WO2021143128A1 - 一种基于深度学习的轴承状态识别方法及系统 - Google Patents

一种基于深度学习的轴承状态识别方法及系统 Download PDF

Info

Publication number
WO2021143128A1
WO2021143128A1 PCT/CN2020/109914 CN2020109914W WO2021143128A1 WO 2021143128 A1 WO2021143128 A1 WO 2021143128A1 CN 2020109914 W CN2020109914 W CN 2020109914W WO 2021143128 A1 WO2021143128 A1 WO 2021143128A1
Authority
WO
WIPO (PCT)
Prior art keywords
bearing
characteristic frequency
frequency
vibration signal
rolling
Prior art date
Application number
PCT/CN2020/109914
Other languages
English (en)
French (fr)
Inventor
张彩霞
胡绍林
王向东
Original Assignee
佛山科学技术学院
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 佛山科学技术学院 filed Critical 佛山科学技术学院
Publication of WO2021143128A1 publication Critical patent/WO2021143128A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction
    • G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
    • 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 diagnosis, in particular to a bearing state recognition method and system based on deep learning.
  • Rolling bearings are the most widely used in rotating machinery and are also one of the most vulnerable components. Many failures of rotating machinery are related to rolling bearings. The performance of the bearing has a great impact on the working state of the machinery. Its defects can cause abnormal vibration and noise of the equipment, and even damage the equipment in severe cases. Correct state monitoring and diagnosis of rolling bearings is an important aspect of equipment optimization management and predictive maintenance of modern enterprises.
  • Rolling bearing is a kind of precision mechanical component that changes the sliding friction between the running shaft and the shaft seat into rolling friction, thereby reducing friction loss.
  • Rolling bearings are generally composed of four parts: inner ring, outer ring, rolling elements and cage.
  • the function of the inner ring is to cooperate with the shaft and rotate with the shaft; the function of the outer ring is to cooperate with the bearing seat and play a supporting role;
  • the cage evenly distributes the rolling elements between the inner ring and the outer ring, and its shape, size and quantity directly affect the performance and life of the rolling bearing; the cage can evenly distribute the rolling elements and guide the rolling elements to rotate for lubrication.
  • the failure phenomenon of rolling bearings is generally manifested as excessive bearing temperature, bearing noise and bearing wear. Most of the bearing failures are not easy to detect. They can only be noticed when the machine has high temperature, large beating amplitude, abnormal noise, etc., but when people find out, most of the rolling shafts have worn out, causing the machine to stop.
  • the present invention provides a bearing state recognition method and system based on deep learning, which can recognize the bearing state in real time and accurately.
  • the present invention provides a bearing state recognition method based on deep learning, including:
  • the characteristic frequency includes: the characteristic frequency of the outer ring of the rolling bearing, the characteristic frequency of the inner ring of the rolling bearing, the characteristic frequency of the rolling element, and the characteristic frequency of the cage;
  • the determining the characteristic frequency of the bearing according to the physical parameters of the bearing is specifically:
  • the obtaining the vibration signal of the bearing and determining in real time whether the amplitude of the bearing deviates from the threshold range is specifically:
  • the crest factor of the vibration signal is obtained
  • the performing state recognition according to the deviation degree of the current operating frequency of the bearing from the characteristic frequency to obtain the current state of the bearing includes:
  • the fast Fourier transform is used to convert the vibration signal of the bearing into a frequency domain signal
  • the fault type of the characteristic frequency is taken as the state of the bearing.
  • the method further includes: when the vibration signal of the bearing deviates from the threshold range, determining the overall failure of the bearing.
  • the present invention also provides a bearing state recognition system based on deep learning.
  • the system includes a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor executes all The computer program runs in the modules of the following systems:
  • the characteristic frequency determination module is used to determine the characteristic frequency of the bearing according to the physical parameters of the bearing.
  • the characteristic frequency includes: the characteristic frequency of the outer ring of the rolling bearing, the characteristic frequency of the inner ring of the rolling bearing, the characteristic frequency of the rolling element, and the characteristic frequency of the cage;
  • the judgment module is used to obtain the vibration signal of the bearing and judge whether the amplitude of the bearing deviates from the threshold range in real time;
  • the state recognition module is used to perform state recognition according to the deviation degree of the current operating frequency of the bearing from the characteristic frequency when the amplitude of the bearing is within the threshold range, and obtain the current state of the bearing.
  • the present invention discloses a bearing state recognition method and system based on deep learning.
  • the method is: firstly determine the characteristic frequency of the bearing according to the physical parameters of the bearing, and the characteristic frequency includes: the outer ring failure of the rolling bearing The characteristic frequency, the characteristic frequency of the inner ring fault of the rolling bearing, the characteristic frequency of the rolling element fault, the characteristic frequency of the cage fault; then the vibration signal of the bearing is obtained to determine whether the amplitude of the bearing deviates from the threshold range in real time; when the amplitude of the bearing is within the threshold range, The degree of deviation between the current operating frequency of the bearing and the characteristic frequency is identified to obtain the current state of the bearing.
  • the invention can recognize the bearing state in real time and accurately.
  • FIG. 1 is a schematic flowchart of a method for recognizing a bearing state based on deep learning according to an embodiment of the present invention
  • Fig. 2 is a schematic structural diagram of a bearing state recognition system based on deep learning according to an embodiment of the present invention.
  • Figure 1 shows a bearing state recognition method based on deep learning, including the following steps:
  • Step S100 Determine the characteristic frequency of the bearing according to the physical parameters of the bearing; wherein the characteristic frequency includes: the characteristic frequency of the outer ring of the rolling bearing, the characteristic frequency of the inner ring of the rolling bearing, the characteristic frequency of the rolling element, and the characteristic frequency of the cage;
  • Step S200 Obtain the vibration signal of the bearing, and determine in real time whether the amplitude of the bearing deviates from the threshold range;
  • Step S300 When the amplitude of the bearing is within the threshold range, perform state recognition according to the deviation degree between the current operating frequency of the bearing and the characteristic frequency, and obtain the current state of the bearing.
  • the physical parameters of the bearing can be measured to determine the characteristic frequency of the bearing, and then the time-domain characteristics of the bearing can be analyzed to determine whether the amplitude of the bearing deviates from the threshold range, so as to initially determine whether the time-domain characteristics of the bearing are normal
  • the frequency domain characteristics of the bearing are analyzed, and the current operating frequency of the bearing is compared with the characteristic frequency to identify the current state of the bearing.
  • the technical solution can be quickly executed by software, it can be recognized in real time. It can be seen that the technical solution provided by the present disclosure can be real-time , Accurately identify the bearing status.
  • the characteristic frequency of the bearing is determined according to the physical parameters of the bearing in the step S100, specifically:
  • step S200 is specifically:
  • the crest factor of the vibration signal is calculated based on the average peak value and the root mean square value
  • N denote the number of sampling points of the vibration signal
  • n denote the number of segments of the vibration signal
  • ⁇ X Pj ⁇ denote the peak set in a segment of the vibration signal
  • the step S300 includes:
  • the fast Fourier transform is used to convert the vibration signal of the bearing into a frequency domain signal
  • the frequency domain signal can be decomposed by fast Fourier transform to obtain the frequency domain characteristics of the vibration signal, thereby obtaining multiple vibration frequency domains, and matching the multiple vibration frequency domains with the characteristic frequency , Calculate the deviation of the two, determine whether there is a characteristic frequency in the vibration signal, so as to obtain the state of the bearing when the matching condition is met.
  • the state of the bearing can be quickly learned.
  • the method further includes: when the vibration signal of the bearing deviates from the threshold range, determining the overall failure of the bearing.
  • an embodiment of the present invention also provides a bearing state recognition system based on deep learning.
  • the system includes a memory, a processor, and a computer program stored in the memory and running on the processor, The processor executes the computer program and runs in the modules of the following system:
  • the characteristic frequency determining module 100 is used to determine the characteristic frequency of the bearing according to the physical parameters of the bearing.
  • the characteristic frequency includes: the characteristic frequency of the outer ring of the rolling bearing, the characteristic frequency of the inner ring of the rolling bearing, the characteristic frequency of the rolling element, and the characteristic frequency of the cage. ;
  • the judging module 200 is used to obtain the vibration signal of the bearing and judge whether the amplitude of the bearing deviates from the threshold range in real time;
  • the state recognition module 300 is configured to perform state recognition according to the deviation degree of the current operating frequency of the bearing from the characteristic frequency when the amplitude of the bearing is within the threshold range, and obtain the current state of the bearing.
  • the processor may be a central processing unit (Central-Processing-Unit, CPU), or other general-purpose processors, digital signal processors (Digital-Signal-Processor, DSP), and application-specific integrated circuits (Application-Specific-Integrated -Circuit, ASIC), Field-Programmable-Gate-Array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc.
  • the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.
  • the processor is the control center of the deep learning-based bearing state recognition system, which uses various interfaces and lines to connect All parts of the bearing state recognition system based on deep learning.
  • the memory may be used to store the computer program and/or module, and the processor can implement the computer program and/or module stored in the memory by running or executing the computer program and/or module stored in the memory and calling the data stored in the memory to implement the computer program and/or module.
  • the memory may mainly include a program storage area and a data storage area.
  • the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, and a smart memory card (Smart-Media- Card, SMC), Secure-Digital (SD) card, Flash-Card, at least one magnetic disk storage device, flash memory device, or other volatile solid-state storage device.
  • Smart-Media- Card SMC
  • SD Secure-Digital

Abstract

本发明涉及故障诊断技术领域，具体涉及一种基于深度学习的轴承状态识别方法及系统，所述方法为：首先根据轴承的物理参数确定轴承的特征频率，所述特征频率包括：滚动轴承外圈故障特征频率、滚动轴承内圈故障特征频率、滚动体故障特征频率、保持架故障特征频率；接着获取轴承的振动信号，实时判断轴承的振幅是否偏离阈值范围；当轴承的振幅在阈值范围内时，根据所述轴承的当前运行频率与所述特征频率的偏离度进行状态识别，得出轴承的当前状态，本发明能够实时、准确的进行轴承状态识别。

Description

一种基于深度学习的轴承状态识别方法及系统 技术领域
本发明涉及故障诊断技术领域，具体涉及一种基于深度学习的轴承状态识别方法及系统。
背景技术
滚动轴承在旋转机械中应用最为广泛，同时也是最易损坏的元件之一。旋转机械的许多故障都与滚动轴承有关，轴承的工作好坏对机械的工作状态有很大影响，其缺陷会导致设备产生异常振动和噪声，严重时甚至损坏设备。对滚动轴承进行正确的状态监测及诊断，是现代化企业的设备优化管理及预知维修的一个重要方面。
滚动轴承是将运转的轴与轴座之间的滑动摩擦变为滚动摩擦，从而减少摩擦损失的一种精密的机械元件。滚动轴承一般由内圈、外圈、滚动体和保持架四部分组成，内圈的作用是与轴相配合并与轴一起旋转；外圈作用是与轴承座相配合，起支撑作用；滚动体是借助于保持架均匀的将滚动体分布在内圈和外圈之间，其形状大小和数量直接影响着滚动轴承的使用性能和寿命；保持架能使滚动体均匀分布，引导滚动体旋转起润滑作用。
滚动轴承的故障现象一般表现为轴承温度过高、轴承噪音和轴承磨损。大部分的轴承故障不易察觉，只有出现机器高温、跳动幅度大、异响等情况时，才会引起察觉，但是到人们发觉时，大部分滚动轴都已磨损，从而造成机器停机。
因此，为实现对轴承的可靠监测，预防轴承出现故障，降低事故发生情况，有必要提供一种实时、准确的轴承状态识别方案。
发明内容
为解决上述问题，本发明提供一种基于深度学习的轴承状态识别方法及系统，能够实时、准确的进行轴承状态识别。
为了实现上述目的，本发明提供以下技术方案：
本发明提供一种基于深度学习的轴承状态识别方法，包括：
根据轴承的物理参数确定轴承的特征频率，所述特征频率包括：滚动轴承外圈故障特征频率、滚动轴承内圈故障特征频率、滚动体故障特征频率、保持架故障特征频率；
获取轴承的振动信号，实时判断轴承的振幅是否偏离阈值范围；
当轴承的振幅在阈值范围内时，根据所述轴承的当前运行频率与所述特征频率的偏离度 进行状态识别，得出轴承的当前状态。
优选地，所述根据轴承的物理参数确定轴承的特征频率，具体为：
获取轴承中旋转轴的转频f n、轴承节径D、滚动体直径d、滚动体个数z、滚动轴承的接触角α；
通过以下公式计算滚动轴承外圈故障特征频率f o：
Figure PCTCN2020109914-appb-000001
通过以下公式计算滚动轴承内圈故障特征频率f i：
Figure PCTCN2020109914-appb-000002
通过以下公式计算滚动体故障特征频率f bsf：
Figure PCTCN2020109914-appb-000003
通过以下公式计算保持架故障特征频率f ftf：
Figure PCTCN2020109914-appb-000004
优选地，所述获取轴承的振动信号，实时判断轴承的振幅是否偏离阈值范围，具体为：
获取轴承的振动信号，以轴承的振动周期作为时间段，对进行分段，提取每段振动信号的峰值；
根据每段振动信号的峰值计算振动信号的平均峰值和均方根值；
根据平均峰值和均方根值得出振动信号的峰值因子；
当峰值因子小于1.5时，判定轴承的振幅在阈值范围内。
优选地，所述根据所述轴承的当前运行频率与所述特征频率的偏离度进行状态识别，得出轴承的当前状态，包括：
采用快速傅里叶变换将轴承的振动信号转换为频域信号；
将所述频域信号进行分解，得到多个频域特征；
分别将多个频域特征与所述特征频率进行比较，判断频域特征与所述特征频率的偏离度是否小于0.1；
若是，则将所述特征频率的故障类型作为所述轴承的状态。
优选地，所述方法还包括：当轴承的振动信号偏离阈值范围时，则判定所述轴承的整体故障。
本发明还提供一种基于深度学习的轴承状态识别系统，所述系统包括：存储器、处理器以及存储在所述存储器中并可在所述处理器上运行的计算机程序，所述处理器执行所述计算机程序运行在以下系统的模块中：
特征频率确定模块，用于根据轴承的物理参数确定轴承的特征频率，所述特征频率包括：滚动轴承外圈故障特征频率、滚动轴承内圈故障特征频率、滚动体故障特征频率、保持架故障特征频率；
判断模块，用于获取轴承的振动信号，实时判断轴承的振幅是否偏离阈值范围；
状态识别模块，用于当轴承的振幅在阈值范围内时，根据所述轴承的当前运行频率与所述特征频率的偏离度进行状态识别，得出轴承的当前状态。
本发明的有益效果是：本发明公开一种基于深度学习的轴承状态识别方法及系统，所述方法为：首先根据轴承的物理参数确定轴承的特征频率，所述特征频率包括：滚动轴承外圈故障特征频率、滚动轴承内圈故障特征频率、滚动体故障特征频率、保持架故障特征频率；接着获取轴承的振动信号，实时判断轴承的振幅是否偏离阈值范围；当轴承的振幅在阈值范围内时，根据所述轴承的当前运行频率与所述特征频率的偏离度进行状态识别，得出轴承的当前状态。本发明能够实时、准确的进行轴承状态识别。
附图说明
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是本发明实施例一种基于深度学习的轴承状态识别方法的流程示意图；
图2是本发明实施例一种基于深度学习的轴承状态识别系统的结构示意图。
具体实施方式
以下将结合实施例和附图对本公开的构思、具体结构及产生的技术效果进行清楚、完整的描述，以充分地理解本公开的目的、方案和效果。需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互组合。
参考图1，如图1所示为一种基于深度学习的轴承状态识别方法，包括以下步骤：
步骤S100、根据轴承的物理参数确定轴承的特征频率；其中，所述特征频率包括：滚动轴承外圈故障特征频率、滚动轴承内圈故障特征频率、滚动体故障特征频率、保持架故障特征频率；
步骤S200、获取轴承的振动信号，实时判断轴承的振幅是否偏离阈值范围；
步骤S300、当轴承的振幅在阈值范围内时，根据所述轴承的当前运行频率与所述特征频率的偏离度进行状态识别，得出轴承的当前状态。
本实施例中，可通过实现测量轴承的物理参数，以确定轴承的特征频率，接着，分析轴承的时域特征，判断轴承的振幅是否偏离阈值范围，以初步确定轴承的时域特征是否在正常状态，当轴承的时域特征正常时，对轴承的频域特征进行分析，将轴承的当前运行频率与所述特征频率进行比对，从而识别轴承的当前状态，在轴承状态识别过程中，可以在滚动轴承的故障现象在萌芽阶段就及时发现，从而准确的判断轴承的故障状态，由于本技术方案可通 过软件的方式进行快速执行，能够做到实时识别，可见，本公开提供的技术方案能够实时、准确的进行轴承状态识别。
在一个优选的实施例中，所述步骤S100中根据轴承的物理参数确定轴承的特征频率，具体为：
获取轴承中旋转轴的转频f n、轴承节径D、滚动体直径d、滚动体个数z、滚动轴承的接触角α；
通过以下公式计算滚动轴承外圈故障特征频率f o：
Figure PCTCN2020109914-appb-000005
通过以下公式计算滚动轴承内圈故障特征频率f i：
Figure PCTCN2020109914-appb-000006
通过以下公式计算滚动体故障特征频率f bsf：
Figure PCTCN2020109914-appb-000007
通过以下公式计算保持架故障特征频率f ftf：
Figure PCTCN2020109914-appb-000008
在一个优选的实施例中，所述步骤S200具体为：
(1)获取轴承的振动信号，以轴承的振动周期作为时间段，对进行分段，提取每段振动信号的峰值；
(2)根据每段振动信号的峰值计算振动信号的平均峰值和均方根值；
(3)根据平均峰值和均方根值得出振动信号的峰值因子；
(4)当所述峰值因子小于1.5时，判定轴承的振幅在阈值范围内。
在一个具体的示例中，设N表示对振动信号的采样点数量，n表示振动信号的分段数量，{X Pj}表示一段振动信号中的峰值集合；通过公式
Figure PCTCN2020109914-appb-000009
计算每段振动信号的峰值X P，通过公式
Figure PCTCN2020109914-appb-000010
计算每段振动信号的峰值，通过公式
Figure PCTCN2020109914-appb-000011
计算振动信号的峰值因子C。
在一个优选的实施例中，所述步骤S300包括：
首先，采用快速傅里叶变换将轴承的振动信号转换为频域信号；
接着，将所述频域信号进行分解，得到多个频域特征；
最后，分别将多个频域特征与所述特征频率进行比较，判断频域特征与所述特征频率的偏离度是否小于0.1，若是，则将所述特征频率对应的故障类型作为所述轴承的状态。
本实施例中，可通过快速傅里叶变换将所述频域信号进行分解，得到振动信号的频域特征，从而获得多个振动频域，将多个振动频域与所述特征频率进行匹配，计算二者的偏离度，判断振动信号中是否存在特征频率，从而在满足匹配条件的情况下得出轴承的状态。通过本 实施例提供的方法，可以快速的获知轴承的状态。
在一个优选的实施例中，所述方法还包括：当轴承的振动信号偏离阈值范围时，则判定所述轴承的整体故障。
当承的振动信号偏离阈值范围时，说明轴承的振幅已经过大，无需判断轴承的频域特征即可判断其整体已出现故障，从而快速的获知轴承的状态。
参考图2，本发明实施例还提供一种基于深度学习的轴承状态识别系统，所述系统包括：存储器、处理器以及存储在所述存储器中并可在所述处理器上运行的计算机程序，所述处理器执行所述计算机程序运行在以下系统的模块中：
特征频率确定模块100，用于根据轴承的物理参数确定轴承的特征频率，所述特征频率包括：滚动轴承外圈故障特征频率、滚动轴承内圈故障特征频率、滚动体故障特征频率、保持架故障特征频率；
判断模块200，用于获取轴承的振动信号，实时判断轴承的振幅是否偏离阈值范围；
状态识别模块300，用于当轴承的振幅在阈值范围内时，根据所述轴承的当前运行频率与所述特征频率的偏离度进行状态识别，得出轴承的当前状态。
可见，上述方法实施例中的内容均适用于本系统实施例中，本系统实施例所具体实现的功能与上述方法实施例相同，并且达到的有益效果与上述方法实施例所达到的有益效果也相同。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施例方法可借助软件的实现方式，以软件的形式加载到处理器中，以进行轴承状态识别。基于这样的理解，本发明的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来。
所述处理器可以是中央处理单元(Central-Processing-Unit，CPU)，还可以是其他通用处理器、数字信号处理器(Digital-Signal-Processor，DSP)、专用集成电路(Application-Specific-Integrated-Circuit，ASIC)、现场可编程门阵列(Field-Programmable-Gate-Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等，所述处理器是所述一种基于深度学习的轴承状态识别系统的控制中心，利用各种接口和线路连接整个基于深度学习的轴承状态识别系统的各个部分。
所述存储器可用于存储所述计算机程序和/或模块，所述处理器通过运行或执行存储在所述存储器内的计算机程序和/或模块，以及调用存储在存储器内的数据，实现所述基于深度学习的轴承状态识别系统的各种功能。所述存储器可主要包括存储程序区和存储数据区， 存储器可以包括高速随机存取存储器，还可以包括非易失性存储器，例如硬盘、内存、插接式硬盘，智能存储卡(Smart-Media-Card，SMC)，安全数字(Secure-Digital，SD)卡，闪存卡(Flash-Card)、至少一个磁盘存储器件、闪存器件、或其他易失性固态存储器件。
尽管本公开的描述已经相当详尽且特别对几个所述实施例进行了描述，但其并非旨在局限于任何这些细节或实施例或任何特殊实施例，而是应当将其视作是通过参考所附权利要求，考虑到现有技术为这些权利要求提供广义的可能性解释，从而有效地涵盖本公开的预定范围。此外，上文以发明人可预见的实施例对本公开进行描述，其目的是为了提供有用的描述，而那些目前尚未预见的对本公开的非实质性改动仍可代表本公开的等效改动。

Claims (6)

  1. 一种基于深度学习的轴承状态识别方法，其特征在于，包括：
    根据轴承的物理参数确定轴承的特征频率，所述特征频率包括：滚动轴承外圈故障特征频率、滚动轴承内圈故障特征频率、滚动体故障特征频率、保持架故障特征频率；
    获取轴承的振动信号，实时判断轴承的振幅是否偏离阈值范围；
    当轴承的振幅在阈值范围内时，根据所述轴承的当前运行频率与所述特征频率的偏离度进行状态识别，得出轴承的当前状态。
  2. 根据权利要求1所述的一种基于深度学习的轴承状态识别方法，其特征在于，所述根据轴承的物理参数确定轴承的特征频率，具体为：
    获取轴承中旋转轴的转频f n、轴承节径D、滚动体直径d、滚动体个数z、滚动轴承的接触角α；
    通过以下公式计算滚动轴承外圈故障特征频率f o：
    Figure PCTCN2020109914-appb-100001
    通过以下公式计算滚动轴承内圈故障特征频率f i：
    Figure PCTCN2020109914-appb-100002
    通过以下公式计算滚动体故障特征频率f bsf：
    Figure PCTCN2020109914-appb-100003
    通过以下公式计算保持架故障特征频率f ftf：
    Figure PCTCN2020109914-appb-100004
  3. 根据权利要求2所述的一种基于深度学习的轴承状态识别方法，其特征在于，所述获取轴承的振动信号，实时判断轴承的振幅是否偏离阈值范围，具体为：
    获取轴承的振动信号，以轴承的振动周期作为时间段，对进行分段，提取每段振动信号的峰值；
    根据每段振动信号的峰值计算振动信号的平均峰值和均方根值；
    根据平均峰值和均方根值得出振动信号的峰值因子；
    当峰值因子小于1.5时，判定轴承的振幅在阈值范围内。
  4. 根据权利要求3所述的一种基于深度学习的轴承状态识别方法，其特征在于，所述根据所述轴承的当前运行频率与所述特征频率的偏离度进行状态识别，得出轴承的当前状态，包括：
    采用快速傅里叶变换将轴承的振动信号转换为频域信号；
    将所述频域信号进行分解，得到多个频域特征；
    分别将多个频域特征与所述特征频率进行比较，判断频域特征与所述特征频率的偏离度是否小于0.1；
    若是，则将所述特征频率的故障类型作为所述轴承的状态。
  5. 根据权利要求4所述的一种基于深度学习的轴承状态识别方法，其特征在于，所述方法还包括：当轴承的振动信号偏离阈值范围时，则判定所述轴承的整体故障。
  6. 一种基于深度学习的轴承状态识别系统，其特征在于，所述系统包括：存储器、处理器以及存储在所述存储器中并可在所述处理器上运行的计算机程序，所述处理器执行所述计算机程序运行在以下系统的模块中：
    特征频率确定模块，用于根据轴承的物理参数确定轴承的特征频率，所述特征频率包括：滚动轴承外圈故障特征频率、滚动轴承内圈故障特征频率、滚动体故障特征频率、保持架故障特征频率；
    判断模块，用于获取轴承的振动信号，实时判断轴承的振幅是否偏离阈值范围；
    状态识别模块，用于当轴承的振幅在阈值范围内时，根据所述轴承的当前运行频率与所述特征频率的偏离度进行状态识别，得出轴承的当前状态。
PCT/CN2020/109914 2020-01-16 2020-08-19 一种基于深度学习的轴承状态识别方法及系统 WO2021143128A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010045618.5 2020-01-16
CN202010045618.5A CN111178327A (zh) 2020-01-16 2020-01-16 一种基于深度学习的轴承状态识别方法及系统

Publications (1)

Publication Number Publication Date
WO2021143128A1 true WO2021143128A1 (zh) 2021-07-22

Family

ID=70658075

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/109914 WO2021143128A1 (zh) 2020-01-16 2020-08-19 一种基于深度学习的轴承状态识别方法及系统

Country Status (2)

Country Link
CN (1) CN111178327A (zh)
WO (1) WO2021143128A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113899572A (zh) * 2021-09-18 2022-01-07 中电华创电力技术研究有限公司 一种汽轮发电机组转子失稳故障的诊断方法
CN114781450A (zh) * 2022-04-24 2022-07-22 华东交通大学 一种基于参数优化momeda-mia-cnn的列车滚动轴承状态识别方法
CN117150349A (zh) * 2023-10-31 2023-12-01 济南嘉宏科技有限责任公司 一种智能化设备基础故障自主定位与量化评估方法及系统

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111178327A (zh) * 2020-01-16 2020-05-19 佛山科学技术学院 一种基于深度学习的轴承状态识别方法及系统
CN111678698B (zh) * 2020-06-17 2022-03-04 沈阳建筑大学 一种基于声振信号融合的滚动轴承故障检测方法
CN111833318A (zh) * 2020-06-30 2020-10-27 佛山科学技术学院 一种基于图像处理的轴承异常检测方法及装置
CN111812511A (zh) * 2020-06-30 2020-10-23 佛山科学技术学院 一种基于大数据的电机故障诊断方法及装置
CN114565000B (zh) * 2022-04-28 2022-08-12 西安航天动力研究所 一种球轴承的故障确定方法、设备以及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2002022617A (ja) * 2000-07-05 2002-01-23 Mitsubishi Electric Corp 軸受診断装置
CN104535323A (zh) * 2015-01-12 2015-04-22 石家庄铁道大学 一种基于角域-时域-频域的机车轮对轴承故障诊断方法
CN110057583A (zh) * 2019-03-01 2019-07-26 西人马(西安)测控科技有限公司 一种轴承故障识别方法、装置及计算机设备
CN110672326A (zh) * 2019-09-29 2020-01-10 上海联影智能医疗科技有限公司 轴承故障检测方法与计算机可读存储介质
CN111178327A (zh) * 2020-01-16 2020-05-19 佛山科学技术学院 一种基于深度学习的轴承状态识别方法及系统

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106404396A (zh) * 2016-08-30 2017-02-15 四川中烟工业有限责任公司 一种滚动轴承的故障诊断方法
CN107560851B (zh) * 2017-08-28 2019-04-23 合肥工业大学 滚动轴承微弱故障特征早期提取方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2002022617A (ja) * 2000-07-05 2002-01-23 Mitsubishi Electric Corp 軸受診断装置
CN104535323A (zh) * 2015-01-12 2015-04-22 石家庄铁道大学 一种基于角域-时域-频域的机车轮对轴承故障诊断方法
CN110057583A (zh) * 2019-03-01 2019-07-26 西人马(西安)测控科技有限公司 一种轴承故障识别方法、装置及计算机设备
CN110672326A (zh) * 2019-09-29 2020-01-10 上海联影智能医疗科技有限公司 轴承故障检测方法与计算机可读存储介质
CN111178327A (zh) * 2020-01-16 2020-05-19 佛山科学技术学院 一种基于深度学习的轴承状态识别方法及系统

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113899572A (zh) * 2021-09-18 2022-01-07 中电华创电力技术研究有限公司 一种汽轮发电机组转子失稳故障的诊断方法
CN113899572B (zh) * 2021-09-18 2024-04-05 中电华创电力技术研究有限公司 一种汽轮发电机组转子失稳故障的诊断方法
CN114781450A (zh) * 2022-04-24 2022-07-22 华东交通大学 一种基于参数优化momeda-mia-cnn的列车滚动轴承状态识别方法
CN114781450B (zh) * 2022-04-24 2024-03-29 华东交通大学 一种基于参数优化momeda-mia-cnn的列车滚动轴承状态识别方法
CN117150349A (zh) * 2023-10-31 2023-12-01 济南嘉宏科技有限责任公司 一种智能化设备基础故障自主定位与量化评估方法及系统
CN117150349B (zh) * 2023-10-31 2024-02-02 济南嘉宏科技有限责任公司 一种智能化设备基础故障自主定位与量化评估方法及系统

Also Published As

Publication number Publication date
CN111178327A (zh) 2020-05-19

Similar Documents

Publication Publication Date Title
WO2021143128A1 (zh) 一种基于深度学习的轴承状态识别方法及系统
US10805151B2 (en) Method, apparatus, and storage medium for diagnosing failure based on a service monitoring indicator of a server by clustering servers with similar degrees of abnormal fluctuation
Purarjomandlangrudi et al. A data mining approach for fault diagnosis: An application of anomaly detection algorithm
CN109542740B (zh) 异常检测方法及装置
Sugumaran et al. Feature selection using decision tree and classification through proximal support vector machine for fault diagnostics of roller bearing
Jia et al. A deviation based assessment methodology for multiple machine health patterns classification and fault detection
Hu et al. Dynamic degradation observer for bearing fault by MTS–SOM system
WO2022134495A1 (zh) 基于深度学习的设备异常检测方法、系统及计算机介质
CA2865761C (en) Time series analytics
EP3788328B1 (en) System and process for pattern matching bearing vibration diagnostics
Ma et al. A novel probability confidence CNN model and its application in mechanical fault diagnosis
CN113804388A (zh) 基于时域分析的机械设备旋转冲击故障检测方法及系统
CN117195139A (zh) 基于机器学习的慢性病健康数据动态监测方法
CN117193164B (zh) 一种数控机床的故障监测方法及系统
WO2021042749A1 (zh) 一种基于有监督lle算法的轴承故障诊断方法及装置
CN115861318A (zh) 一种棉花加工生产质量检测方法
CN114861749A (zh) 一种基于深度原型网络的少样本轴承故障诊断方法
Tao et al. A novel model of one-class bearing fault detection using SVDD and genetic algorithm
CN114136617A (zh) 高速主轴轴承动态监测方法
TW201825229A (zh) 綜合加工機之迴轉體預兆診斷方法
Hu et al. Incipient machine degradation assessment by MF-MTS system
Almatheel et al. Bearing Element Fault Diagnosis Using Support Vector Machine
WO2019140613A1 (en) Method, apparatus and system for wind converter management
Zhu et al. Multi-Class Imbalanced Fault Diagnosis Method Based on IMWMOTE and MFO-Bayes-LS-SVM
CN110836786B (zh) 机械故障监测方法、装置、系统、介质和计算设备

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: 20913317

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20913317

Country of ref document: EP

Kind code of ref document: A1