用大模型为油井异常检测提供可解释的决策说明和命名。
An Explainable LLM Agent Layer for Open-World Anomaly Detection in Oil Wells
- 在已有检测流程后加大模型代理，用自然语言解释判断依据。
- 对9类异常识别准确率达35.1%（Top-1），7类异常验证精度0.91。
- 能稳定给未知异常聚类命名，帮助工程师理解与审计结果。
针对油井异常检测的开放世界学习（OWL）流程，现有方法虽能识别异常类型，但缺乏解释性，无法说明为何判定、也不提供操作建议，且难以为人理解发现的新型异常集群。本文在上游检测流程下游引入一个大语言模型（LLM）代理，作为补充而非替代。基于通过NVIDIA NIM服务的Qwen3.5-397B-A17B MoE模型，该代理接收结构化传感器数据及上游分类或新颖性判断结果，输出自然语言解释、置信度排序的批判意见，以及对检测到的新颖性事件进行统一命名。在3W数据集中的989个真实井段上，代理在全部9类中达到35.1%（Top-1）/63.9%（Top-3，95% CI [56.9, 70.4]）分类准确率，在7个探测类别中验证准确率达71.7%（Top-2，95% CI [64.8, 77.6]），精确率为0.91（[0.84, 0.95]），对7个隐藏类别中的5个实现89.7%（[87.0, 91.9]）新颖性检测并保持稳定的聚类命名。该代理不独立分类，而是：(1) 在传感器证据支持时确认上游判断；(2) 用可审计的语言解释决策；(3) 标出上游标签不合理的情况；(4) 为未标注事件赋予统一的人类可读名称。目标是弥合当前阻碍部署的可解释性鸿沟。
原文摘要 · Abstract (English)
Open-World Learning (OWL) pipelines for oil well anomaly detection have recently been shown to combine autoencoder-based detection, multiclass classification, and Mahalanobis-based novelty detection on the public 3W dataset. These pipelines answer \textit{what happened}, but they do not explain \textit{why the model believes it} or \textit{what the operator should do next}, and they do not put a human-readable name on the novelty clusters they discover. This paper evaluates a Large Language Model (LLM) agent layer placed downstream of the OWL pipeline, designed as a \textbf{companion} to the published upstream methods rather than a replacement. Using the Qwen3.5-397B-A17B Mixture-of-Experts model served via NVIDIA NIM, the agent receives structured sensor metrics and upstream classification or novelty assertions, and returns natural-language justifications, confidence-ranked critiques, and consolidated names for detected novelties. Across three studies spanning 989 real well-file segments from the 3W dataset, the agent achieved $35.1\%$ top-1 / $63.9\%$ top-3 (95\% CI [56.9, 70.4]) classification on all nine classes, $71.7\%$ top-2 validation [64.8, 77.6] with precision $0.91$ [0.84, 0.95] across 7 probed classes, and $89.7\%$ novelty detection [87.0, 91.9] with stable cluster naming on 5 of 7 hidden classes. The agent is not a standalone classifier. Its role is to: (1) confirm upstream decisions when sensor evidence supports them, (2) justify decisions in sensor-grounded language operators can audit, (3) flag disagreement when upstream labels are implausible, and (4) name novelties so that clustered unlabeled events arrive at the engineer with a consolidated human-readable label. The goal is to close the explainability gap that currently blocks deployment of OWL pipelines in operational settings.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。