轴承振动信号压缩重构与故障诊断算法数据集

实验平台基于一个2马力（1.5kW）的电机驱动系统，电机通过柔性联轴器带动转子，转子由被测轴承（测试对象）支撑在电机驱动端和风扇端。平台使用电火花加工技术在轴承的内圈、外圈和滚动体上精确模拟了不同尺寸（0.007英寸至0.04英寸）的单点故障，并利用安装在电机壳体的加速度传感器，在驱动端和风扇端位置，以12kHz与48kHz的高采样率采集了轴承在正常状态和多种故障类型/损伤程度下的振动信号。该数据集因其标准化实验设置、丰富故障工况和公开可获取性，被广泛应用于轴承故障检测、诊断与寿命预测算法的验证和比较研究。
48Khz采样频率下的驱动端轴承故障直径又分为0.007英寸、0.014英寸、0.028英寸三种类别，每种故障下负载又分为0马力、1马力、2马力、3马力。在每种故障的每种马力下有轴承内圈故障、轴承滚动体故障、轴承外环故障（由于轴承外环位置一般比较固定，因此外环故障又分为3点钟、6点钟和12点钟三种类别）。

The experimental platform is based on a 2-horsepower (1.5 kW) motor drive system. The motor drives the rotor via a flexible coupling, and the rotor is supported by the tested bearing (test object) at both the drive end and fan end of the motor. The platform uses electrical discharge machining (EDM) technology to precisely simulate single-point faults with sizes ranging from 0.007 inches to 0.04 inches on the inner rings, outer rings and rolling elements of the bearings. Acceleration sensors mounted on the motor housing are employed to collect vibration signals of the bearings under normal conditions and various fault types/severity levels at both the drive end and fan end, with high sampling rates of 12 kHz and 48 kHz.

This dataset has been widely utilized for the validation and comparative research of bearing fault detection, diagnosis and remaining useful life prediction algorithms due to its standardized experimental setup, abundant fault operating conditions and public accessibility.

For the drive end bearings under 48 kHz sampling frequency, the fault diameters are further categorized into three types: 0.007 inches, 0.014 inches and 0.028 inches. For each fault type, the applied load is divided into four levels: 0 HP, 1 HP, 2 HP and 3 HP.

Under each load level of each fault type, there are three fault categories: bearing inner ring fault, bearing rolling element fault, and bearing outer ring fault. Since the position of the bearing outer ring is generally fixed, the outer ring faults are further divided into three subtypes: 3 o'clock, 6 o'clock and 12 o'clock.

搜集汇总

背景与挑战

背景概述

该数据集是一个用于轴承故障诊断和压缩重构算法研究的振动信号集合，基于标准化的电机驱动实验平台，通过电火花加工模拟了多种尺寸和类型的轴承单点故障（包括内圈、外圈和滚动体），并以高采样率（12kHz和48kHz）采集了振动信号。它涵盖了正常状态和不同故障工况下的数据，具有丰富性和公开可获取性，主要用于验证和比较轴承故障检测、诊断与寿命预测算法。