WO2023279382A1 - 一种电机轴承运行状态故障检测方法及系统 - Google Patents

一种电机轴承运行状态故障检测方法及系统 Download PDF

Info

Publication number
WO2023279382A1
WO2023279382A1 PCT/CN2021/105526 CN2021105526W WO2023279382A1 WO 2023279382 A1 WO2023279382 A1 WO 2023279382A1 CN 2021105526 W CN2021105526 W CN 2021105526W WO 2023279382 A1 WO2023279382 A1 WO 2023279382A1
Authority
WO
WIPO (PCT)
Prior art keywords
frequency
sound wave
motor bearing
vibration direction
server
Prior art date
Application number
PCT/CN2021/105526
Other languages
English (en)
French (fr)
Inventor
徐萌萌
赵宁
Original Assignee
徐萌萌
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 徐萌萌 filed Critical 徐萌萌
Priority to PCT/CN2021/105526 priority Critical patent/WO2023279382A1/zh
Publication of WO2023279382A1 publication Critical patent/WO2023279382A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M13/00—Testing of machine parts
    • G01M13/04—Bearings
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M13/00—Testing of machine parts
    • G01M13/04—Bearings
    • G01M13/045—Acoustic or vibration analysis

Definitions

  • the invention relates to the technical field of fault prediction, in particular to a method and system for detecting faults in the operating state of motor bearings.
  • manual auscultation has the problems of high labor intensity for auscultation workers, low accuracy rate, and difficulty in training auscultation talents.
  • manual auscultation can only detect problems, but cannot judge the severity of the accident and when problems will occur.
  • the object of the present invention is to provide a method and system for detecting faults in the running state of motor bearings, which improves the accuracy of fault detection.
  • the present invention provides a method for detecting faults in the operating state of motor bearings, including:
  • the sound wave frequency is not within the set threshold range, upload the sound wave frequency and the vibration direction to the server according to the second set frequency; the second set frequency is higher than the first set frequency;
  • sample data of the motor bearing includes the acoustic frequency, vibration direction, fault type, fault level and remaining life of the motor bearing;
  • the convolutional neural network model is trained to obtain the motor bearing fault detection model.
  • the fault types include: bearing eccentricity, bearing inner ring faults and bearing outer ring faults.
  • the server is a cloud server.
  • the present invention also discloses a motor bearing operating state fault detection system, the motor bearing operating state fault detection system applies the motor bearing operating state fault detection method, and the motor bearing operating state fault detection system includes: an information collection module and server;
  • the information collection module is arranged on the casing of the motor to be detected, the information collection module is used to collect the sound wave frequency and the vibration direction of the motor bearing to be detected, the server is used to receive the sound wave frequency and the vibration direction, the The server includes a motor bearing fault detection model, and the motor bearing fault detection model is used to determine the fault type, fault level and remaining life of the motor to be detected according to the sound wave frequency and the vibration direction;
  • the information collection module is also used to judge whether the sound wave frequency is within the set threshold range, and if the sound wave frequency is within the set threshold range, the sound wave frequency and the vibration direction will be compared according to the first set frequency. upload to the server; if the sound wave frequency is not within the set threshold range, then upload the sound wave frequency and the vibration direction to the server according to the second set frequency; the second set frequency is higher than the first set frequency
  • the fixed frequency is high.
  • the information collection module includes a housing and an acoustic wave sensor, a vibration sensor, a temperature sensor, a humidity sensor, a communication unit, a storage unit and a controller arranged in the housing; the acoustic wave sensor, the vibration sensor, The temperature sensor, the humidity sensor, the communication unit and the storage unit are all connected to the controller;
  • the acoustic wave sensor is used to detect the sound frequency of the motor to be detected; the vibration sensor is used to detect the vibration direction of the motor to be detected; the temperature sensor is used to collect temperature signals, and the humidity sensor is used to collect humidity signals;
  • the storage unit is used to store the sound frequency, the vibration direction, the temperature signal and the humidity signal, and the controller is used to pass the sound frequency, the vibration direction, the temperature signal and the humidity signal through The communication unit sends to the server.
  • the server is further configured to send a feedback signal to the controller after receiving the sound frequency, the vibration direction, the temperature signal, and the humidity signal, and the controller is also configured to After receiving the feedback signal from the server, delete the sound frequency, the vibration direction, the temperature signal and the humidity signal that have been sent to the server in the storage unit.
  • the server is a cloud server.
  • the vibration sensor is a gyroscope.
  • an alarm module is also included, the alarm module is connected to the server, and the alarm module is used when the fault level output by the motor bearing fault detection model is greater than the set level value or the remaining life is greater than the set life , send out an alarm signal.
  • the invention discloses the following technical effects:
  • the present invention uploads the sound wave frequency and vibration direction of the motor bearing to be detected collected by the information collection module to the server.
  • the upload frequency is increased, and the fault type and fault are output through the motor bearing fault detection model in the server.
  • Level and remaining life reducing the influence of manual judgment, improving the accuracy and efficiency of fault detection.
  • Fig. 1 is a schematic flow chart of a method for detecting faults in the operating state of a motor bearing of the present invention
  • Fig. 2 is a schematic structural diagram of a motor bearing operating state fault detection system according to the present invention.
  • the object of the present invention is to provide a method and system for detecting faults in the operating state of motor bearings, which improves the accuracy of fault detection.
  • Fig. 1 is a schematic flow chart of a method for detecting a fault in the operating state of a motor bearing of the present invention. As shown in Fig. 1, a method for detecting a fault in the operating state of a motor bearing includes:
  • Step 101 Obtain the sound wave frequency and vibration direction of the motor bearing to be tested through the information collection module.
  • the information collection module includes an acoustic wave sensor and a gyroscope.
  • the acoustic wave frequency of the motor bearing to be detected is collected by the acoustic wave sensor, and the vibration direction of the motor bearing to be detected is collected by the gyroscope. Vibration amplitude is determined by the frequency of the sound waves.
  • Step 102 Determine whether the sound wave frequency is within a set threshold range.
  • step 103 is executed.
  • Step 103 Upload the sound wave frequency and vibration direction to the server according to the first set frequency.
  • step 104 is executed.
  • Step 104 Upload the sound wave frequency and vibration direction to the server according to the second set frequency; the second set frequency is higher than the first set frequency.
  • the server is a cloud server.
  • the data collected by the information collection module is continuously uploaded to the cloud server to realize real-time monitoring.
  • the specific process of processing the acoustic frequency includes:
  • xi(n) represents the collected acoustic signal
  • si(n) represents the original signal of the sound source
  • hi(n) represents the environmental impulse response between the sound source and the i-th industrial acoustic wave detector
  • wi(n) represents the i-th The environmental noise around the industrial acoustic wave detector
  • n represents the number of detection probes in the industrial acoustic wave detector.
  • the acoustic signal is preprocessed to obtain the signal after removing the DC component.
  • Discrete Fourier transform is performed on the preprocessed acoustic signal for spectrum analysis, and the subband energy ratio Si is extracted.
  • the sub-band energy ratio Si extracted from the signal received by the industrial sound wave detector does not belong to the [L, H] range, it is determined that an abnormal sound is detected. At this time, the sound wave frequency and the vibration direction are uploaded to the server according to the second set frequency.
  • the monitoring results are transmitted to the server, and the server continuously pushes the abnormal data in real time according to the big data analysis, continuously analyzes its characteristic curve, and judges the important abnormality and the secondary important abnormality through the motor bearing fault detection model.
  • the motor bearing fault detection model outputs a fault level of An alarm signal is sent when there is an important abnormality, and a reminder signal is sent when the output fault level of the motor bearing fault detection model is a minor abnormality.
  • Step 105 Input the sound wave frequency and vibration direction into the motor bearing fault detection model obtained through machine learning to obtain the fault type, fault level and remaining life.
  • the motor bearing fault detection model realizes accurate prediction of motor bearing faults, detects faults as early as possible, and reasonably predicts the remaining life of the machine.
  • a method for detecting faults in the operating state of a motor bearing also includes:
  • sample data of motor bearings includes the acoustic frequency, vibration direction, fault type, fault level and remaining life of motor bearings.
  • the sample data is enhanced by synthesizing data with simulation tools to obtain the enhanced sample data. Solve the problem of insufficient data and reduce model training time through data augmentation.
  • the convolutional neural network model is trained to obtain the motor bearing fault detection model.
  • the fault type, fault level, and remaining life are marked as labels by means of human labeling.
  • Fault types include: bearing eccentricity, bearing inner ring fault and bearing outer ring fault.
  • Fig. 2 is a schematic structural diagram of a motor bearing operating state fault detection system of the present invention.
  • the motor bearing operating state fault detection system includes: an information collection module 201 and a server 202;
  • the information collection module 201 is arranged on the motor casing to be detected, and the information collection module 201 is used to collect the sound wave frequency and the vibration direction of the motor bearing to be detected, and the server 202 is used to receive the sound wave frequency and the vibration direction, and the server 202 includes a motor bearing fault detection model,
  • the motor bearing fault detection model is used to determine the fault type, fault level and remaining life of the motor to be detected according to the sound wave frequency and vibration direction.
  • the information collection module 201 is also used to judge whether the sound wave frequency is within the set threshold range, if the sound wave frequency is within the set threshold value range, then upload the sound wave frequency and the vibration direction to the server 202 according to the first set frequency; If within the set threshold range, the sound wave frequency and vibration direction are uploaded to the server 202 according to the second set frequency; the second set frequency is higher than the first set frequency.
  • the information acquisition module 201 includes a housing and an acoustic sensor, a vibration sensor, a temperature sensor, a humidity sensor, a communication unit, a storage unit and a controller arranged in the housing; an acoustic sensor, a vibration sensor, a temperature sensor, a humidity sensor, a communication unit and a storage unit are connected to the controller.
  • the information collection module 201 also includes an antenna and a battery, the battery provides power for each component in the information collection module 201 , and the antenna is used for communication with the server 202 .
  • the shell adopts sound guide design and design method to prevent surrounding noise, and has a sound insulation layer of polymer material.
  • the communication unit uses 4G+WIFI, and there are many metals inside the factory, which is not suitable for 5G transmission.
  • the acoustic wave sensor is used to detect the sound frequency of the motor to be detected; the vibration sensor is used to detect the vibration direction of the motor to be detected; the temperature sensor is used to collect temperature signals, and the humidity sensor is used to collect humidity signals; the storage unit is used to store sound frequency, vibration direction, For the temperature signal and the humidity signal, the controller is used to send the sound frequency, vibration direction, temperature signal and humidity signal to the server 202 through the communication unit.
  • the server 202 is also used to send a feedback signal to the controller after receiving the sound frequency, vibration direction, temperature signal and humidity signal. 202 sound frequency, vibration direction, temperature signal and humidity signal.
  • the server 202 is a cloud server.
  • the vibration sensor is a gyroscope.
  • the motor bearing operating state fault detection system also includes an alarm module, which is connected to the server 202.
  • the alarm module is used to send an alarm signal when the fault level output by the motor bearing fault detection model is greater than the set level value or the remaining life is greater than the set life. .
  • the storage unit and the battery are to ensure that the data is not lost.
  • the operating data is sent at a low frequency (the first set frequency), for example, 5 minutes
  • the method of continuously sending data sending data at the first set frequency
  • the storage unit cache is needed to save the latest data that has not yet been transmitted.
  • the device after receiving the instruction from the cloud server to receive the data, it will send a feedback signal to the device (information collection module) and then delete the data, which solves the problem of accidents caused by accidents.
  • the data retention problem before the shutdown and the data transmission problem after the shutdown are caused by the shutdown.
  • the frequency data, amplitude and vibration direction of the sound wave are clarified.
  • Data can be recorded for machine learning.
  • Manual auscultation requires regular inspections, and the machine works 365*24 hours without cost, greatly reducing labor costs and the probability of failure.
  • the present invention can be based on artificial intelligence based on big data to assist the evaluation of the remaining life of the mechanical operation, with high accuracy and performance. Manual auscultation can only find problems, but cannot judge the severity of the accident and when the problem will occur.

Abstract

一种电机轴承运行状态故障检测方法及系统，包括：通过信息采集模块获得待检测电机轴承的声波频率和振动方向（S101）；判断该声波频率是否在设定阈值范围内（S102）；若该声波频率在设定阈值范围内，则按照第一设定频率将该声波频率和该振动方向上传至服务器（S103）；若该声波频率不在设定阈值范围内，则按照第二设定频率将该声波频率和该振动方向上传至服务器；该第二设定频率比该第一设定频率高（S104）；将该声波频率和该振动方向输入通过机器学习获得的电机轴承故障检测模型，获得故障类型、故障级别和剩余寿命（S105）。该方案提高了故障检测的准确性。

Description

一种电机轴承运行状态故障检测方法及系统 技术领域
本发明涉及故障预测技术领域，特别是涉及一种电机轴承运行状态故障检测方法及系统。
背景技术
目前，发电厂电机轴承的检测是通过人工听诊进行检测的，而人工听诊存在听诊工人劳动强度大，准确率低，听诊人才难培养的问题。而且人工听诊只是可以发现问题，但是不可以判断事故的严重性以及何时会出现问题。
发明内容
基于此，本发明的目的是提供一种电机轴承运行状态故障检测方法及系统，提高了故障检测的准确性。
为实现上述目的，本发明提供了一种电机轴承运行状态故障检测方法，包括：
通过信息采集模块获得待检测电机轴承的声波频率和振动方向；
判断所述声波频率是否在设定阈值范围内；
若所述声波频率在设定阈值范围内，则按照第一设定频率将所述声波频率和所述振动方向上传至服务器；
若所述声波频率不在设定阈值范围内，则按照第二设定频率将所述声波频率和所述振动方向上传至服务器；所述第二设定频率比所述第一设定频率高；
将所述声波频率和所述振动方向输入通过机器学习获得的电机轴承故障检测模型，获得故障类型、故障级别和剩余寿命。
可选地，还包括：
采集电机轴承样本数据，所述样本数据包括电机轴承的声波频率、振动方向、故障类型、故障级别和剩余寿命；
对所述样本数据进行数据增强，获得数据增强后的样本数据；
以数据增强后的样本数据中的声波频率和振动方向为输入，与声波频率和振动方向对应的故障类型、故障级别和剩余寿命为输出，训练卷积神 经网络模型，获得电机轴承故障检测模型。
可选地，所述故障类型包括：轴承偏心、轴承内环故障和轴承外环故障。
可选地，所述服务器为云服务器。
本发明还公开了一种电机轴承运行状态故障检测系统，所述电机轴承运行状态故障检测系统应用所述电机轴承运行状态故障检测方法，所述电机轴承运行状态故障检测系统包括：信息采集模块和服务器；
所述信息采集模块设置在待检测电机外壳上，所述信息采集模块用于采集待检测电机轴承的声波频率和振动方向，所述服务器用于接收所述声波频率和所述振动方向，所述服务器包括电机轴承故障检测模型，所述电机轴承故障检测模型用于根据所述声波频率和所述振动方向确定所述待检测电机的故障类型、故障等级和剩余寿命；
所述信息采集模块还用于判断所述声波频率是否在设定阈值范围内，若所述声波频率在设定阈值范围内，则按照第一设定频率将所述声波频率和所述振动方向上传至服务器；若所述声波频率不在设定阈值范围内，则按照第二设定频率将所述声波频率和所述振动方向上传至服务器；所述第二设定频率比所述第一设定频率高。
可选地，所述信息采集模块包括外壳和设置在所述外壳内的声波传感器、振动传感器、温度传感器、湿度传感器、通信单元、储存单元和控制器；所述声波传感器、所述振动传感器、所述温度传感器、所述湿度传感器、所述通信单元和所述储存单元均与所述控制器连接；
所述声波传感器用于检测待检测电机的声音频率；所述振动传感器用于检测待检测电机的振动方向；所述温度传感器用于采集温度信号，所述湿度传感器用于采集湿度信号；所述储存单元用存储所述声音频率、所述振动方向、所述温度信号和所述湿度信号，所述控制器用于将所述声音频率、所述振动方向、所述温度信号和所述湿度信号通过所述通信单元发送到所述服务器。
可选地，所述服务器还用于当接收到所述声音频率、所述振动方向、所述温度信号和所述湿度信号后向所述控制器发送反馈信号，所述控制器还用于当接收到服务器的反馈信号后，删除所述储存单元中已发送到所述 服务器的所述声音频率、所述振动方向、所述温度信号和所述湿度信号。
可选地，所述服务器为云服务器。
可选地，所述振动传感器为陀螺仪。
可选地，还包括报警模块，所述报警模块与所述服务器连接，所述报警模块用于当所述电机轴承故障检测模型输出的故障等级大于设定等级值或剩余寿命大于设定寿命时，发出报警信号。
根据本发明提供的具体实施例，本发明公开了以下技术效果：
本发明将信息采集模块采集的待检测电机轴承的声波频率和振动方向上传至服务器，当声波频率大于设定阈值时，提高上传频率，并通过服务器中的电机轴承故障检测模型输出故障类型、故障级别和剩余寿命，降低了人工判断影响，提高了故障检测的准确性和效率。
说明书附图
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1为本发明一种电机轴承运行状态故障检测方法流程示意图；
图2为本发明一种电机轴承运行状态故障检测系统结构示意图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
本发明的目的是提供一种电机轴承运行状态故障检测方法及系统，提高了故障检测的准确性。
为使本发明的上述目的、特征和优点能够更加明显易懂，下面结合附图和具体实施方式对本发明作进一步详细的说明。
图1为本发明一种电机轴承运行状态故障检测方法流程示意图，如图1所示，一种电机轴承运行状态故障检测方法，包括：
步骤101：通过信息采集模块获得待检测电机轴承的声波频率和振动方向。
信息采集模块包括声波传感器和陀螺仪。通过声波传感器采集待检测电机轴承的声波频率，通过陀螺仪采集待检测电机轴承的振动方向。通过声波频率确定振动幅度。
步骤102：判断声波频率是否在设定阈值范围内。
若声波频率在设定阈值范围内，执行步骤103。
步骤103：按照第一设定频率将声波频率和振动方向上传至服务器。
若声波频率不在设定阈值范围内，执行步骤104。
步骤104：按照第二设定频率将声波频率和振动方向上传至服务器；第二设定频率比第一设定频率高。
服务器为云服务器。
信息采集模块采集的数据持续上传至云服务器，实现实时监控。
对声波频率进行处理的具体过程包括：
声波检测器采集的声信号表示为：xi(n)＝hi(n)*s(n)+wi(n)。
xi(n)表示采集的声信号，si(n)表示声源原始信号，hi(n)表示声源与第i个工业声波检测器之间的环境冲激响应，wi(n)代表第i个工业声波检测器周围的环境噪声，n表示工业声波检测器中检测探头的个数。
对声信号预处理，得到去直流分量后的信号。
对预处理后的声信号进行离散傅里叶变换进行频谱分析，提取子带能量比Si。
设定检测阈值的上限H与下限L。
当工业声波检测器所接收的信号提取的子带能量比Si不属于[L，H]范围时，则判定检测到异常声音。此时，按照第二设定频率将声波频率和振动方向上传至服务器。
将监测结果传输至服务器，服务器根据大数据分析对异常数据要求实时连续推送，持续分析其特征曲线，通过电机轴承故障检测模型判定重要异常和次重要异常，当电机轴承故障检测模型输出故障等级为重要异常时发出报警信号，当电机轴承故障检测模型输出故障等级为次要异常时发出提醒信号。
步骤105：将声波频率和振动方向输入通过机器学习获得的电机轴承故障检测模型，获得故障类型、故障级别和剩余寿命。
电机轴承故障检测模型实现对电机轴承故障的精准预测，尽早发现故障，合理预估机械剩余寿命。
一种电机轴承运行状态故障检测方法还包括：
采集电机轴承样本数据，样本数据包括电机轴承的声波频率、振动方向、故障类型、故障级别和剩余寿命。
采用仿真工具合成数据对样本数据进行数据增强，获得数据增强后的样本数据。通过数据增强解决数据不足问题，减少模型训练时间。
以数据增强后的样本数据中的声波频率和振动方向为输入，与声波频率和振动方向对应的故障类型、故障级别和剩余寿命为输出，训练卷积神经网络模型，获得电机轴承故障检测模型。
故障类型、故障级别和剩余寿命作为标签通过人为标注的方式进行标注。
故障类型包括：轴承偏心、轴承内环故障和轴承外环故障。
图2为本发明一种电机轴承运行状态故障检测系统结构示意图，如图2所示，电机轴承运行状态故障检测系统包括：信息采集模块201和服务器202；
信息采集模块201设置在待检测电机外壳上，信息采集模块201用于采集待检测电机轴承的声波频率和振动方向，服务器202用于接收声波频率和振动方向，服务器202包括电机轴承故障检测模型，电机轴承故障检测模型用于根据声波频率和振动方向确定待检测电机的故障类型、故障等级和剩余寿命。
信息采集模块201还用于判断声波频率是否在设定阈值范围内，若声波频率在设定阈值范围内，则按照第一设定频率将声波频率和振动方向上传至服务器202；若声波频率不在设定阈值范围内，则按照第二设定频率将声波频率和振动方向上传至服务器202；第二设定频率比第一设定频率高。
信息采集模块201包括外壳和设置在外壳内的声波传感器、振动传感器、温度传感器、湿度传感器、通信单元、储存单元和控制器；声波传感器、振动传感器、温度传感器、湿度传感器、通信单元和储存单元均与控制器连接。信息采集模块201还包括天线和电池，电池为信息采集模块201中各个部件提供电源，天线用于与服务器202通讯。
外壳采取导音设计和防止周围噪音的设计方式，有高分子材料隔音层。
通信单元采用4G+WIFI，工厂内部金属多，不适合采用5G传输。
声波传感器用于检测待检测电机的声音频率；振动传感器用于检测待检测电机的振动方向；温度传感器用于采集温度信号，湿度传感器用于采集湿度信号；储存单元用存储声音频率、振动方向、温度信号和湿度信号，控制器用于将声音频率、振动方向、温度信号和湿度信号通过通信单元发送到服务器202。
服务器202还用于当接收到声音频率、振动方向、温度信号和湿度信号后向控制器发送反馈信号，控制器还用于当接收到服务器202的反馈信号后，删除储存单元中已发送到服务器202的声音频率、振动方向、温度信号和湿度信号。
服务器202为云服务器。
振动传感器为陀螺仪。
电机轴承运行状态故障检测系统还包括报警模块，报警模块与服务器202连接，报警模块用于当电机轴承故障检测模型输出的故障等级大于设定等级值或剩余寿命大于设定寿命时，发出报警信号。
储存单元和电池是为了保证数据不丢失，在正常时候，即边缘计算在合理区间的时候(声波频率在设定阈值范围内)，低频率(第一设定频率)发送运行数据，例如5分钟一次，当声波频率高于设定阀值时候，采取连续发送数据的方式(以第一设定频率发送数据)。无论何时都需要储存单元缓存来保存还没有发射的最新数据，同时在接到云服务器接到数据的指令后，给设备(信息采集模块)一个回馈信号再删去数据，解决了由于意外当机而形成的当机前的数据保留问题以及当机后的数据发射问题。
本发明一种电机轴承运行状态故障检测方法及系统的技术效果包括：
1.明确了声波频率数据，振幅以及振动方向。2.可以记录数据便于机器学习。3.由于系统对振动数据的采集，可以很好的规避振动噪音和环境噪音带来的干扰，分辨率更加准确。4.人工听诊需要定期巡防，机器无需成本365*24小时工作，大大减少人力成本和发生故障的概率。5.本发明可以基于大数据的人工智能辅助机械运行剩余寿命评估，准确性能高，人工听诊只是可以发现问题，但是不可以判断事故的严重性以及何时会出现问题。
本说明书中各个实施例采用递进的方式描述，每个实施例重点说明的都是与其他实施例的不同之处，各个实施例之间相同相似部分互相参见即可。
本文中应用了具体个例对本发明的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本发明的方法及其核心思想；同时，对于本领域的一般技术人员，依据本发明的思想，在具体实施方式及应用范围上均会有改变之处。综上所述，本说明书内容不应理解为对本发明的限制。

Claims (10)

  1. 一种电机轴承运行状态故障检测方法，其特征在于，包括：
    通过信息采集模块获得待检测电机轴承的声波频率和振动方向；
    判断所述声波频率是否在设定阈值范围内；
    若所述声波频率在设定阈值范围内，则按照第一设定频率将所述声波频率和所述振动方向上传至服务器；
    若所述声波频率不在设定阈值范围内，则按照第二设定频率将所述声波频率和所述振动方向上传至服务器；所述第二设定频率比所述第一设定频率高；
    将所述声波频率和所述振动方向输入通过机器学习获得的电机轴承故障检测模型，获得故障类型、故障级别和剩余寿命。
  2. 根据权利要求1所述的电机轴承运行状态故障检测方法，其特征在于，还包括：
    采集电机轴承样本数据，所述样本数据包括电机轴承的声波频率、振动方向、故障类型、故障级别和剩余寿命；
    对所述样本数据进行数据增强，获得数据增强后的样本数据；
    以数据增强后的样本数据中的声波频率和振动方向为输入，与声波频率和振动方向对应的故障类型、故障级别和剩余寿命为输出，训练卷积神经网络模型，获得电机轴承故障检测模型。
  3. 根据权利要求2所述的电机轴承运行状态故障检测方法，其特征 在于，所述故障类型包括：轴承偏心、轴承内环故障和轴承外环故障。
  4. 根据权利要求1所述的电机轴承运行状态故障检测方法，其特征在于，所述服务器为云服务器。
  5. 一种电机轴承运行状态故障检测系统，其特征在于，所述电机轴承运行状态故障检测系统应用权利要求1-4任意一项所述电机轴承运行状态故障检测方法，所述电机轴承运行状态故障检测系统包括：信息采集模块和服务器；
    所述信息采集模块设置在待检测电机外壳上，所述信息采集模块用于采集待检测电机轴承的声波频率和振动方向，所述服务器用于接收所述声波频率和所述振动方向，所述服务器包括电机轴承故障检测模型，所述电机轴承故障检测模型用于根据所述声波频率和所述振动方向确定所述待检测电机的故障类型、故障等级和剩余寿命；
    所述信息采集模块还用于判断所述声波频率是否在设定阈值范围内，若所述声波频率在设定阈值范围内，则按照第一设定频率将所述声波频率和所述振动方向上传至服务器；若所述声波频率不在设定阈值范围内，则按照第二设定频率将所述声波频率和所述振动方向上传至服务器；所述第二设定频率比所述第一设定频率高。
  6. 根据权利要求5所述的电机轴承运行状态故障检测系统，其特征在于，所述信息采集模块包括外壳和设置在所述外壳内的声波传感器、振动传感器、温度传感器、湿度传感器、通信单元、储存单元和控制器；所述声波传感器、所述振动传感器、所述温度传感器、所述湿度传感器、所 述通信单元和所述储存单元均与所述控制器连接；
    所述声波传感器用于检测待检测电机的声音频率；所述振动传感器用于检测待检测电机的振动方向；所述温度传感器用于采集温度信号，所述湿度传感器用于采集湿度信号；所述储存单元用存储所述声音频率、所述振动方向、所述温度信号和所述湿度信号，所述控制器用于将所述声音频率、所述振动方向、所述温度信号和所述湿度信号通过所述通信单元发送到所述服务器。
  7. 根据权利要求6所述的电机轴承运行状态故障检测系统，其特征在于，所述服务器还用于当接收到所述声音频率、所述振动方向、所述温度信号和所述湿度信号后向所述控制器发送反馈信号，所述控制器还用于当接收到服务器的反馈信号后，删除所述储存单元中已发送到所述服务器的所述声音频率、所述振动方向、所述温度信号和所述湿度信号。
  8. 根据权利要求5所述的电机轴承运行状态故障检测系统，其特征在于，所述服务器为云服务器。
  9. 根据权利要求6所述的电机轴承运行状态故障检测系统，其特征在于，所述振动传感器为陀螺仪。
  10. 根据权利要求5所述的电机轴承运行状态故障检测系统，其特征在于，还包括报警模块，所述报警模块与所述服务器连接，所述报警模块用于当所述电机轴承故障检测模型输出的故障等级大于设定等级值或剩余寿命大于设定寿命时，发出报警信号。
PCT/CN2021/105526 2021-07-09 2021-07-09 一种电机轴承运行状态故障检测方法及系统 WO2023279382A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/105526 WO2023279382A1 (zh) 2021-07-09 2021-07-09 一种电机轴承运行状态故障检测方法及系统

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/105526 WO2023279382A1 (zh) 2021-07-09 2021-07-09 一种电机轴承运行状态故障检测方法及系统

Publications (1)

Publication Number Publication Date
WO2023279382A1 true WO2023279382A1 (zh) 2023-01-12

Family

ID=84800185

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/105526 WO2023279382A1 (zh) 2021-07-09 2021-07-09 一种电机轴承运行状态故障检测方法及系统

Country Status (1)

Country Link
WO (1) WO2023279382A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115859210A (zh) * 2023-02-10 2023-03-28 宁波慧声智创科技有限公司 一种带电机设备停机尾音智能检测方法及系统
CN116857172A (zh) * 2023-09-05 2023-10-10 诺沃机械南通有限公司 一种基于大数据的水泵故障预测方法及系统
CN117409816A (zh) * 2023-12-14 2024-01-16 湖南华夏特变股份有限公司 一种基于声音信号的设备故障检测方法及系统

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103969046A (zh) * 2014-05-20 2014-08-06 北京康拓红外技术股份有限公司 一种与轮对轴承磨合机联用的轴承声学诊断系统和方法
CN105588720A (zh) * 2015-12-15 2016-05-18 广州大学 基于声信号形态分量分析的滚动轴承故障诊断装置及方法
CN109827777A (zh) * 2019-04-01 2019-05-31 哈尔滨理工大学 基于偏最小二乘法极限学习机的滚动轴承故障预测方法
US20190250066A1 (en) * 2016-09-19 2019-08-15 Schaeffler Technologies AG & Co. KG Monitoring method and monitoring apparatus for determining remaining life of a bearing
CN110987434A (zh) * 2019-12-18 2020-04-10 中国民航大学 一种基于去噪技术的滚动轴承早期故障诊断方法
US20200149993A1 (en) * 2018-11-13 2020-05-14 GM Global Technology Operations LLC Method and apparatus for monitoring a machine bearing on-vehicle
CN113049252A (zh) * 2021-03-25 2021-06-29 成都天佑路航轨道交通科技有限公司 一种列车轴承箱的故障检测方法

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103969046A (zh) * 2014-05-20 2014-08-06 北京康拓红外技术股份有限公司 一种与轮对轴承磨合机联用的轴承声学诊断系统和方法
CN105588720A (zh) * 2015-12-15 2016-05-18 广州大学 基于声信号形态分量分析的滚动轴承故障诊断装置及方法
US20190250066A1 (en) * 2016-09-19 2019-08-15 Schaeffler Technologies AG & Co. KG Monitoring method and monitoring apparatus for determining remaining life of a bearing
US20200149993A1 (en) * 2018-11-13 2020-05-14 GM Global Technology Operations LLC Method and apparatus for monitoring a machine bearing on-vehicle
CN109827777A (zh) * 2019-04-01 2019-05-31 哈尔滨理工大学 基于偏最小二乘法极限学习机的滚动轴承故障预测方法
CN110987434A (zh) * 2019-12-18 2020-04-10 中国民航大学 一种基于去噪技术的滚动轴承早期故障诊断方法
CN113049252A (zh) * 2021-03-25 2021-06-29 成都天佑路航轨道交通科技有限公司 一种列车轴承箱的故障检测方法

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115859210A (zh) * 2023-02-10 2023-03-28 宁波慧声智创科技有限公司 一种带电机设备停机尾音智能检测方法及系统
CN116857172A (zh) * 2023-09-05 2023-10-10 诺沃机械南通有限公司 一种基于大数据的水泵故障预测方法及系统
CN116857172B (zh) * 2023-09-05 2023-12-22 诺沃机械南通有限公司 一种基于大数据的水泵故障预测方法及系统
CN117409816A (zh) * 2023-12-14 2024-01-16 湖南华夏特变股份有限公司 一种基于声音信号的设备故障检测方法及系统
CN117409816B (zh) * 2023-12-14 2024-03-26 湖南华夏特变股份有限公司 一种基于声音信号的设备故障检测方法及系统

Similar Documents

Publication Publication Date Title
WO2023279382A1 (zh) 一种电机轴承运行状态故障检测方法及系统
CN110647133B (zh) 轨道交通设备状态检测维护方法及系统
CN103115789B (zh) 金属结构损伤剩余寿命的第二代小波支持向量机评估方法
CN104048825B (zh) 一种多传感器融合的风电机组齿轮箱故障定位方法
CN110177017B (zh) 一种减速机智能故障诊断系统及其诊断方法
CN111400961B (zh) 风力发电机组叶片故障判断方法及装置
CN103364189A (zh) 一种风电机组齿轮箱在线故障诊断系统
CN110095179A (zh) 一种电机振动检测与诊断方法
CN113298134B (zh) 一种基于bpnn的风机叶片远程非接触健康监测系统和方法
CN111351567A (zh) 一种基于加速度传感器的船舶设备振动状态监测装置及方法
CN113567162A (zh) 基于声传感器的风机故障智能诊断装置及方法
CN112052712B (zh) 一种电力设备状态监测与故障识别方法及系统
Sánchez et al. Wind Turbines Acoustic Inspections performed with UAV and sound frequency domain analysis
CN110056640B (zh) 基于加速度信号和边缘计算的减速机无线故障诊断方法
CN115238785A (zh) 基于图像融合和集成网络的旋转机械故障诊断方法和系统
CN111879479B (zh) 集输气管道非高斯噪声环境微泄漏信号监测系统及方法
CN113776794A (zh) 一种嵌入式边缘计算的故障诊断方法、装置和系统
CN212513308U (zh) 一种水上运输工具振动报警装置
Zhang et al. Fault diagnosis method of belt conveyor idler based on sound signal
CN114739667A (zh) 一种多模态信息融合轴承润滑状态监测装置及方法
CN203011692U (zh) 一种便携式滚动轴承故障诊断器
CN113218653B (zh) 一种基于消除低频扰动的海上风电齿轮箱监测方法及装置
CN113516023A (zh) 设备振动异常诊断方法和系统
Shen et al. Fault detection method based on multi-scale convolutional neural network for wind turbine gearbox
CN214330806U (zh) 风力发电机系统和风力发电机的故障分析设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21948876

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE