MustaphaL/pipeline-thickness-dataset

该数据集是一个模拟工业管道厚度测量的合成数据集，用于预测腐蚀和监测管道完整性。数据集包含8440个测量数据，覆盖250个工业管道段，时间跨度为2005年至2025年。数据生成基于行业认可的腐蚀模型，如Waard-Milliams模型、API 570 / ASME B31.8标准和NACE SP0775。数据集的结构包括静态特征（如管道段ID、材料、直径、厚度等）、时间特征（如温度、压力、流量等）、预测目标（如测量厚度、腐蚀速率、剩余使用寿命等）和派生特征（如历史值、厚度变化、腐蚀速率滚动平均等）。数据集还提供了详细的统计信息，如厚度范围、腐蚀速率范围、不同腐蚀机制和材料类型等。

This dataset is a synthetic simulation of industrial pipeline thickness measurements for corrosion prediction and integrity monitoring. It contains **8,440 measurements** across **250 segments** of industrial piping, covering the period from **2005 to 2025**. The data is generated from industry-recognized corrosion models: - **Waard-Milliams model** for CO₂ corrosion - **API 570 / ASME B31.8** for integrity criteria - **NACE SP0775** for internal corrosion. The dataset structure includes static features (e.g., segment ID, material, diameter, thickness), temporal features (e.g., temperature, pressure, flow rate), prediction targets (e.g., measured thickness, corrosion rate, remaining useful life), and derived features (e.g., historical values, thickness change, rolling average of corrosion rate). The dataset also provides detailed statistics such as thickness range, corrosion rate range, different corrosion mechanisms, and material types.