of thermo-mechanical wafer deformation is essential for sub-nanometer overlay in extreme ultraviolet (EUV) lithography. Localized heating expands the wafer, while friction between the wafer and its supporting elastic pillars causes hysteresis. Modeling this is challenging because of the system's large scale, the computational cost of friction-induced hysteresis and the sparse deformation measurements. This work augments a linear thermo-mechanical baseline model with a learning-based structure in a linear fractional representation (LFR) framework, identified from sparse measurements. Simulation results show that the augmented model captures the unmodeled dynamics with hysteretic effects and predicts the deformation significantly more accurately than the baseline.

0.29

Safely Mitigating Grid Imbalance: A Hybrid Distributional Wide-Margin Recourse Approach

eess.SY ·2026-11-17 ·Shoaib Hussain, Mostafa Farrokhabadi, Hamidreza Zareipour
Abstract · 17 pages, 10 figures, 3 tables

This paper proposes the Multi-Agent Distributional Wide-Margin Policy Gradient (MADWMPG) framework: a dual-policy hybrid imitation and safe reinforcement learning (IL-SRL) method designed for safely mitigating grid imbalance in multiphase active distribution networks with distributed soft open points (SOPs). The framework comprises a wide-margin constraint barrier recourse (WCBR) module that seeks the maximal separation of constraint-compliant and unsafe actions and learns to proactively suppress exploratory violations. The framework addresses operational volatility using a novel adaptation of noncrossing quantile regression (NQR) distributional critics that track the complete distribution of policy returns. The set of behavior and recourse policies integrates an instantaneous physics-based and differentiable action projection mechanism that generates SOP constraint-compliant actions without relying on active learning. Policy safety is made responsive using an enhancement of conventional experience replay, and the safety performance of the framework is supported by probabilistic bounds derived in the paper. The proposed framework was trained and tested on a modified multiphase IEEE 123-bus test feeder. Results demonstrate its superiority to conventional reinforcement learning and SRL methods and show that it is highly competitive against the nonlinear AC optimal power flow.

0.29

Data-driven model reduction: An identification perspective using Generalized Bases of Orthonormal Functions

eess.SY ·2026-11-17 ·Bernard Vau, Henri Bourlès
Abstract

This paper proposes a data-driven model order reduction approach based on system identification concepts. A revised equation error formulation is introduced, in which the model is parameterized using Generalized Bases of Orthonormal Transfer Functions (GBOFs). The generalized reduction problem is formulated in the Hambo frequency domain associated with the GBOF transform. The frequency-domain fit of the reduced-order model is adjusted through the selection of the basis poles. A fit indicator depending on these poles is introduced, providing valuable heuristic guidance for their selection.

0.29

On the recursive feasibility and stability of MPC without terminal constraints

eess.SY math.OC ·2026-11-17 ·Mara Perez, Ignacio Sanchez, Filiberto Fele, Daniel Limon
Abstract · 6 pages, 1 figure. Comments are welcome!

In this paper, we analyze stability and recursive feasibility of a model predictive control (MPC) scheme that does not rely on terminal ingredients -- such as terminal constraints or a terminal cost based on a terminal controller. In the proposed formulation, we augment the cost function by a weighted finite tail which is optimized online, and we show it plays the role of a stabilizing terminal cost. Under a relaxed, local dynamic programming condition on the tail, we certify a family of regions -- parameterized by the tail's weight coefficient -- on which the problem is recursively feasible and the origin is asymptotically stable. These regions are larger as the weight increases: in particular, we show that any state that can be steered within the control horizon to a predefined neighborhood of the origin is contained in a certified region for a sufficiently large weight. Finally, we characterize the closed-loop performance over the certified region, bounding the suboptimality of the closed loop with respect to the infinite-horizon optimal cost. We conclude the discussion by illustrating these results on a simple numerical example.

0.28

Facet-offset Polytope Expansion for Dynamic Operating Envelopes of Aggregated Distribution Grid Resources

eess.SY ·2026-11-17 ·Feng Chen, Kumarsinh Jhala, Kiran Kumar Challa, Zongjie Wang
Abstract · 2027 IEEE PES Grid Edge Technologies Conference & Exposition

This paper presented a facet-offset polytope expansion framework with sensitivity-informed directions (FOPE-sen) for constructing DOEs in three-phase unbalanced distribution systems. The facet directions are selected from DGR coordinate directions and sensitivity-informed network-critical directions, while the facet locations are determined through a two-step bisection procedure guided by a power-flow feasibility oracle. Numerical results demonstrate that FOPE-sen captures at least 106% more flexibility per dimension than box-shaped and random-direction expansion methods. Moreover, FOPE-sen constructs DOEs comparable to or larger than those of VBPE while using fewer than 1/40 as many facets in dimensions greater than four. Future work will focus on stronger safety verification for nonconvex AC feasible regions, scalability in high-dimensional spaces, and extensions to multi-aggregator allocation and market-participation applications.

0.28

SR-VLA: Stall-Reduced VLA Execution via Predicted-Input Lookahead and Selective KV Cache Repair

cs.RO eess.SY ·2026-11-17 ·Tao Zhang, Yi Zhong, Jinhao Ruan, Yipeng Gao, Linfeng Zhong +2 more
Abstract

Vision-language-action (VLA) policies incur control stalls when inference blocks the next action chunk. Asynchronous early inference can hide latency, but its results may be inconsistent with the arriving actual inputs. Unconditional reuse risks acting on mismatched inputs, whereas full recomputation discards useful precomputation. We propose SR-VLA, combining predicted-input lookahead with selective KV cache repair. During current chunk execution, a lightweight action-conditioned predictor generates next-boundary visual observations and proprioceptive state. These predictions, together with the task instruction, drive the frozen VLA to precompute a speculative KV cache and action chunk. After real inputs are available to VLA, predicted--real differences and task conditions determine whether to accept the action chunk or trigger selective KV repair. On LIBERO, SR-VLA achieves $2.67\times$ and $1.27\times$ query speedups over the respective synchronous $π_{0.5}$ and OpenVLA-OFT baselines, while maintaining competitive average task success rates. Under target-motion perturbation, success remains baseline-comparable, with $1.52\times$--$1.93\times$ query speedup.

0.26

Causal Input--Output Modeling with Prior Information

eess.SY math.OC ·2026-11-17 ·Amirhossein Taghvaei, Daran Xu
theory
Abstract

We study the problem of inferring a causal input--output model from unpaired marginal observations and a given prior coupling. We formulate this problem using entropic causal optimal transport, which seeks a coupling that preserves the observed marginals, satisfies a one--sided causality constraint, and remains close to the prior. For Gaussian marginals and general quadratic costs, we prove that the optimal coupling is Gaussian and reduce the problem to a finite--dimensional optimization over covariance matrices. Using a Cholesky transformation to simplify the causality constraint, we derive a Lagrangian dual formulation and develop a block coordinate descent algorithm with explicit update formulas. We establish linear convergence of the algorithm, with explicit convergence rates, and prove stability of the optimal coupling under marginal covariance perturbations and derive a statistical error bound. We further show its equivalence to Sinkhorn iterations. Numerical experiments demonstrate the computational efficiency of the proposed algorithm and illustrate the predictive advantages of causal couplings.

0.25

Technical Report of Mobile Manipulator Robot for Industrial Environments

cs.RO cs.CVeess.SY ·2026-11-17 ·Aria Dadnavi, Matin Jamshidi, Kian Khatibi, Shahriar Karimi Makhsous, Saina Najafi Shabestari +6 more
Abstract · 25 pages, 22 figures, 4 tables. Technical report of Team Auriga's 2025 mobile manipulator (@Work) platform (Robotics and Intelligent Automat…

This paper presents the design and implementation of Auriga Robotics' mobile manipulator robot, developed at the Robotics and Intelligent Automation Laboratory, Department of Electrical Engineering, Shahid Beheshti University. Hereafter, this platform is referred to as the @Work Robot. The robot targets industrial tasks, with an emphasis on improving efficiency in repetitive or hazardous work environments. It uses a 4-wheel Mecanum drive for omnidirectional movement and a 5-degree-of-freedom manipulator arm with a 3D-printed gripper for object handling and navigation. The electronics consist of custom PCB boards built around ESP32 microcontrollers, together with an NVIDIA Jetson Orin NX that handles high-level processing and real-time system control. On the software side, Hector SLAM provides mapping and localization, A* handles path planning, and YOLO performs real-time object detection. These modules draw on fused sensor data to deliver robust navigation and reliable collision avoidance. Overall, the system shows effective hardware-software integration, with consistent performance in autonomous navigation, object handling, and task execution in industrial-like settings.

0.23

Transmission Expansion and Energy Storage Planning under Great Britain's Uncertain Energy Transition Pathways

eess.SY math.OC ·2026-11-17 ·Shengming Cao, Runsheng He, Waqquas Bukhsh
benchmark
Abstract · 49 pages, 10 figures

Long term power system planning must account for uncertainty in electricity demand and generation, while recognising the different investment and commissioning characteristics of transmission and energy storage. This paper develops a multi year stochastic framework for coordinated transmission expansion and energy storage planning and applies it to the Great Britain electricity system to assess investment needs through to 2050. The framework integrates multiple energy transition pathways with an enhanced representation of the GB transmission network and captures the different characteristics of transmission and storage investment over time. The coordinated treatment of transmission and storage is considered as a system level benchmark rather than a centralised investment mechanism. In GB, transmission investment is subject to regulated forward looking planning, while storage is largely developed by market participants in response to market signals. The framework therefore quantifies the potential value of improved coordination and identifies where existing investment signals may not encourage complementary storage deployment. Results across alternative transmission only, storage only and coordinated investment strategies show substantial benefits from joint planning. Compared with transmission only planning, coordinated investment reduces additional transmission capacity by 2%, cumulative expected load shedding by 97%, and total expected system cost by 36%, primarily through avoided unserved energy. Transmission preparation periods affect the timing, but not the final portfolio, of investments. Finally, the energy transition pathway materially influences investment requirements, with Electrification and Efficiency requiring 7% more transmission capacity and 45% more storage energy capacity than Falling Behind.

0.23

Grid-Aware Assessment of Forced Oscillations from Spatially Distributed Large Loads

eess.SY ·2026-11-17 ·Min-Seung Ko, Hao Zhu
Abstract

This paper develops a unified analytical framework for assessing forced oscillations induced by spatially distributed large-load fluctuations to determine their grid-aware fluctuation magnitudes. Emerging computational loads, such as hyperscale datacenters, present cyclic power variations with dynamic impacts depending on the waveform characteristics, forcing frequency, and source location. Our grid-aware framework represents these fluctuations using deterministic harmonic and probabilistic spectral models, and analyzes the resulting bus-wise frequency deviations based on a generic grid frequency-response model. Deterministic, risk-aware, and band-based probabilistic severity measures are further developed to obtain admissible fluctuation limits under different assumptions on frequency and phase uncertainty. We also formulate a security-constrained (SC) optimization of locational fluctuation magnitudes among existing large loads, accounting for their aggregated frequency deviation level. Case studies on the IEEE 39-bus and 14-generator Australian systems validate the proposed severity measures and admissible limit analysis. The SC optimization results further show that distributing a given amount of aggregate fluctuation toward dynamically less sensitive locations can substantially reduce system-wide oscillation severity.

0.20

Feedback-Linearised Trajectory Tracking via Model-Informed Robust Geometric DDPC

cs.RO eess.SY ·2026-11-17 ·Pranav Gupta, Deepan Muthirayan, Ravi Banavar
Abstract · This manuscript has been submitted to the 9th Annual Learning for Dynamics & Control Conference, taking place in Stockholm, Sweden, in June …

Data-driven predictive control (DDPC) predicts system responses from recorded trajectories, but nonlinear dynamics can limit the accuracy of fixed linear predictors. This paper studies geometric robust DDPC for trajectory tracking of a fully actuated manipulator using model-based feedback linearisation. A mechanical model maps virtual acceleration commands into joint torques, while recorded data capture residual dynamics arising from imperfect model parameters. The controller minimises worst-case prediction error over an open Grassmannian ball centred at the fitted trajectory subspace. Simulations on a planar 2R manipulator under gravity compare the proposed method with double-integrator model predictive control (MPC) and direct regularised data-enabled predictive control (DeePC). Results show improved tracking over double-integrator MPC despite increased torque effort, while performance relative to DeePC depends on the reference trajectory.

0.20

An Explainable Agentic AI Framework for Distribution System Load Restoration

eess.SY ·2026-11-17 ·S. M. Monzurul Hoque Chowdhury, Qinghua Li
Abstract

Distribution system load restoration is a sequential decision-making problem in which restoration actions must maximize served load, prioritize critical loads, and preserve power-flow feasibility. This paper proposes an explainable agentic artificial intelligence (AI) framework for distribution system load restoration. The proposed framework uses multiple large language model (LLM) agents for planner guidance, strategic plan generation, decision arbitration, memory-based feedback, and explanation generation. Unlike unconstrained LLM-based control, the proposed method restricts agent decisions to feasible restoration actions and evaluates all candidate plans using deterministic AC power-flow feasibility checks. The framework is evaluated on IEEE 33-bus and 69-bus distribution feeders, as well as a larger IEEE 118-bus test case, and compared against reinforcement learning and planning baselines. The results show that the proposed agentic framework achieves better restoration performance while producing structured explanations for operator-facing decision support. A direct comparison with a Single-LLM-Agent baseline further shows that the proposed multi-agent decomposition improves weighted restored load and critical-load recovery compared with a monolithic LLM decision structure.

0.20

Explicit Dual Model Predictive Path Integral Control

eess.SY ·2026-11-17 ·Christopher Narr, Martin Buss
Abstract

In this work, we present a fully sampling-based explicit dual control algorithm that combines a particle filter and model predictive path integral control for partially observed nonlinear systems. The key idea is to roll out multiple sampled input sequences from all particles of the particle filter to obtain a belief-averaged control cost that induces caution. The same rollouts are used to sample imagined future measurements, which are drawn from the predictive distribution of the particle belief. This results in a Monte Carlo information term that gives a sample estimate of the expected information gain of each sampled input sequence and thereby induces probing. In contrast to existing explicit dual control schemes, the proposed scheme neither relies on gradients of the dynamics or measurement models nor requires a Gaussian belief. Furthermore, all computations can be parallelized across input sequences, particles, and imagined measurements, which makes the scheme amenable to real-time implementation on suitable hardware. We demonstrate the effectiveness of the proposed method in a simulation example of a planar terrain-aided navigation task.

0.20

Grid-Responsive Pricing of Dynamic Wireless Charging with Coupled Residential Demand

eess.SY math.OC ·2026-11-17 ·Zhiheng Wang, Dipanjan Ghose, S. Sivaranjani, Junjie Qin, Qin Lv
Abstract · 5 pages, 2 figures. Submitted to the IEEE Power & Energy Society General Meeting (PESGM) 2027

Dynamic wireless charging (DWC) is an emerging technology that enables electric vehicles (EVs) to recharge while traveling. Its unique operational model creates two revenue streams for DWC operators through lane access tolls and electricity sales. However, these prices influence not only traffic flow and roadway energy delivery but also how much EV charging moves to homes instead, potentially within the same distribution network. Neglecting this coupling can underestimate grid loading and compromise pricing decisions. We develop a coupled traffic, EV energy, driver-response, and AC distribution network model with a model predictive control (MPC) framework that jointly optimizes DWC tolls and charging prices while explicitly accounting for displaced residential charging and network losses. Using 78 days of I-210 West traffic data, we show that conventional MPC exceeds feeder loading limits on 52 days, whereas the proposed residential load aware controller eliminates these violations with only 0.6% lower operator revenue. Under a more constrained feeder configuration, it eliminates 50 overloaded days with a 0.9% revenue reduction. Sensitivity analyses highlight the importance of accurate residential demand estimation and feeder loss modeling, while peak-aware pricing further reduces daily peak loading by 3.7% relative to the base load. These results demonstrate the value of coordinated pricing that captures the interaction between roadway and residential charging.

0.19

Benchmarking a Grid Foundation Model for AC Optimal Power Flow: Feasibility and Warm-Start Results on Synthetic Texas and South Carolina Grids

eess.SY ·2026-11-17 ·Tam Kemabonta, Brighton Ombuki, Vladimir Abdelnour
Abstract · 5 pages, 3 figures, 2 tables. Submitted to the 2027 IEEE PES General Meeting

Learning-based surrogates for alternating-current optimal power flow (AC-OPF) are often judged by the gap between their objective value and that of a nonlinear solver, which does not show that a prediction satisfies power balance and network limits. This paper tests whether predictions from the pretrained foundation model GridSFM-Open 1.1.0, recently released by Microsoft, are feasible AC-OPF operating points and whether they are useful initial values for a nonlinear solver. The model was evaluated zero-shot and after fine-tuning on 10, 100, and 1,000 labeled scenarios (FT10, FT100, and FT1000) on 500 scenarios for each of two synthetic grids, ACTIVSg2000 (Texas) and ACTIVSg500 (South Carolina), with Ipopt, MATPOWER MIPS, and pandapower PIPS as solver references. The lowest Texas median cost gap, 0.44\% for FT100, coincided with thermal violations in 24.1\% of cases and a median maximum bus power mismatch of 89 MW and 197 Mvar. Texas zero-shot predictions had a median maximum mismatch of 648 MW and 1,641 Mvar, and every South Carolina case violated a thermal limit for every model variant. Seeding Ipopt with FT1000 predictions gave geometric-mean solve-time speedups of 1.55x on South Carolina and 1.12x on Texas (p = 0.06, not significant at the 5\% level), while zero-shot seeds slowed South Carolina solves (0.26x). A small objective gap therefore did not indicate feasibility: although median zero-shot inference took only 1.32 s on Texas and 0.157 s on South Carolina on a shared host, every prediction still required a nonlinear solver, and fine-tuned predictions gave modest, grid-dependent warm-start speedups while zero-shot predictions slowed convergence on one grid. GridSFM may therefore be useful for initializing a nonlinear solver, but obtaining feasible AC-OPF solutions still required the solver.

0.18

Capacity Credit of Data-Centre Flexibility: The Role of Workload Scheduling and Usable Battery Energy

eess.SY ·2026-11-17 ·Pengyu Ren, Wei Sun, Jiakai Wu, Yifan Wang
Abstract

Data centres that offer the same reduction in grid demand can make different contributions to supply adequacy. We estimate the effective load-carrying capability of workload scheduling and on-site batteries by matching expected unserved energy across paired simulations. Longer task deadlines increase capacity credit by giving operators more choice over when to defer and recover work. A fixed evening commitment shifts more work than deadline-based scheduling but delivers a smaller reduction in expected unserved energy. Protected backup reserves leave uninterruptible power supplies with limited capacity credit, while longer discharge durations raise the credit of dedicated batteries at the same response rating. These findings suggest assessing recovery deadlines and dispatchable energy alongside offered demand reductions.

0.18

Optimal Energy and LLM Task Scheduling for Multiple Cloud Data Centers

eess.SY ·2026-11-17 ·Zhongqi Zhao, Lei Fan, Zhu Han
Abstract

This paper studies coordinated Large Language Model (LLM) training-task scheduling and facility operation for geographically distributed cloud data centers (CDCs). The central question is how two scheduling flexibilities - cross-CDC task transfer and task reordering - affect service and economic outcomes when remote execution remains subject to compute, bandwidth, energy, and zone-temperature constraints. We formulate a mixed-integer optimization model that jointly determines task outcomes, heterogeneous-node assignment, execution timing, pre-execution data transfer, node activation, and facility operation. A deterministic 24-hour case study contains 3 CDCs, 15 heterogeneous graphics processing unit (GPU) nodes, 4 thermal zones per CDC, and 103 tasks. In this instance, enabling transfer increases net profit by 57.9% relative to the otherwise matched no-transfer variant and increases completed tasks from 76 to 103. The more restrictive no-transfer FIFO-prefix variant completes 14 tasks and yields negative profit. The proposed schedule transfers 39 tasks, achieves 70.3% node utilization, and maintains zone temperatures within 24°C-27°C. These fixed-instance results indicate that cross-CDC transfer and task-reordering flexibility can improve the modeled profit and task-completion outcomes under the studied conditions.

0.17

Solver-Free and Cost-Aware Online Scheduling of AI Data Center Tasks

eess.SY cs.AI ·2026-11-17 ·Xin Qin, Dawei Qiu, Fei Teng
benchmark
Abstract

AI data centers receive computing tasks continuously, whereas solver-based optimization takes time to run and electricity is priced per settlement period. This paper bridges this timing gap through a solver-free online optimization rule inspired by dual decomposition. The rule assigns each task a start period upon arrival to optimize energy cost subject to operational constraints, shifting flexible tasks toward low-price periods while keeping high-priority tasks on schedule. Simulations on the A100 GPU trace (3,456 GPUs, 1,343 tasks) show that the proposed rule keeps GPU and CPU use within capacity while scheduling tasks at low cost. Compared with a multi-timescale optimization benchmark, it achieves a lower cost and, because it needs no solver, much less computation time.

0.15