xpertsystems/oil008-sample

OIL-008 — 合成井眼轨迹数据集（样本）是一个免费的、模式相同的预览数据集，基于XpertSystems.ai的企业级井眼轨迹数据集，用于定向钻井、地质导向、测量质量控制和防撞机器学习。该样本覆盖了全球10个盆地的200口井，包含306,250个测量站，链接到11个表格中。数据集包括井主数据、计划轨迹、实际轨迹、地质导向目标、碰撞监测、测量不确定性、钻井分段、BHA定向数据、扭矩阻力效应、测量QC标志和井间距标签等文件，总行数为306,250行，磁盘大小约为16.3 MB。数据生成采用最小曲率法，并经过行业标准校准，验证得分为100.0/100（A+级）。数据集适用于轨迹异常检测、地质导向目标命中预测、防撞风险评分、测量QC机器学习、计划与实际交付分析、分段分类、扭矩阻力预测和多表关系机器学习等用例。OIL-008 — Synthetic Wellbore Trajectory Dataset (Sample) is a free, schema-identical preview of XpertSystems.ais enterprise wellbore-trajectory dataset for directional drilling, geosteering, survey QC, and anti-collision ML. The sample covers 200 wells across 10 global basins with 306,250 surveyed stations linked across 11 tables. The dataset includes files such as wells master, planned trajectory, actual trajectory, geosteering targets, collision monitoring, survey uncertainty, drilling sections, BHA directional data, torque drag effects, survey QC flags, and well spacing labels, totaling 306,250 rows and approximately 16.3 MB on disk. Data generation uses the minimum-curvature method and is calibrated to industry standards, with a validation score of 100.0/100 (Grade A+). It is suitable for use cases like trajectory anomaly detection, geosteering target-hit prediction, anti-collision risk scoring, survey QC ML, planned-vs-actual delivery analytics, section classification, torque-drag prediction, and multi-table relational ML.