Evolution of Surface Defects on Ball Screw Drive Spindles for intelligent Prognostics and Health Management Systems

The dataset shows the development of 82 surface defects (pits) over the operating time of Ball Screw Drives. The name of the images is structured as follows: XX_XX_YYMMDDHHMMSS_XX_XX. Here X is some identifier, which is not important in this context. The dataset is especially suited to investigate the development of surface defects on ball screw drive spindles. The dataset mainly addresses the machine learning research community for engineering and computer science to build intelligent models for surface defect detection and forecasting in the context of prognostics and health management (PHM). Each folder consists the evolution of one pit.本数据集展示了82个表面凹坑（pits）缺陷在滚珠丝杠副（Ball Screw Drives）全运行周期内的演化过程。图像的命名格式如下：XX_XX_YYMMDDHHMMSS_XX_XX，其中X为任意标识符，在此研究场景中无实际参考价值。本数据集尤其适用于研究滚珠丝杠主轴表面缺陷的演化规律。本数据集主要面向工程与计算机科学领域的机器学习研究社区，用于构建面向预测与健康管理（Prognostics and Health Management, PHM）场景的表面缺陷检测与预测智能模型。每个文件夹对应一处凹坑缺陷的完整演化序列。

搜集汇总

背景与挑战

背景概述

该数据集记录了82个滚珠丝杠驱动主轴表面缺陷（凹坑）随运行时间的演变过程，每个缺陷的演变数据存储于独立文件夹中，图像文件按时间戳命名。数据集专为机器学习和工程研究社区设计，用于开发智能模型以进行缺陷检测和预测，支持预测与健康管理（PHM）系统的应用。数据发布于2023年，总容量为56.4 GB，适用于智能制造和状态监测领域的研究。