arXiv subjects

Wei Guan

Publications and source records attributed to Wei Guan.

13 recordsLinked to original sources

arXiv2026

Stabilization of premixed NH3/H2/air flames via bluff-body flame holders

The stabilization mechanisms of fully premixed NH3/H2/air flames anchored behind a bluff body are investigated using combined experiments and direct numerical simulations. Particular attention is given to the interplay between preferential diffusion, heat release, flow recirculation, and turbulence-flame interaction. Comparison between non-reactive and reactive cases shows that thermal expansion strongly alters the flow field, increasing the recirculation zone length by about 40% and the shear layer width by roughly 50% near the end of the recirculation region. Excellent agreement between measurements and simulations for mean and fluctuating axial velocities validates the numerical approach. Analysis of the flame structure reveals a distinctive stabilization mechanism at the flame root: preferential hydrogen diffusion generates a localized diffusion flame branch that enhances radical production and promotes robust anchoring. Combustion proceeds sequentially, with hydrogen mainly consumed in the shear layer, followed by ammonia cracking and the main heat-release region. Near the bluff body, heat release is concentrated within the recirculation zone, while downstream regions are increasingly influenced by turbulence and velocity fluctuations. The roles of curvature and strain are quantified to assess stretch effects along the flame front. Convex curvature near the flame root enhances hydrogen enrichment and locally increases burning rates, reinforcing stabilization. In contrast, concave curvature and higher stretch near the end of the recirculation zone weaken the flame and mark a transition toward a turbulence-dominated regime. Overall, stabilization results from a coupled feedback between recirculation-driven heat exchange and rapid hydrogen oxidation, sustaining an intermediate ammonia reaction zone and enabling robust anchoring of carbon-free NH3/H2 flames.

physics.flu-dyn↗

arXiv2025

GLOW: Graph-Language Co-Encoding for Agentic Workflow Performance Prediction

Agentic Workflows (AWs) have emerged as a promising paradigm for solving complex tasks. However, automatically generating high-quality AWs remains expensive because AW optimization requires evaluating a large number of candidate AWs via execution, resulting in high computational cost and latency. Recently, AW performance prediction has become a hot research topic to avoid costly execution-based evaluation, but existing methods primarily use Graph Neural Networks (GNNs) to model workflow structures and insufficiently capture the semantic relationships among agents. To address this limitation, we propose GLOW, a unified framework for AW performance prediction that combines the graph-structure modeling ability of GNNs with the topology-aware semantic encoding capability of LLMs. Specifically, a graph-oriented LLM is first built through instruction-tuning on graph understanding tasks to extract topology-aware semantic representations from descriptive text of AWs. Meanwhile, a GNN explicitly models the structural information of AWs and produces corresponding structural representations. The semantic and structural representations are then fused in a shared latent space using a Transformer-based fusion module. A contrastive learning strategy is further introduced to learn more discriminative representations for AWs. Experiments on the FLORA-Bench benchmark demonstrate that GLOW consistently outperforms state-of-the-art baselines in both prediction accuracy and ranking utility. Moreover, when integrated into the AFLOW, an automatic AW generation framework, GLOW reduces optimization time by 98.7% with only a 0.031 average score decrease across three datasets, showing its effectiveness as an efficient surrogate evaluator for AW optimization.

cs.LG↗

arXiv2025

Falcon: A Comprehensive Chinese Text-to-SQL Benchmark for Enterprise-Grade Evaluation

We introduce Falcon, a cross-domain Chinese text-to-SQL benchmark grounded in an enterprise-compatible dialect (MaxCompute/Hive). It contains 600 Chinese questions over 28 databases; 77% require multi-table reasoning and over half touch more than four tables. Each example is annotated along SQL-computation features and Chinese semantics. For evaluation, we release a robust execution comparator and an automated evaluation pipeline, under which all current state-of-the-art large-scale models (including Deepseek) achieve accuracies of at most 50%. Major errors originate from two sources: (1) schema linking in large enterprise landscapes - hundreds of tables, denormalized fields, ambiguous column names, implicit foreign-key relations and domain-specific synonyms that make correct join/column selection difficult; and (2) mapping concise, colloquial Chinese into the exact operators and predicates required for analytics - e.g., choosing the correct aggregation and group-by keys, expressing time windows and granularities, applying unit conversions, handling NULLs and data-quality rules, and formulating nested or windowed subqueries. Falcon therefore targets Chinese-specific semantics and enterprise dialects (abbreviations, business jargon, fuzzy entity references) and provides a reproducible middle ground before full production deployment by using realistic enterprise schemas, query templates, an execution comparator, and an automated evaluation pipeline for end-to-end validation.

cs.CL↗

arXiv2025

EMIT: Enhancing MLLMs for Industrial Anomaly Detection via Difficulty-Aware GRPO

Industrial anomaly detection (IAD) plays a crucial role in maintaining the safety and reliability of manufacturing systems. While multimodal large language models (MLLMs) show strong vision-language reasoning abilities, their effectiveness in IAD remains limited without domain-specific adaptation. In this work, we propose EMIT, a unified framework that enhances MLLMs for IAD via difficulty-aware group relative policy optimization (GRPO). EMIT constructs a multi-task IAD dataset and utilizes GPT-generated object text descriptions to compensate for missing defective images. For few-shot anomaly detection, it integrates a soft prompt and heatmap-guided contrastive embeddings derived from patch-level comparisons. To better handle difficult data samples, i.e., cases where the MLLM struggles to generate correct answers, we propose a difficulty-aware GRPO that extends the original GRPO by incorporating a response resampling strategy to ensure the inclusion of correct answers in the sampled responses, as well as an advantage reweighting mechanism to strengthen learning from such difficult data samples. Extensive experiments on the MMAD benchmark demonstrate that EMIT significantly enhances the IAD performance of MLLMs, achieving an average improvement of 7.77\% over the base model (InternVL3-8B) across seven tasks.

cs.CV↗

arXiv2024

LogLLM: Log-based Anomaly Detection Using Large Language Models

Software systems often record important runtime information in logs to help with troubleshooting. Log-based anomaly detection has become a key research area that aims to identify system issues through log data, ultimately enhancing the reliability of software systems. Traditional deep learning methods often struggle to capture the semantic information embedded in log data, which is typically organized in natural language. In this paper, we propose LogLLM, a log-based anomaly detection framework that leverages large language models (LLMs). LogLLM employs BERT for extracting semantic vectors from log messages, while utilizing Llama, a transformer decoder-based model, for classifying log sequences. Additionally, we introduce a projector to align the vector representation spaces of BERT and Llama, ensuring a cohesive understanding of log semantics. Unlike conventional methods that require log parsers to extract templates, LogLLM preprocesses log messages with regular expressions, streamlining the entire process. Our framework is trained through a novel three-stage procedure designed to enhance performance and adaptability. Experimental results across four public datasets demonstrate that LogLLM outperforms state-of-the-art methods. Even when handling unstable logs, it effectively captures the semantic meaning of log messages and detects anomalies accurately.

cs.SE↗

arXiv2024

DABL: Detecting Semantic Anomalies in Business Processes Using Large Language Models

Detecting anomalies in business processes is crucial for ensuring operational success. While many existing methods rely on statistical frequency to detect anomalies, it's important to note that infrequent behavior doesn't necessarily imply undesirability. To address this challenge, detecting anomalies from a semantic viewpoint proves to be a more effective approach. However, current semantic anomaly detection methods treat a trace (i.e., process instance) as multiple event pairs, disrupting long-distance dependencies. In this paper, we introduce DABL, a novel approach for detecting semantic anomalies in business processes using large language models (LLMs). We collect 143,137 real-world process models from various domains. By generating normal traces through the playout of these process models and simulating both ordering and exclusion anomalies, we fine-tune Llama 2 using the resulting log. Through extensive experiments, we demonstrate that DABL surpasses existing state-of-the-art semantic anomaly detection methods in terms of both generalization ability and learning of given processes. Users can directly apply DABL to detect semantic anomalies in their own datasets without the need for additional training. Furthermore, DABL offers the capability to interpret the causes of anomalies in natural language, providing valuable insights into the detected anomalies.

cs.CL↗

arXiv2023

Temporal link prediction methods based on behavioral synchrony

Link prediction -- to identify potential missing or spurious links in temporal network data -- has typically been based on local structures, ignoring long-term temporal effects. In this chapter, we propose link-prediction methods based on agents' behavioral synchrony. Since synchronous behavior signals similarity and similar agents are known to have a tendency to connect in the future, behavioral synchrony could function as a precursor of contacts and, thus, as a basis for link prediction. We use four data sets of different sizes to test the algorithm's accuracy. We compare the results with traditional link prediction models involving both static and temporal networks. Among our findings, we note that the proposed algorithm is superior to conventional methods, with the average accuracy improved by approximately 2% - 5%. We identify different evolution patterns of four network topologies -- a proximity network, a communication network, transportation data, and a collaboration network. We found that: (1) timescale similarity contributes more to the evolution of the human contact network and the human communication network; (2) such contribution is not observed through a transportation network whose evolution pattern is more dependent on network structure than on the behavior of regional agents; (3) both timescale similarity and local structural similarity contribute to the collaboration network.

physics.soc-ph↗

arXiv2020

AIM 2020 Challenge on Real Image Super-Resolution: Methods and Results

This paper introduces the real image Super-Resolution (SR) challenge that was part of the Advances in Image Manipulation (AIM) workshop, held in conjunction with ECCV 2020. This challenge involves three tracks to super-resolve an input image for $\times$2, $\times$3 and $\times$4 scaling factors, respectively. The goal is to attract more attention to realistic image degradation for the SR task, which is much more complicated and challenging, and contributes to real-world image super-resolution applications. 452 participants were registered for three tracks in total, and 24 teams submitted their results. They gauge the state-of-the-art approaches for real image SR in terms of PSNR and SSIM.

cs.CV↗

arXiv2016

Distributed Optimization of Hierarchical Small Cell Networks: A GNEP Framework

Deployment of small cell base stations (SBSs) overlaying the coverage area of a macrocell BS (MBS) results in a two-tier hierarchical small cell network. Cross-tier and inter-tier interference not only jeopardize primary macrocell communication but also limit the spectral efficiency of small cell communication. This paper focuses on distributed interference management for downlink small cell networks. We address the optimization of transmit strategies from both the game theoretical and the network utility maximization (NUM) perspectives and show that they can be unified in a generalized Nash equilibrium problem (GNEP) framework. Specifically, the small cell network design is first formulated as a GNEP, where the SBSs and MBS compete for the spectral resources by maximizing their own rates while satisfying global quality of service (QoS) constraints. We analyze the GNEP via variational inequality theory and propose distributed algorithms, which only require the broadcasting of some pricing information, to achieve a generalized Nash equilibrium (GNE). Then, we also consider a nonconvex NUM problem that aims to maximize the sum rate of all BSs subject to global QoS constraints. We establish the connection between the NUM problem and a penalized GNEP and show that its stationary solution can be obtained via a fixed point iteration of the GNE. We propose GNEP-based distributed algorithms that achieve a stationary solution of the NUM problem at the expense of additional signaling overhead and complexity. The convergence of the proposed algorithms is proved and guaranteed for properly chosen algorithm parameters. The proposed GNEP framework can scale from a QoS constrained game to a NUM design for small cell networks by trading off signaling overhead and complexity.

cs.IT↗

arXiv2015

A Nonlinear Pairwise Swapping Dynamics to Model the Selfish Rerouting Evolutionary Game

In this paper, a nonlinear revision protocol is proposed and embedded into the traffic evolution equation of the classical proportional-switch adjustment process (PAP), developing the present nonlinear pairwise swapping dynamics (NPSD) to describe the selfish rerouting evolutionary game. It is demonstrated that i) NPSD and PAP require the same amount of network information acquisition in the route-swaps, ii) NPSD is able to prevent the over-swapping deficiency under a plausible behavior description; iii) NPSD can maintain the solution invariance, which makes the trial and error process to identify a feasible step-length in a NPSD-based swapping algorithm is unnecessary, and iv) NPSD is a rational behavior swapping process and the continuous-time NPSD is globally convergent. Using the day-to-day NPSD, a numerical example is conducted to explore the effects of the reaction sensitivity on traffic evolution and characterize the convergence of discrete-time NPSD.

math.OC↗

arXiv2015

Alpha-reliable combined mean traffic equilibrium model with stochastic travel times

Based on the reliability budget and percentile travel time (PTT) concept, a new travel time index named combined mean travel time (CMTT) under stochastic traffic network was proposed. CMTT here was defined as the convex combination of the conditional expectations of PTT-below and PTT-excess travel times. The former was designed as a risk-optimistic travel time index, and the latter was a risk-pessimistic one. Hence, CMTT was able to describe various routing risk-attitudes. The central idea of CMTT was comprehensively illustrated and the difference among the existing travel time indices was analysed. The Wardropian combined mean traffic equilibrium (CMTE) model was formulated as a variational inequality and solved via an alternating direction algorithm nesting extra-gradient projection process. Some mathematical properties of CMTT and CMTE model were rigorously proved. In the end, a numerical example was performed to characterize the CMTE network. It is founded that that risk-pessimism is of more benefit to a modest (or low) congestion and risk network, however, it changes to be risk-optimism for a high congestion and risk network.

math.OC↗

arXiv2015

Individual departure time decision considering departure scheduling utility

The scheduling utility plays a fundamental role in addressing the commuting travel behaviours. In this paper, a new scheduling utility, termed as DMRD-SU, was suggested based on some recent research findings in behavioural economics. DMRD-SU admitted the existence of positive arrival-caused utility. In addition, besides the travel-time-caused utility and arrival-caused utility, DMRD-SU firstly took the departure utility into account. The necessity of the departure utility in trip scheduling was analysed comprehensively, and the corresponding individual trip scheduling model was presented. Based on a simple network, an analytical example was executed to characterize DMRD-SU. It can be found from the analytical example that: 1) DMRD-SU can predict the accumulation departure behaviors at NDT, which explains the formation of daily serious short-peak-hours in reality, while MRD-SU cannot; 2) compared with MRD-SU, DMRD-SU predicts that people tend to depart later and its gross utility also decrease faster. Therefore, the departure utility should be considered to describe the traveler's scheduling behaviors better.

math.OC↗

arXiv2013

Route guidance strategies revisited: Comparison and evaluation in an asymmetric two-route traffic system

To alleviate traffic congestion, a variety of route guidance strategies has been proposed for intelligent transportation systems. A number of the strategies are proposed and investigated on a symmetric two-route traffic system over the past decade. To evaluate the strategies in a more general scenario, this paper conducts eight prevalent strategies on a asymmetric two-route traffic network with different slowdown behaviors on alternative routes. The results show that only mean velocity feedback strategy is able to equalize travel time, i.e., approximate user optimality; while the others fail due to incapability of establishing relations between the feedback parameters and travel time. The paper helps better understand these strategies, and suggests mean velocity feedback strategy if the authority intends to achieve user optimality.

physics.soc-ph↗