Multimodal Bearing Health Management (MBHM)

该数据集名为MBHM，包含了135,516对振动信号段与故障类型的配对，以及542,064对文本提示与响应的配对，支持轴承健康管理系统中的各种任务。该数据集涵盖了从九个公开可获取的数据集中收集的262个工作条件，旨在促进多模态多任务模型的教学与发展。作为一个大规模数据集，它包含了135,516个振动信号段和542,064对文本数据。其任务范围包括异常检测、故障诊断、维护建议以及潜在风险分析。

This dataset, named MBHM, contains 135,516 pairs of vibration signal segments and fault types, as well as 542,064 pairs of text prompts and responses, supporting various tasks in bearing health management systems. It covers 262 operating conditions collected from nine publicly available datasets, aiming to facilitate the teaching and development of multimodal and multitask models. As a large-scale dataset, it includes 135,516 vibration signal segments and 542,064 text data pairs. Its task scope encompasses anomaly detection, fault diagnosis, maintenance recommendations, and potential risk analysis.

搜集汇总

背景与挑战

背景概述

MBHM数据集是一个多模态轴承健康管理数据集，包含振动信号和健康管理语料，旨在支持轴承健康状态识别研究。数据集来源于9个公开数据集，覆盖数千种工况，持续更新中。