In-House Multi-Sensor-Location Vibration Dataset for PMSM-Based Drive Modules with Physically Induced Disturbances 核心 · 已核验

mxdsmrdx57b77a96

Note: This dataset is provided as the public dataset associated with a manuscript submitted for peer review. This dataset contains tri-axial vibration signals collected from permanent magnet synchronous motor (PMSM)-based drive modules under multiple sensor-location and physically induced disturbance conditions. It is designed to evaluate cross-location generalization and false alarm suppression in vibration-based fault diagnosis. [Experimental Setup] Motor: Frameless inrunner PMSM (CubeMars RI60 KV120) Operating condition: 500 rpm, no-load condition Sensor: LSM6DS3TR-C tri-axial MEMS accelerometer (STMicroelectronics) Sampling rate: Effective logging rate of 5 kHz Recording duration: 10 s per recording Diagnostic classes: Normal (NR), Bearing Fault (BF), Magnet Fault (MF), Stator Assembly Fault (SF), and Eccentricity Fault (EF) Temporal variability: Data acquisition for Loc S, Loc A, Loc B, and Loc C was conducted over separate time intervals to introduce additional temporal variability. The folder names use "Noise" for historical file-naming consistency, but the corresponding signals were collected under physically induced disturbance conditions rather than mathematically injected noise. [Dataset Structure] LocS_Train_Clean:Data measured at the reference sensor location, Loc S (12 o'clock), under clean conditions without intentionally introduced external disturbances. Five independent 10-s recordings were acquired for each diagnostic class. LocS_Train_Noise:Data measured at Loc S (12 o'clock) under five physically induced disturbance types (D1-D5), including contact vibration and hammer-induced impacts. One 10-s recording was acquired for each combination of diagnostic class and disturbance type. LocS_Val_Clean:An independent held-out dataset measured at Loc S under clean conditions, strictly separated from the training set. LocC_Val_Test_Noise:Data measured after relocating the sensor to Loc C (2 o'clock) under the five disturbance types (D1-D5). This subset was used for validation-based model selection and was also included in the final evaluation; therefore, it should be interpreted as a validation-exposed evaluation location rather than a strictly unseen test location. LocA_Test_Noise:Data measured at Loc A (9 o'clock) under the five disturbance types (D1-D5). This subset represents a strictly unseen test location that was not used during training or validation. LocB_Test_Noise:Data measured at Loc B (11 o'clock) under the five disturbance types (D1-D5). This subset also represents a strictly unseen test location that was not used during training or validation.

本卡描述如含来自官方页的原文片段,其版权归原作者,不在本站 CC-BY 4.0 许可范围(见关于与许可)。

落地页
https://zenodo.org/records/19550319
许可证
CC-BY-4.0 (置信:verified_official)
国内可访问性
国内直连:直连超时(慢或不稳定,非封锁证据) (2026-07-11 检测) 非直连:可达 (2026-07-11 检测)
「直连超时」表示检测窗口内未完成,系慢或不稳定证据,不构成封锁证据。
设备类型
synchronous_motor
PHM 任务
fault_diagnosis

故障工况

fault_type: healthy_baseline
description: 诊断类 BF,部位未述fault_type: bearing_fault_unspecified

传感器

sensor_type: accelerometer
溯源(13 条)
来源链接: https://zenodo.org/records/19550319 日期: 2026-07-09
日期: 2026-07-10
日期: 2026-07-10
日期: 2026-07-10
日期: 2026-07-10
日期: 2026-07-10
日期: 2026-07-10
日期: 2026-07-10
日期: 2026-07-10
日期: 2026-07-12
日期: 2026-07-14
日期: 2026-07-25
日期: 2026-07-26