arXiv subjects

Qingchun Hou

Publications and source records attributed to Qingchun Hou.

8 recordsLinked to original sources

arXiv2026

Spatial LLM Workload Shifting Needs Foresight: Model Commitment for AI Data Center Operation under Power Grid Constraints

AI data centers may face power supply shortages during certain periods, requiring operators to shift large language model (LLM) inference workloads spatially to maintain service rates. However, existing workload-shifting methods typically assume that any data center with sufficient computing resources can immediately serve shifted requests, which may lead to infeasible transfers and unserved demand. This letter proposes model commitment (MC), a mixed-integer linear programming framework that jointly schedules model deployment and cross-site request routing under power constraints and electricity-price signals. First, MC formulates the intertemporal coupling introduced by model replica loading. Second, it translates prefill and decode latency requirements into the amount of demand that each replica can serve. Case studies based on real-world data show that MC enables AI data center operators to achieve a 100% service rate under time-varying grid conditions and reduce total operating cost by 29.0%.

eess.SY↗

arXiv2026

From Tokens to Energy Flexibility: Quantization-Enabled Demand Response for Data Centers with LLM Inference Workloads

The rapid growth of large language model (LLM) inference is creating significant data-center loads that face increasing energy-management challenges under tightening grid conditions and demand response (DR) requirements. Conventional data-center energy management mainly relies on temporal and spatial workload shifting and campus-level energy asset scheduling, but it usually treats LLM inference demand as an aggregate load. As a result, these approaches fail to exploit the internal characteristics of LLM serving and therefore overlook the flexibility offered by LLM-specific techniques such as model quantization. To unlock this flexibility, this paper proposes a quantization-enabled energy management framework for grid-responsive LLM inference data centers. First, a quantization-to-power model is established to map each model--quantization configuration to a compact set of dispatchable parameters. Second, a two-stage quantization-enabled DR model is developed to account for model instance switching, request routing, and precision selection. Third, a multi-campus co-optimization method is introduced for DR participation by integrating grid-side electricity and carbon signals with the quantization-enabled DR model. Case studies show that the proposed framework reduces total data-center operating cost by 34.3\% without curtailing served token volume, validating model quantization as an effective flexibility lever for grid-responsive LLM data-center energy management.

eess.SY↗

arXiv2025

T-SKM-Net: Trainable Neural Network Framework for Linear Constraint Satisfaction via Sampling Kaczmarz-Motzkin Method

Neural network constraint satisfaction is crucial for safety-critical applications such as power system optimization, robotic path planning, and autonomous driving. However, existing constraint satisfaction methods face efficiency-applicability trade-offs, with hard constraint methods suffering from either high computational complexity or restrictive assumptions on constraint structures. The Sampling Kaczmarz-Motzkin (SKM) method is a randomized iterative algorithm for solving large-scale linear inequality systems with favorable convergence properties, but its argmax operations introduce non-differentiability, posing challenges for neural network applications. This work proposes the Trainable Sampling Kaczmarz-Motzkin Network (T-SKM-Net) framework and, for the first time, systematically integrates SKM-type methods into neural network constraint satisfaction. The framework transforms mixed constraint problems into pure inequality problems through null space transformation, employs SKM for iterative solving, and maps solutions back to the original constraint space, efficiently handling both equality and inequality constraints. We provide theoretical proof of post-processing effectiveness in expectation and end-to-end trainability guarantees based on unbiased gradient estimators, demonstrating that despite non-differentiable operations, the framework supports standard backpropagation. On the DCOPF case118 benchmark, our method achieves 4.27ms/item GPU serial forward inference with 0.0025% max optimality gap with post-processing mode and 5.25ms/item with 0.0008% max optimality gap with joint training mode, delivering over 25$\times$ speedup compared to the pandapower solver while maintaining zero constraint violations under given tolerance.

cs.LG↗

arXiv2021

Topology and Admittance Estimation: Precision Limits and Algorithms

Distribution grid topology and admittance information are essential for system planning, operation, and protection. In many distribution grids, missing or inaccurate topology and admittance data call for efficient estimation methods. However, measurement data may be insufficient or contaminated with large noise, which will introduce fundamental limits to the estimation accuracy. This work explores the theoretical precision limits of the topology and admittance estimation (TAE) problem, with different measurement devices, noise levels, and the number of measurements. On this basis, we propose a conservative progressive self-adaptive (CPS) algorithm to estimate the topology and admittance. Results on IEEE 33 and 141-bus systems validate that the proposed CPS method can approach the theoretical precision limits under various measurement settings.

cs.IT↗

arXiv2020

Embedding Lithium-ion Battery Scrapping Criterion and Degradation Model in Optimal Operation of Peak-shaving Energy Storage

Lithium-ion battery systems have been used in practical power systems for peak-shaving, demand response, and frequency regulation. However, a lithium-ion battery is degrading while cycling and would be scrapped when the capacity reduces to a certain threshold (e.g. 80%). Such scrapping criterion may not explore the maximum benefit from the battery storage. In this paper, we propose a novel scrapping criterion for peak-shaving energy storage based on battery efficiency, time-of-use price, and arbitrage benefit. A new battery life model with scrapping parameters is then derived using this criterion. Embedded with the life model, an optimal operation method for peak-shaving energy storage system is presented. The results of case study show that the operation method could maximize the benefits of peak-shaving energy storage while delaying battery degradation. Compared with the traditional 80% capacity-based scrapping criterion, our efficiency-based scrapping criterion can significantly improve the lifetime benefit of the battery.

eess.SP↗

arXiv2020

Sparse Oblique Decision Tree for Power System Security Rules Extraction and Embedding

Increasing the penetration of variable generation has a substantial effect on the operational reliability of power systems. The higher level of uncertainty that stems from this variability makes it more difficult to determine whether a given operating condition will be secure or insecure. Data-driven techniques provide a promising way to identify security rules that can be embedded in economic dispatch model to keep power system operating states secure. This paper proposes using a sparse weighted oblique decision tree to learn accurate, understandable, and embeddable security rules that are linear and can be extracted as sparse matrices using a recursive algorithm. These matrices can then be easily embedded as security constraints in power system economic dispatch calculations using the Big-M method. Tests on several large datasets with high renewable energy penetration demonstrate the effectiveness of the proposed method. In particular, the sparse weighted oblique decision tree outperforms the state-of-art weighted oblique decision tree while keeping the security rules simple. When embedded in the economic dispatch, these rules significantly increase the percentage of secure states and reduce the average solution time.

eess.SY↗

arXiv2019

A Linear LMP Model for Active and Reactive Power with Power Loss

Pricing the reactive power is more necessary than ever before because of the increasing challenge of renewable energy integration on reactive power balance and voltage control. However, reactive power price is hard to be efficiently calculated because of the non-linear nature of optimal AC power flow equation. This paper proposes a linear model to calculate active and reactive power LMP simultaneously considering power loss. Firstly, a linearized AC power flow equation is proposed based on an augmented Generation Shift Distribution Factors (GSDF) matrix. Secondly, a linearized LMP model is derived using GSDF and loss factors. The formulation of LMP is further decomposed into four components: energy, congestion, voltage limitation and power loss. Finally, an iterate algorithm is proposed for calculating LMP with the proposed model. The performance of the proposed model is validated by the IEEE-118 bus system.

eess.SY↗

arXiv2019

Probabilistic duck curve in high PV penetration power system: Concept, modeling, and empirical analysis in China

The high penetration of photovoltaic (PV) is reshaping the electricity net-load curve and has a significant impact on power system operation and planning. The concept of duck curve is widely used to describe the timing imbalance between peak demand and PV generation. The traditional duck curve is deterministic and only shows a single extreme or typical scenario during a day. Thus, it cannot capture both the probability of that scenario and the uncertainty of PV generation and loads. These weaknesses limit the application of the duck curve on power system planning under high PV penetration. To address this issue, the novel concepts of probabilistic duck curve (PDC) and probabilistic ramp curve (PRC) are proposed to accurately model the uncertainty and variability of electricity net load and ramp under high PV penetration. An efficient method is presented for modeling PDC and PRC using kernel density estimation, copula function, and dependent discrete convolution. Several indices are designed to quantify the characteristics of the PDC and PRC. For the application, we demonstrate how the PDC and PRC will benefit flexible resource planning. Finally, an empirical study on the Qinghai provincial power system of China validates the effectiveness of the presented method. The results of PDC and PRC intuitively illustrate that the ramp demand and the valley of net load face considerable uncertainty under high PV penetration. The results of flexible resource planning indicate that retrofitting coal-fired units has remarkable performance on enhancing the power system flexibility in Qinghai. In average, reducing the minimal output of coal-fired units by 1 MW will increase PV accommodation by over 4 MWh each day.

eess.SY↗