CN116448236B

CN116448236B - Edge-end vibration monitoring system and method, and computer-readable storage medium

Info

Publication number: CN116448236B
Application number: CN202310729174.0A
Authority: CN
Inventors: 贾维银; 冯坤; 宋海峰; 郝文平; 王勇; 宗承治
Original assignee: Anhui Ronds Science & Technology Inc Co
Current assignee: Anhui Ronds Science & Technology Inc Co
Priority date: 2023-06-20
Filing date: 2023-06-20
Publication date: 2023-09-12
Anticipated expiration: 2043-06-20
Also published as: CN116448236A

Abstract

The application provides an edge-end vibration monitoring system and method and a computer readable storage medium. The edge end vibration monitoring system comprises an upper computer and an edge computing and collecting system; the edge calculation acquisition system acquires time domain waveform data of a vibration signal generated by mechanical equipment, calculates first frequency domain data according to the time domain waveform data, determines second frequency domain data positioned in a part of frequency bands in the first frequency domain data, and sends the second frequency domain data to the upper computer; the first frequency domain class data comprises at least one of a frequency spectrum, an envelope spectrum, an order spectrum and an order envelope spectrum of the vibration signal, wherein the frequency spectrum comprises an amplitude spectrum and/or a power spectrum; and the upper computer monitors the fault condition of the mechanical equipment according to the second frequency domain data. The system monitors vibration based on frequency domain data, is beneficial to reducing the calculated amount, storage and uploading pressure of the data in the monitoring process and relieves the pressure of an upper computer.

Description

Edge-end vibration monitoring system and method, and computer-readable storage medium

Technical Field

The application relates to the technical field of vibration monitoring, in particular to an edge-end vibration monitoring system and method and a computer readable storage medium.

Background

The vibration monitoring of mechanical equipment mainly utilizes state quantity data such as vibration, temperature and the like acquired by a data acquisition system at the edge end, an equipment nursing system is formed by combining an intelligent alarm system and a manual judgment mode, and the data acquisition system arranged at the edge end provides a key data basis for the nursing system.

With the development of edge calculation, the requirement on the length of waveform data is longer and longer, and the frequency is higher and higher. The waveform data itself has complete fault data information, but longer waveform data will have higher and higher requirements on data storage, uploading, calculation and the like of the monitoring system. In addition, the index data can be used for recording certain critical data information by comparing the waveform data, and the pressure of uploading the waveform data is reduced, but partial information is lost in the index.

Disclosure of Invention

An objective of the embodiments of the present application is to provide a system and a method for monitoring edge vibration, and a computer readable storage medium for improving the above technical problems. The specific technical scheme is as follows:

in a first aspect, an embodiment of the present application provides an edge-end vibration monitoring system, including: the upper computer and the edge computing and collecting system; the edge calculation acquisition system is used for acquiring time domain waveform data of vibration signals generated by mechanical dynamic equipment, calculating first frequency domain data according to the time domain waveform data, determining second frequency domain data in a part of frequency bands in the first frequency domain data, and sending the second frequency domain data to the upper computer; wherein the first frequency domain class data comprises at least one of a frequency spectrum, an envelope spectrum, an order spectrum and an order envelope spectrum of the vibration signal, the frequency spectrum comprising an amplitude spectrum and/or a power spectrum; and the upper computer is used for monitoring the fault condition of the mechanical equipment according to the second frequency domain data.

In the scheme, the upper computer performs vibration monitoring according to the frequency domain data of the partial frequency band sent by the edge calculation acquisition system, on one hand, because the frequency domain data contains most of the time domain waveform data and information critical to fault analysis, and the information loss is less, the operations of secondary index calculation, secondary alarm strategy application, algorithm iteration and the like related to the vibration monitoring can be well performed based on the frequency domain data; on the other hand, the data volume of the frequency domain data is smaller than that of the corresponding time domain waveform data, and only a part of the frequency domain data is transmitted by the scheme, so that the data volume transmitted by the edge computing acquisition system to the upper computer is reduced, the data storage pressure is reduced, or high-density data transmission is facilitated; on the other hand, if the upper computer needs to perform secondary calculation based on the data sent by the edge calculation acquisition system in the vibration monitoring process, the calculation based on the frequency domain data can avoid the operation with higher calculation amount such as Fast Fourier Transform (FFT) and the like, so that the calculation pressure of the upper computer is effectively relieved.

Therefore, the second frequency domain data is selected to be uploaded to the upper computer, so that the requirement of fault monitoring on the data is met, and the data processing pressure of the monitoring system is effectively reduced.

In an implementation form of the first aspect, the vibration signal comprises a velocity signal and/or an acceleration signal. The edge end vibration monitoring system can perform vibration monitoring through the speed signal and/or the acceleration signal, so that the edge end vibration monitoring system can be suitable for more application scenes, and the applicability of the edge end vibration monitoring system is improved.

In an implementation manner of the first aspect, if the mechanical moving device is a variable speed device, the first frequency domain class data includes the order spectrum and/or the order envelope spectrum. The inventor researches and discovers that for the variable speed equipment, the frequency spectrum is easy to be in a pile-shaped structure, which is not beneficial to fault analysis and index extraction calculation, and the effect of adopting the clear order spectrum and/or the order envelope spectrum is better.

In an implementation manner of the first aspect, when the first frequency domain class data includes the frequency spectrum, the edge computing and collecting system is configured to determine a portion of the frequency spectrum located in a first low frequency band as second frequency domain class data corresponding to the frequency spectrum; wherein the frequency range of the first low frequency range is from W1Hz to W2Hz, W1 and W2 are configurable values, W1 is more than 0, and W2 is more than or equal to 5000 and more than W1; and determining a portion of the envelope spectrum located within a second low frequency band as second frequency domain class data corresponding to the envelope spectrum when the first frequency domain class data includes the envelope spectrum; wherein the frequency range of the second low frequency range is from W3Hz to W4Hz, W3 and W4 are configurable values, W3 is more than 0, and 5000 is more than or equal to W4 and more than W3. In the above scheme, the edge calculation acquisition system only transmits the spectrum data in the first low frequency band and the envelope spectrum data in the second low frequency band to the upper computer, so that on one hand, the data transmission amount and the calculation amount are reduced, and on the other hand, most of fault characteristic information is contained in the first low frequency band of the spectrum and the second low frequency band of the envelope spectrum.

In one implementation manner of the first aspect, the values of W2 and W4 are one of the following values: 100. 500, 1000, 2000 and 5000.

In an implementation manner of the first aspect, if the first frequency domain class data includes the envelope spectrum, the edge calculation acquisition system is configured to calculate the envelope spectrum by: calculating the frequency spectrum according to the time domain waveform data; determining each sub-band of the target demodulation frequency band; the frequency range of the target demodulation frequency band is configurable, if the number of the sub-frequency bands is one, the sub-frequency band is the target demodulation frequency band, and if the number of the sub-frequency bands is a plurality of, each sub-frequency band is generated by dividing the target demodulation frequency band, and the frequency range of each sub-frequency band is configurable; and respectively carrying out envelope demodulation on the frequency spectrums in each sub-frequency band to obtain envelope spectrums in each sub-frequency band. The target demodulation frequency band in the scheme can be freely configured, so that the envelope spectrum can be acquired as required.

In an implementation manner of the first aspect, the frequency range of the target demodulation frequency band is from 1000Hz to 20000Hz, and the sub-frequency bands of the target demodulation frequency band include frequency bands from 0.1Hz to 100Hz, 100 to 1000Hz, 1000Hz to 5000Hz, frequency bands from 5000Hz to 10000Hz, frequency bands from 10000Hz to 15000Hz, and frequency bands from 15000Hz to 20000 Hz.

In an implementation manner of the first aspect, if the first frequency domain class data includes the frequency spectrum and the envelope spectrum, and the second frequency domain class data corresponding to the frequency spectrum is a portion of the frequency spectrum located in a first low frequency band, and a frequency range of the first low frequency band is from W1Hz to W2Hz, W1 and W2 are configurable values, W1 > 0, 5000 is greater than or equal to W2 > W1, then a frequency range of the target demodulation frequency band is from W2Hz to W5Hz, W5 is a configurable value, and W5 > W2. The scheme realizes the combination of the low-frequency transmission spectrum and the high-frequency transmission envelope spectrum, and is beneficial to improving the fault monitoring effect of most mechanical moving equipment.

In an implementation manner of the first aspect, when the first frequency domain class data includes the order spectrum, the edge computing and collecting system is configured to determine a portion of the order spectrum located in a first low-order segment as second frequency domain class data corresponding to the order spectrum; the first low-order section has an order range from M1 order to M2 order, M1 and M2 are configurable values, M1 is more than 0, 5000/reference frequency is more than or equal to M2 and more than M1, and the reference frequency is a frequency corresponding to 1 order in the order spectrum; and determining, when the first frequency domain class data includes the order envelope spectrum, a portion of the order envelope spectrum within a second lower order segment as second frequency domain class data corresponding to the order envelope spectrum; wherein the order of the second low-order section ranges from M3 order to M4 order, M3 and M4 are configurable values, M3 is more than 0, and 5000/reference frequency is more than or equal to M4 and more than M3. In the scheme, the edge computing and collecting system only transmits the order spectrum data in the first low-order section and the order envelope spectrum data in the second low-order section to the upper computer, so that the data transmission quantity and the calculation quantity are reduced on one hand, and on the other hand, most of fault characteristic information is contained in the first low-order section of the order spectrum and the second low-order section of the order envelope spectrum.

In one implementation manner of the first aspect, the values of M2 and M4 are one of the following values: 100/reference frequency, 500/reference frequency, 1000/reference frequency, 2000/reference frequency, and 5000/reference frequency.

In an implementation manner of the first aspect, if the first frequency domain class data includes the order envelope spectrum, the edge calculation acquisition system is configured to calculate the order envelope spectrum by: calculating the frequency spectrum according to the time domain waveform data; calculating the order spectrum according to the frequency spectrum and the rotating speed time scale data of the mechanical moving equipment; determining each sub-order of the target demodulation order segment; wherein, the order range of the target demodulation order section is configurable, if the number of the sub-order sections is one, the sub-order section is the target demodulation order section, if the number of the sub-order sections is a plurality of sub-order sections, each sub-order section is generated by dividing the target demodulation order section, and the order range of each sub-order section is configurable; and respectively carrying out envelope demodulation on the order spectrums in each sub-order section to obtain order envelope spectrums in each sub-order section. The target demodulation stage in the scheme can be freely configured, so that the order envelope spectrum can be acquired as required.

In one implementation manner of the first aspect, the order of the target demodulation order section ranges from (1000/fc) order to (20000/fc) order, and the sub-order section of the target demodulation order section includes an order section from (1000/fc) order to (5000/fc) order, an order section from (5000/fc) order to (10000/fc) order, an order section from (10000/fc) order to (15000/fc) order, and an order section from (15000/fc) order to (20000/fc) order, where fc is a reference frequency, and fc is a frequency corresponding to 1 order in the order spectrum.

In an implementation manner of the first aspect, if the first frequency domain class data includes the order spectrum and the order envelope spectrum, and the second frequency domain class data corresponding to the order spectrum is a portion of the order spectrum located in a first low-order segment, and an order range of the first low-order segment is from M1 order to M2 order, M1 and M2 are configurable values, M1 > 0, 5000/reference frequency is greater than or equal to M2 > M1, then an order range of the target demodulation order segment is from M2 order to M5 order, M5 is a configurable value, M5 > M2, and the reference frequency is a frequency corresponding to 1 order in the order spectrum. The scheme realizes the combination of the low-order transmission order spectrum and the high-order transmission order envelope spectrum, and is beneficial to improving the fault monitoring effect of most mechanical moving equipment (particularly speed changing equipment).

In an implementation manner of the first aspect, the edge computing and collecting system is further configured to send the time domain waveform data to the upper computer; the upper computer is used for monitoring the fault condition of the mechanical equipment according to the time domain waveform data and the second frequency domain class data; the period of the edge computing and collecting system for sending the second frequency domain data to the upper computer is smaller than the period of the time domain waveform data to the upper computer. The above scheme also monitors equipment faults through the time domain waveform data, the time domain waveform data is the most original data collected, contains the most abundant information, and can sometimes improve the monitoring effect, but the time domain waveform data can be uploaded according to lower frequency because of larger data quantity, and the second frequency domain data can be uploaded according to higher frequency because of smaller data quantity.

In one implementation manner of the first aspect, the period of sending the second frequency domain data to the upper computer by the edge computing and collecting system is m minutes, m is a configurable value, and 120 is greater than or equal to m is greater than or equal to 1.

In one implementation manner of the first aspect, the value of m is one of the following values: 5. 10, 15, 20, 25, 30, 60, 120.

In an implementation manner of the first aspect, the edge computing and collecting system is further configured to compute first index class data according to the time domain waveform data, and send the first index class data to the upper computer; wherein the first index class data comprises a first single class index and/or a first comprehensive class index; and the device is used for judging whether the mechanical equipment needs to perform fault alarm according to the first index type data and an edge end alarm strategy, and sending a corresponding alarm result to the upper computer if the mechanical equipment needs to perform fault alarm; the upper computer is used for monitoring the fault condition of the mechanical movable equipment according to the first index class data and the second frequency domain class data when receiving the first index class data; or the method is used for monitoring the fault condition of the mechanical movable equipment according to the first index class data, the alarm result and the second frequency domain class data when the first index class data and the alarm result are received. Compared with the mode of only calculating the comprehensive index in the related art, the first index type data calculated by the edge calculation acquisition system can comprise the first single type index and/or the first comprehensive type index, so that the upper computer can freely perform combined calculation according to the first single type index, and the calculation amount of the edge calculation acquisition system is reduced.

In an implementation manner of the first aspect, the upper computer is configured to calculate second index class data according to the second frequency domain class data, and determine whether to perform fault alarm of the mechanical moving device according to the second index class data and an alarm policy of the upper computer; wherein the second index class data comprises a second single class index and/or a second comprehensive class index. The upper computer in the scheme is provided with the upper computer alarm strategy, judges whether the fault alarm of the mechanical equipment is required according to the second index data calculated by the upper computer and the upper computer alarm strategy, relieves the calculation pressure of the edge calculation acquisition system, and can well support the index calculation because the second frequency domain data itself contains most of information in the time domain waveform data.

In an implementation manner of the first aspect, the upper computer is configured to calculate a third comprehensive class index according to the first single class index combination, and determine whether to perform fault alarm of the mechanical moving device according to the third comprehensive class index and an alarm policy of the upper computer. According to the scheme, the upper computer can freely perform combination calculation according to the first single-class index to perform secondary fault alarm, and the equipment fault condition is monitored more accurately.

In one implementation manner of the first aspect, the edge computing acquisition system is a wired acquisition system or a wireless acquisition system. The edge end vibration monitoring system in the scheme is not limited in data acquisition mode of the edge calculation acquisition system, and has a wide application range.

In an implementation manner of the first aspect, if the edge computing and acquiring system is a wireless acquiring system, the edge computing and acquiring system includes a communication station and a wireless vibration sensor installed on the mechanical moving device; the wireless vibration sensor is used for acquiring the time domain waveform data and carrying out data interaction with the communication station in a wireless communication mode, the communication station is used for carrying out data interaction with the upper computer, and edge calculation in the vibration monitoring process is independently carried out by the wireless vibration sensor, or is independently carried out by the communication station, or is shared and carried out by the communication station and the wireless vibration sensor. In the scheme, the edge calculation can be distributed between the communication station and the wireless vibration sensor according to the need, so that the flexibility is high.

In an implementation manner of the first aspect, the upper computer is configured to perform at least one of the following operations: performing data display based on the second frequency domain data, performing signal analysis according to the second frequency domain data, and performing signal confirmation according to the second frequency domain data; wherein the data presentation operation comprises presenting at least one of: the second frequency domain data comprises frequency band information corresponding to the second frequency domain data, envelope segment information corresponding to the second frequency domain data and association relation between index data and the second frequency domain data. In the scheme, the upper computer can perform a series of monitoring operations based on the second frequency domain data, and has a relatively perfect monitoring function.

In one implementation of the first aspect, the system further includes an algorithm platform; the upper computer is also used for sending the second frequency domain data to the algorithm platform; the algorithm platform is used for iterating a model and/or algorithm used by the edge computing and collecting system according to the second frequency domain data; the model used by the edge computing and collecting system comprises an equipment model of the mechanical equipment, and the algorithm used by the edge computing and collecting system comprises at least one of an index data computing algorithm, a data collecting strategy, an edge end alarming strategy and a data uploading strategy. According to the scheme, the algorithm platform iterates the model and/or the algorithm according to the second frequency domain data, so that the iteration speed and the iteration quality of the model and/or the algorithm are improved, and the standardization of the index data calculation algorithm, the improvement of the data acquisition strategy, the edge end alarm strategy and the data uploading strategy are promoted.

In a second aspect, an embodiment of the present application provides a method for monitoring edge-end vibration, including:

the edge calculation acquisition system acquires time domain waveform data of vibration signals generated by mechanical dynamic equipment;

The edge computing and collecting system computes first frequency domain data according to the time domain waveform data; wherein the first frequency domain class data comprises at least one of a frequency spectrum, an envelope spectrum, an order spectrum and an order envelope spectrum of the vibration signal, the frequency spectrum comprising an amplitude spectrum and/or a power spectrum;

the edge computing and collecting system determines second frequency domain data in a partial frequency band in the first frequency domain data and sends the second frequency domain data to an upper computer;

and the upper computer monitors the fault condition of the mechanical movable equipment according to the second frequency domain data.

In a third aspect, embodiments of the present application provide a computer readable storage medium having stored thereon computer program instructions which, when read and executed by a processor, perform the method provided by the first aspect or any one of the possible implementations of the first aspect.

In a fourth aspect, embodiments of the present application also provide a computer program product comprising computer program instructions which, when read and executed by a processor, perform the method provided by the first aspect or any one of the possible implementations of the first aspect.

Additional features and advantages of the application will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of the embodiments of the application. The objectives and other advantages of the application will be realized and attained by the structure particularly pointed out in the written description and claims thereof as well as the appended drawings.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and should not be considered as limiting the scope, and other related drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a schematic structural diagram of an edge-end vibration monitoring system according to an embodiment of the present application;

fig. 2 is a flow chart of an edge vibration monitoring method according to an embodiment of the present application.

Detailed Description

Embodiments of the technical scheme of the present application will be described in detail below with reference to the accompanying drawings. The following examples are only for more clearly illustrating the technical aspects of the present application, and thus are merely examples, and are not intended to limit the scope of the present application.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having" and any variations thereof in the description of the application and the claims and the description of the drawings above are intended to cover a non-exclusive inclusion.

In the description of embodiments of the present application, the technical terms "first," "second," and the like are used merely to distinguish between different objects and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated, a particular order or a primary or secondary relationship. In the description of the embodiments of the present application, the meaning of "plurality" is two or more unless explicitly defined otherwise.

Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments.

The mechanical power device refers to a rotating device driven by a driving machine, namely, a device with energy consumption, such as a pump, a compressor, a fan and the like, and the consumed energy can be electric power, aerodynamic force, steam power and the like. Generally, they can be classified into fluid transport machines, heterogeneous separation machines, stirring and mixing machines, freezing machines, crystallization and drying equipment, and the like. The static equipment corresponds to mechanical equipment, and refers to non-rotating or moving equipment without driving machine, such as chemical reactors, towers, heat exchange equipment, separation equipment, storage equipment and other equipment.

At present, in the related art, when vibration monitoring is performed on mechanical moving equipment, a mode that an upper computer is matched with edge computing equipment is generally adopted, wherein the edge computing equipment is used for providing time domain waveform data for vibration monitoring of the upper computer, and the upper computer performs vibration monitoring according to the time domain waveform data provided by the edge computing equipment. The upper computer is insufficient in vibration monitoring according to time domain waveform data or index data, the time domain waveform data has high requirements on data storage, uploading, calculation and the like of a monitoring system of the system, and the index data is only critical data, so that the problems of data loss, fault report missing and the like can be caused.

Based on this, the embodiment of the application provides an edge vibration monitoring system, in which an edge computing and collecting system calculates first frequency domain data according to time domain waveform data of a collected vibration signal, and sends frequency domain data of a part of frequency bands in second frequency domain data to an upper computer, so that the upper computer can monitor fault conditions of mechanical equipment according to the second frequency domain data.

In addition, if the upper computer in the embodiment of the application needs to perform secondary calculation based on the data sent by the edge calculation acquisition system in the vibration monitoring process, the calculation based on the second frequency domain data can avoid the operation with higher calculation amount such as Fast Fourier Transform (FFT) and the like, so that the calculation pressure of the upper computer is relieved; and because the second frequency domain data contains most of information critical to fault analysis in the time domain waveform data, the information loss is less, and therefore, operations such as secondary index calculation, secondary alarm strategy application, algorithm iteration and the like related to vibration monitoring can be well performed based on the frequency domain data. Before describing the specific content of the edge-end vibration monitoring system provided by the embodiment of the present application, explanation is first made on the related terms related to the embodiment of the present application:

Data density: it is understood that the frequency or period of data collection, such as several hours or minutes, collects a piece of data;

data resolution: the method refers to the fineness of single data, for example, the more seconds a piece of data is continuously collected, the more the number of seconds is, the higher the data resolution is;

time and frequency domains: the time domain and the frequency domain are two concepts in signal processing, which describe characteristics of signals on a time axis and a frequency axis respectively, wherein a time domain signal can be represented by a waveform diagram and the like, a frequency domain signal can be represented by a spectrogram and the like, the time domain signal can intuitively reflect the characteristics of periodicity, amplitude, phase and the like of the signals, the spectrogram of the frequency domain signal can intuitively reflect the characteristics of frequency distribution, amplitude spectrum, phase spectrum and the like of the signals, and the time domain signal can be converted into the frequency domain signal by a Fourier transform and the like.

Referring to fig. 1, an embodiment of the present application provides an edge-end vibration monitoring system 100, which includes: the host computer 110 and the edge computing and collecting system 120, wherein:

as shown in fig. 1, the edge computing and collecting system 120 is configured to collect time domain waveform data of a vibration signal generated by a mechanical device, calculate first frequency domain class data according to the time domain waveform data, determine second frequency domain class data located in a part of frequency bands in the first frequency domain class data, and send the second frequency domain class data to the upper computer 110; the first frequency domain data comprises at least one of a frequency spectrum, an envelope spectrum, an order spectrum and an order envelope spectrum of the vibration signal, wherein the frequency spectrum comprises an amplitude spectrum and/or a power spectrum;

The upper computer 110 is configured to monitor a fault condition of the mechanical device according to the second frequency domain data.

It will be appreciated that the vibration signal generated by the mechanical device may be acquired by a vibration sensor (e.g., a piezoelectric vibration acceleration sensor) disposed on the mechanical device housing, which is part of the edge computing acquisition system 120. The vibration signal acquired by the vibration sensor may be in the form of a current and/or voltage signal.

Optionally, the vibration signal comprises a velocity signal and/or an acceleration signal.

Since the acceleration signal is integrated to obtain a velocity signal, and the velocity signal is integrated to obtain a displacement signal, in an alternative scheme, the first frequency domain data and the second frequency domain data may only include related frequency domain data of the acceleration signal, such as a frequency spectrum of the acceleration signal, and after the upper computer 110 obtains the frequency spectrum of the acceleration signal, the frequency spectrum of the velocity signal is obtained by integrating the frequency spectrum of the acceleration signal. This reduces the amount of data that the edge computing acquisition system 120 needs to transmit.

The possible acquisition methods of the spectrum, the envelope spectrum, the order spectrum and the order envelope spectrum will be described in the following. In brief, the spectrum may be calculated using a fourier transform based on the time domain waveform data, the envelope spectrum and the order spectrum may be calculated based on the spectrum, and the order envelope spectrum may be calculated based on the order spectrum. It should be noted that only a part of the time domain waveform data collected by the edge computing and collecting system 120 may be used to calculate the first frequency domain class data, and another part may be uploaded to the upper computer 110 in an alternative.

The edge computing and collecting system 120 can collect time domain waveform data of vibration signals generated by mechanical moving equipment at fixed time, and can collect time domain waveform data of vibration signals generated by mechanical moving equipment in a working condition triggering mode. The vibration signals generated by the mechanical moving equipment are collected by adopting a working condition triggering mode, for example: if some or some faults of the mechanical moving equipment only occur under a certain rotating speed working condition, a rotating speed sensor can be installed on the mechanical moving equipment to acquire the rotating speed of the mechanical moving equipment, certain rotating speed thresholds can be preset, and when the rotating speed of the equipment reaches the thresholds, time domain waveform data are acquired.

It may be appreciated that the first frequency domain data is data of the vibration signal on different frequencies, and the second frequency domain data is data of a partial frequency band in the first frequency domain data, that is, the second frequency domain data may be considered to be formed by cutting a partial frequency band from the first frequency domain data, where the partial frequency band may be a low frequency band, a middle frequency band, a high frequency band, or one or more other frequency bands determined according to requirements. The possible selection method of the partial frequency band is described in the following.

It is understood that the host computer 110 may store data using a database, as shown in fig. 1. The upper computer 110 monitoring the fault condition of the mechanical moving device according to the second frequency domain data may include: the data display and the alarm result display and pushing may be an alarm result of the edge computing and collecting system 120 performing a primary alarm according to an edge alarm policy, or an alarm result of the upper computer 110 performing a secondary alarm according to an upper computer alarm policy (for obtaining an alarm result, see later for details).

Referring to fig. 1, the upper computer 110 is further configured to issue a configured model and/or algorithm to the edge computing and collecting system 120, and the edge computing and collecting system 120 may perform a series of operations such as data collection, index calculation, fault alarm, and data uploading by using the configured model and/or algorithm, and description of the model and algorithm is described in the following description of the algorithm platform 130.

The technical effects that the edge-side vibration monitoring system 100 can achieve include at least the following:

first, the second frequency domain data keeps most of fault analysis key information in the time domain waveform data, and information loss is less, so that operations such as secondary index calculation, secondary alarm strategy application, algorithm, model iteration and the like related to vibration monitoring can be well performed on the upper computer 110 and even the algorithm platform 130 based on the second frequency domain data, and the problem of missing fault information caused by imperfect index design and too low time domain waveform data density (if the time domain waveform data density is too high, the upper computer 110 cannot process) can be solved.

Second, because the data size of the frequency domain data is smaller than the data size of the time domain waveform data, and the technical scheme in the embodiment of the application only transmits the data of a part of frequency bands in the first frequency domain data, the data size sent by the edge computing and collecting system 120 to the upper computer 110 is reduced, and high-density data transmission is facilitated.

Third, if the upper computer 110 needs to perform secondary calculation based on the data sent by the edge calculation acquisition system 120 in the vibration monitoring process, the calculation based on the frequency domain data can avoid the operation with higher calculation amount such as fast fourier transform FFT, so that the calculation pressure of the upper computer 110 is effectively relieved. As can be seen from the above three points, the scheme of the present application selects the second frequency domain data to upload to the upper computer 110, which takes into account the requirement of fault monitoring on the data itself, and effectively reduces the data processing pressure of the monitoring system.

Optionally, the edge computing and acquiring system 120 is configured to determine, when the first frequency domain class data includes a frequency spectrum (where the first frequency domain class data may also include other data), a portion of the frequency spectrum located in the first low frequency band as second frequency domain class data corresponding to the frequency spectrum; wherein the frequency range of the first low frequency range is from W1Hz to W2Hz, W1 and W2 are configurable values (for example, the configuration can be carried out on the upper computer 110), W1 is more than 0, and 5000 is more than or equal to W2 is more than W1; and determining, when the first frequency domain class data includes the envelope spectrum (at this time, the first frequency domain class data may also include other data), a portion of the envelope spectrum that is located in the second low frequency band as second frequency domain class data corresponding to the envelope spectrum; wherein the second low frequency range is from W3Hz to W4Hz, W3 and W4 are configurable values (e.g., can be configured on the host computer 110), W3 > 0, 5000 > W4 > W3.

The inventor researches and discovers that general mechanical faults are often reflected in middle-low frequency areas of a frequency spectrum and/or an envelope spectrum, so that the edge computing and collecting system 120 in the embodiment of the application can only send the part which can reflect the mechanical fault characteristics most in the frequency spectrum and the envelope spectrum to the upper computer 110, on one hand, the communication cost between the edge computing and collecting system 120 and the upper computer 110 is reduced, on the other hand, the calculated amount of vibration monitoring of the upper computer 110 is reduced, and the vibration monitoring efficiency is improved.

Taking an acceleration spectrum as an example, the acceleration spectrum selects a low-frequency band to upload, including the following reasons: one is that the low frequency band is generally a frequency band with clear spectral lines, rather than pile-like noise, and fault frequency characteristics are easy to identify in the frequency band. The low-frequency acceleration is high in association degree of fault types such as gears, bearings and looseness, and the speed spectrum integrated according to the acceleration low-frequency spectrum is basically and fully covered on power frequency faults in the frequency range. In conclusion, the frequency spectrum of the acceleration low-frequency band has high coverage to fault types, and the fault frequency identification is clear, so that the algorithm identification and the manual analysis are easy.

In addition, the spectrum includes other frequency band data besides the first low frequency band, and the edge computing and collecting system 120 only determines the portion of the spectrum in the first low frequency band as the second frequency domain data corresponding to the spectrum. Similarly, the envelope spectrum includes other frequency band data besides the second low frequency band, and the edge computing and collecting system 120 only determines the portion of the envelope spectrum in the second low frequency band as the second frequency domain data corresponding to the envelope spectrum.

It is understood that the frequency ranges of the first low frequency band and the second low frequency band may be configured identically or differently. Also, if there are a plurality of envelope spectrums (how to generate a plurality of envelope spectrums is described later), the second low frequency band corresponding to each of the envelope spectrums may be the same or different.

Optionally, the values of W2 and W4 are each one of the following values: 100. 500, 1000, 2000 and 5000.

This embodiment is, for example: the frequency range of the first low frequency band may be configured to be 0.1-100 hz, 0.1-500 hz, 0.1-1000 hz, 0.1-2000 hz, 0.1-5000 hz, 2-100 hz, 2-500 hz, 2-1000 hz, 2-2000 hz, 2-5000 hz, 10-100 hz, 10-500 hz, 10-1000 hz, 10-2000 hz, or 10-5000 hz, etc.; the frequency range of the second low frequency band may be configured to be 0.1-100 hz, 0.1-500 hz, 0.1-1000 hz, 0.1-2000 hz, 0.1-5000 hz, 2-100 hz, 2-500 hz, 2-1000 hz, 2-2000 hz, 2-5000 hz, 10-100 hz, 10-500 hz, 10-1000 hz, 10-2000 hz, or 10-5000 hz, and so on.

Optionally, if the first frequency domain class data includes an envelope spectrum, the edge computing acquisition system 120 is configured to calculate the envelope spectrum by:

firstly, calculating a frequency spectrum according to time domain waveform data;

Secondly, determining each sub-frequency band of the target demodulation frequency band; wherein, the frequency range of the target demodulation frequency band is configurable (for example, the frequency range can be configured on the upper computer 110), if the number of the sub-frequency bands is one, the sub-frequency band is the target demodulation frequency band, if the number of the sub-frequency bands is multiple, each sub-frequency band is generated by dividing the target demodulation frequency band, and the frequency range of each sub-frequency band is configurable (for example, the frequency range can be configured on the upper computer 110);

finally, the frequency spectrum in each sub-frequency band is respectively subjected to envelope demodulation, and the envelope spectrum (the number of the envelope spectrums is the same as the number of the sub-frequency bands) in each sub-frequency band is obtained.

It can be appreciated that the frequency spectrum can be calculated according to the time domain waveform data by the FFT algorithm, and the target demodulation frequency band can be a low frequency band, a medium frequency band, a high frequency band or other frequency bands determined according to requirements, and in short, the frequency range of the target demodulation frequency band can be freely configured. The number of the sub-frequency bands in the target demodulation frequency band can be one or a plurality of, when the number of the sub-frequency bands is one, the sub-frequency band is the target demodulation frequency band, and when the number of the sub-frequency bands is a plurality of, the sub-frequency band is obtained by dividing the target demodulation frequency band according to configuration.

For example, if the frequency range of the target demodulation frequency band is 1000 to 20000hz, the sub-frequency band of the target demodulation frequency band may include a frequency band of 1000 to 5000hz, a frequency band of 5000 to 10000hz, a frequency band of 10000 to 15000hz, and a frequency band of 15000 to 20000 hz.

It is understood that the frequency ranges of the target demodulation frequency band and each sub-frequency band may be freely configured, for example, if the frequency range of the target demodulation frequency band is 1000 to 20000hz, the sub-frequency band of the target demodulation frequency band may include a frequency band of 1000 to 3000hz, a frequency band of 3000 to 5000hz, a frequency band of 5000 to 5000hz, a frequency band of 8000 to 11000hz, a frequency band of 11000 to 14000hz, a frequency band of 14000 to 17000hz, and a frequency band of 17000 to 20000 hz.

Optionally, if the first frequency domain class data includes a spectrum and an envelope spectrum, and the second frequency domain class data corresponding to the spectrum is a portion of the spectrum located in the first low frequency band, and the frequency range of the first low frequency band is from W1Hz to W2Hz, the frequency range of the target demodulation frequency band is from W2Hz to W5Hz, W5 is a configurable value (for example, may be configured on the upper computer 110), and W5 > W2.

It will be appreciated that the first low frequency band and the target demodulation frequency band are frequency bands joined together, the first low frequency band has a frequency range from W1Hz to W2Hz, and the target demodulation frequency band has a frequency range from W2Hz to W5Hz.

It can be understood that the above-mentioned target demodulation frequency band corresponds to a high frequency band (relative to the first low frequency band) in the frequency spectrum, so that the above-mentioned target demodulation frequency band may be simply referred to as a low frequency band uploading frequency spectrum and a high frequency band uploading envelope spectrum, because the inventor finds that, for most mechanical moving devices, a relatively serious noise exists in a high frequency band portion of the frequency spectrum, if the noise is relatively difficult to identify a fault, the high frequency noise can be filtered after the envelope demodulation is performed, so as to obtain a relatively clear spectral line (especially in a second low frequency band of the envelope spectrum), that is, a relatively clear fault-related feature is reserved, so that a combination manner of the low frequency band uploading frequency spectrum and the high frequency band uploading envelope spectrum is beneficial for the upper computer 110 to monitor the fault condition of the mechanical moving device.

However, it should be noted that the target demodulation frequency band may be set to other frequency bands according to the requirement, for example, a low-rotation-speed device including a gear, and the fault feature thereof is mainly concentrated in the low frequency band, and in this case, if the envelope spectrum is to be uploaded, the target demodulation frequency band may also be set to the low frequency band.

Taking an acceleration spectrum as an example, firstly, the basic spectral line of the acceleration spectrum above 2000hz (possibly other values) has no obvious clear spectral line, even basically takes noise as a main part, so that the accurate identification of fault frequency in the frequency band is generally difficult to identify, the usability of the high-frequency spectrum in the accurate identification is not high, and the uploading of the acceleration spectrum in the high-frequency band is not suitable.

And secondly, the envelope demodulation is used for filtering high-frequency noise, and clear spectral lines related to faults are obtained, so that the envelope demodulation (namely, the target demodulation frequency band is set to be a high frequency band) is basically used on faults such as bearings, retainers and the like. The frequencies associated with the faults are not high, for example, the bearing fault frequency is generally high, even harmonic characteristics (frequency multiplication) are basically below 2000Hz, and an envelope spectrum above 2000Hz obtained by envelope demodulation has basically no significant characteristic frequency. The envelope spectrum is also in the low frequency band where it is valuable, so that an upload of the envelope spectrum in the second low frequency band can be selected.

Further, the frequency band of 2000-20000 hz is wider (target demodulation frequency band), if only one frequency band is used for envelope demodulation, the obtained characteristic information will be lost much, because selecting the frequency band means that the most significant frequency characteristics are found out in all high frequencies obtained from the frequency band, and the characteristics are actually in different envelope segments, the significance of the frequency characteristics is different, the difference of the significance will affect the location, degree and the like of faults, so that the target demodulation frequency band can be divided into a plurality of sub-frequency bands for envelope demodulation respectively.

For example, the envelope spectrum characteristics obtained by selecting a frequency band of 2000-2000 Hz are mainly frequency conversion characteristics, and the characteristic frequency of the bearing is weak.

If it is chosen to divide 2000-20000 hz into three envelope segments, the features obtained will differ.

2000-8000 Hz, the main characteristic of the envelope spectrum is frequency conversion, and the bearing characteristic is invisible;

8000-14000 Hz, the main characteristic of envelope spectrum demodulation is bearing characteristic frequency, and frequency conversion is invisible;

14000-20000 Hz, envelope demodulation mainly comprises frequency conversion, and bearing characteristic frequency is invisible;

in contrast, it is found that, when the Porro demodulation is performed on the frequency band of 2000-2000 Hz, the bearing characteristics can be identified, otherwise, the bearing characteristics may be inhibited in terms of the significance of the characteristics under the condition that the frequency conversion is dominant.

Furthermore, the inventor researches and discovers that, for the case that the mechanical moving equipment is a speed changing equipment (equipment with variable rotating speed, compared with steady-state equipment), the frequency spectrum is in a pile-shaped structure, and the spectrum line is unclear, so that the positioning of center frequency and the like is difficult, the accuracy rate in calculating indexes is low, the fault analysis is also not facilitated, and the order spectrum can be regarded as a special spectrum, the spectrum line is clearer, and the fault analysis and index calculation are facilitated. Thus for variable speed devices, the first frequency domain class data acquired by the edge computing acquisition system 120 may include an order spectrum and/or an order envelope spectrum of the vibration signal (but does not exclude a spectrum and/or an envelope spectrum including the vibration signal). Of course, the inclusion of the order spectrum and/or the order envelope spectrum of the vibration signal in the first frequency domain class data is not excluded for steady state devices either.

Optionally, the edge computing and acquiring system 120 is configured to determine, when the first frequency domain class data includes an order spectrum (where the first frequency domain class data may also include other data), a portion of the order spectrum located in the first low-order segment as second frequency domain class data corresponding to the order spectrum; wherein, the first low-order stage ranges from M1 order to M2 order, M1 and M2 are configurable values (for example, can be configured on the upper computer 110), M1 is more than 0, 5000/reference frequency is more than or equal to M2 and more than M1, and the reference frequency is the frequency corresponding to 1 order in the order spectrum; and determining, when the first frequency domain class data includes the order envelope spectrum (at this time, the first frequency domain class data may also include other data), a portion of the order envelope spectrum located within the second lower-order segment as second frequency domain class data corresponding to the order envelope spectrum; wherein the order of the second lower-order stage ranges from M3 order to M4 order, M3 and M4 are configurable values (e.g., can be configured on the host computer 110), M3 > 0, and 5000/reference frequency is greater than or equal to M4 > M3 (the reference frequency is defined as the same as the order spectrum because the order envelope spectrum is originally calculated based on the order spectrum).

Alternatively, the reference frequency may refer to a frequency of a certain axis of the device or other frequencies, and the order range may be set with reference to the frequency range due to a linear scaling relationship between the order and the frequency (e.g., M1 may be set with reference to W1, but not necessarily, M2 may be set with reference to W2).

It will be appreciated that the order ranges of the first low order segment and the second low order segment may be configured identically or differently. Also, if there are multiple order envelope spectra (how to generate multiple order envelope spectra is described later), the second low-order segment corresponding to each order envelope spectrum may be the same or different.

Optionally, the values of M2 and M4 are each one of the following values: 100/reference frequency, 500/reference frequency, 1000/reference frequency, 2000/reference frequency, and 5000/reference frequency. If the reference frequency is 50Hz, the values of M2 and M4 are one of the following values: 2. 10, 20, 40 and 100.

This embodiment is, for example: the order range of the first low-order stage may be configured to be 0.002 to 2, 0.002 to 10, 0.002 to 20, 0.002 to 40, 0.002 to 100, 0.04 to 2, 0.04 to 10, 0.04 to 20, 0.04 to 40, 0.04 to 100, 0.2 to 2, 0.2 to 10, 0.2 to 20, 0.2 to 40, or 0.2 to 100, to name a few; the order range of the second lower-order stage may be configured to be 0.002 to 2, 0.002 to 10, 0.002 to 20, 0.002 to 40, 0.002 to 100, 0.04 to 2, 0.04 to 10, 0.04 to 20, 0.04 to 40, 0.04 to 100, 0.2 to 2, 0.2 to 10, 0.2 to 20, 0.2 to 40, or 0.2 to 100, to name a few.

Optionally, if the first frequency domain class data includes an order envelope spectrum, the edge computing acquisition system 120 is configured to compute the order envelope spectrum by:

secondly, calculating an order spectrum according to the frequency spectrum and the rotating speed time scale data of the mechanical moving equipment;

further, determining each sub-stage of the target demodulation stage; wherein, the order range of the target demodulation stage is configurable (for example, the configuration can be performed on the upper computer 110), if the number of the sub-stages is one, the sub-stages are the target demodulation stages, and if the number of the sub-stages is multiple, each sub-stage is generated by dividing the target demodulation stage, and the order range of each sub-stage is configurable (for example, the configuration can be performed on the upper computer 110);

and finally, carrying out envelope demodulation on the order spectrums in each sub-order section respectively to obtain order envelope spectrums in each sub-order section.

It is understood that the target demodulation stage may be a low-order stage, a medium-order stage, a high-order stage, or other stages determined according to requirements, and in any case, the order range of the target demodulation stage may be freely configured. The number of sub-stages in the target demodulation stage may be one or more, and when the number of sub-stages is one, the sub-stages are the target demodulation stages, and when the number of sub-stages is more, the sub-stages are obtained by dividing the target demodulation stages according to the configuration.

For example, if the order of the target demodulation order period ranges from (1000/fc) order to (20000/fc) order, the sub-order of the target demodulation order period may include an order period from (1000/fc) order to (5000/fc) order, an order period from (5000/fc) order to (10000/fc) order, an order period from (10000/fc) order to (150000/fc) order, and an order period from (150000/fc) order to (20000/fc) order, where fc is a reference frequency. It will be appreciated that the order ranges of the target demodulation order segment and the respective sub-order segments described above are freely configurable, and are not limited to the values configured in the examples.

Optionally, if the first frequency domain class data includes an order spectrum and an order envelope spectrum, and the second frequency domain class data corresponding to the order spectrum is a portion of the order spectrum located in the first low-order segment, and the order range of the first low-order segment is from M1 to M2, the order range of the target demodulation order segment is from M2 to M5, M5 is a configurable value (e.g., may be configured on the upper computer 110), and M5 > M2.

It will be appreciated that the first low-order segment and the target demodulation-order segment are segments that are joined together, the first low-order segment having an order ranging from M1 to M2, and the target demodulation-order segment having an order ranging from M2 to M5.

It can be understood that the target demodulation stage corresponds to the higher-order stage (relative to the first lower-order stage) in the order spectrum, so that the target demodulation stage may be simply referred to as the lower-order stage uploading order spectrum and the higher-order stage uploading order envelope spectrum, because the inventor finds that the higher-order stage part of the order spectrum has serious noise, if the higher-order stage part is used for identifying a fault, the higher-frequency noise can be filtered after the envelope demodulation is performed, so as to obtain a clearer spectral line (particularly in the second lower-order stage of the order envelope spectrum), that is, most of fault related features are reserved, so that the combination mode of the lower-order stage uploading order spectrum and the higher-order stage uploading order envelope spectrum is beneficial for the upper computer 110 to monitor the fault condition of the mechanical equipment. It should be noted, however, that the target demodulation stage may be set to other stages as desired.

For the problems that the order spectrum is uploaded by selecting a low-order segment, the high-order segment is subjected to envelope demodulation through a molecular order segment to obtain an order envelope spectrum, the order envelope spectrum is uploaded by selecting a low frequency band, and the like, reference can be made to the example of the acceleration spectrum, and the analysis is not repeated.

Optionally, referring to fig. 1, the edge computing acquisition system 120 is further configured to send time domain waveform data to the upper computer 110; the upper computer 110 is configured to monitor a fault condition of the mechanical device according to at least the time domain waveform data and the second frequency domain data, where the monitoring mode may include data display, analysis, alarm, etc.

It will be appreciated that the time domain waveform data may be time domain waveform data collected by the edge computing and collecting system 120 as a result of a condition trigger to the upper computer 110, or collected at regular time, and may be sent when the edge computing and collecting system 120 alarms or does not alarm. It should be noted that, the time domain waveform data collected by the edge computing and collecting system 120 is not necessarily transmitted, but may be only partially transmitted, and the portion is only used for performing edge computing at the edge computing and collecting system 120 and is not transmitted to the upper computer 110.

In the above scheme, the upper computer 110 can monitor the fault condition of the equipment through the time domain waveform data, and since the time domain waveform data is the most original data collected, the most abundant information is contained, which is beneficial to improving the monitoring effect.

Optionally, the period of the edge computing acquisition system 120 sending the second frequency domain class data to the upper computer 110 is less than the period of sending the time domain waveform data to the upper computer 110.

It will be appreciated that the edge computing acquisition system 120 may upload at a lower frequency due to the greater data volume of the time domain waveform data, while the second frequency domain data may upload at a higher frequency due to the lesser data volume, i.e., the period of the second frequency domain class data may be less than the period of transmitting the time domain waveform data to the host computer 110. On the one hand, the data amount is not too large if the time domain waveform data is uploaded at a low frequency, and the second frequency domain class data is still acceptable because the second frequency domain class data itself is not large in data amount although it is uploaded at a high frequency. Further, since the second frequency domain type data can replace the time domain waveform data in many aspects to realize fault monitoring, the period of uploading the time domain waveform data can be further increased, for example, the original period (when the second frequency domain type data is not uploaded) is 2 hours, and the current period (when the second frequency domain type data is uploaded) is 4 hours.

Optionally, the period of sending the second frequency domain class data to the upper computer by the edge computing and collecting system 120 is m minutes, m is a configurable value, and 120 is more than or equal to m is more than or equal to 1. The value of m is one of the following values: 5. 10, 15, 20, 25, 30, 60, 120.

The edge computing device in the mechanical monitoring system in the related art calculates some index data and alarms based on the alarm policy and the index data, but the index data calculated by the edge computing device is only some simple total value indexes or total indexes generated after combination, and the indexes cannot be disassembled, so that the upper computer 110 cannot process the indexes, and the algorithm platform 130 cannot iterate the indexes by using the indexes.

Referring to fig. 1, optionally, in the edge vibration monitoring system 100 provided by the embodiment of the present application, the edge computing and collecting system 120 is further configured to compute first index class data according to the time domain waveform data, and send the first index class data to the upper computer 110; wherein the first index class data comprises a first single class index and/or a first comprehensive class index; and, the device is configured to determine whether to perform fault alarm on the mechanical equipment according to the first index data and the edge alarm policy, and if the mechanical equipment needs to perform fault alarm, send a corresponding alarm result to the upper computer 110.

The upper computer 110 is configured to monitor a fault condition of the mechanical moving device according to the first index class data and the second frequency domain class data when the first index class data is received (no alarm result is sent); or the method is used for monitoring the fault condition of the mechanical equipment according to the first index type data, the alarm result and the second frequency domain type data when the first index type data and the alarm result are received.

The first index class data includes, for example: frequency spectrum indexes such as 1 frequency multiplication, 2 frequency multiplication and the like; envelope spectrum indexes such as envelope spectrum bearing outer ring energy, envelope spectrum retainer energy and the like; the gear meshing frequency conversion sideband energy, frequency conversion harmonic energy and other order spectrum indexes; kurtosis, skewness, phase and other waveform indexes. The first single-class index may be an original index which is not subjected to combination calculation, and the first comprehensive-class index may be an index obtained by combination calculation of a plurality of single-class indexes.

It may be understood that the edge computing and collecting system 120 is provided with an edge alarm policy (i.e. some alarm rules), the calculated first index class data cooperates with the edge alarm policy to perform fault alarm on the mechanical equipment, if the edge computing and collecting system 120 determines that the fault alarm needs to be performed, the first index class data, the second frequency domain class data and the alarm result are sent to the upper computer 110, and at this time, the upper computer 110 monitors the fault condition of the mechanical equipment according to the first index class data, the second frequency domain class data and the alarm result. If the edge computing and collecting system 120 determines that no fault alarm is required, the first index data and the second frequency domain data are sent to the upper computer 110, and at this time, the upper computer 110 monitors the fault condition of the mechanical moving device according to the first index data and the second frequency domain data.

In the above scheme, the edge computing and collecting system 120 can upload the single-class indexes to the upper computer 110, so that on one hand, the computing pressure of the edge computing and collecting system 120 is relieved (because the single-class indexes are not used for combined computation on the edge computing and collecting system 120), on the other hand, the upper computer 110 can perform combined computation, secondary alarm and the like on the received single-class indexes as required, so as to monitor the equipment fault condition more accurately, and flexibly distribute the edge computing pressure between the edge computing and collecting system 120 and the upper computer 110, and avoid the situations that the edge computing and collecting system 120 is too busy, the upper computer 110 is too idle and the like.

For example: aiming at three single indexes of 1 frequency doubling energy, 2 frequency doubling energy and 3 frequency doubling energy, the related technology is to directly sum the 1 frequency doubling energy, the 2 frequency doubling energy and the 3 frequency doubling energy in the edge computing equipment, and then transmit the summed indexes to the upper computer, while the edge computing and collecting system 120 in the embodiment of the application is to transmit all three single indexes of the 1 frequency doubling energy, the 2 frequency doubling energy and the 3 frequency doubling energy to the upper computer 110, and the upper computer 110 can sum the three indexes or sum any two indexes according to the requirement.

The index calculation tasks in the mechanical movable equipment monitoring system in the related art are all borne by the edge calculation equipment, and the upper computer only carries out simple addition and subtraction operation or even no operation on index data, so that the calculation pressure between the upper computer and the edge calculation equipment is unbalanced; meanwhile, in order to balance the complexity of the index and the performance of hardware, the index data is sometimes limited to a certain extent or even deleted directly, so that the accuracy of fault alarm is affected.

Referring to fig. 1, based on this, the upper computer 110 may directly perform index calculation according to the second frequency domain data to obtain second index data; the second index type data comprises a second single type index and/or a second comprehensive type index, and the meaning of the second index type data can refer to the first single type index and the first comprehensive type index respectively. Optionally, the upper computer 110 may further determine whether to perform fault alarm on the mechanical equipment according to the second index data and the upper computer alarm policy.

The second index data in the above scheme is calculated on the upper computer 110, so that the index calculation pressure of the edge calculation acquisition system 120 is relieved, and the edge calculation pressure is flexibly distributed between the edge calculation acquisition system 120 and the upper computer 110. And because the second frequency domain data itself contains most of information in the time domain waveform data, the index calculation can be well supported, and the calculated index is also beneficial to implementing fault alarm. And since the second frequency domain data is the data subjected to fourier transform, the calculation pressure borne by the upper computer 110 is not great when calculating the index.

Optionally, the upper computer 110 may also directly process the first index data, for example, perform a combination calculation on the first single data in the first index data to obtain a comprehensive index, where the comprehensive index is referred to as a third comprehensive index, and determine whether to perform a fault alarm of the mechanical equipment according to the third comprehensive index and an alarm policy of the upper computer.

It can be appreciated that, compared with the scheme that the upper computer directly uses the alarm result of the edge computing device in the related art, the upper computer 110 in the embodiment of the application can set the alarm policy of the upper computer, and judge whether the fault alarm of the mechanical equipment is required according to the second index data and/or the third comprehensive index and the alarm policy of the upper computer, and the method of the secondary alarm is beneficial to improving the accuracy of the fault alarm. And, the upper computer 110 calculates the second index type data and/or the third comprehensive type index according to the requirement, which is beneficial to perfecting the index coverage rate and the complexity.

Alternatively, the edge computing and acquiring system 120 is a wired or wireless acquiring system, i.e. the implementation of the scheme is not limited to the data acquisition mode of the edge computing and acquiring system 120.

If the edge computing and acquiring system 120 is a wired acquiring system, the edge computing and acquiring system 120 includes a wired vibration sensor and a data acquirer, where the wired vibration sensor is used for acquiring time domain waveform data and sending the time domain waveform data to the data acquirer in a wired manner, and the data acquirer is used for performing data interaction with the upper computer 110. Edge calculations during vibration monitoring may be performed independently by the data collector. The edge calculation here is, for example, calculation of index data, alarm judgment, calculation of frequency domain class data, and the like.

If the edge computing and acquiring system 120 is a wireless acquiring system, the edge computing and acquiring system 120 comprises a communication station and a wireless vibration sensor arranged on mechanical equipment; the wireless vibration sensor is used for acquiring time domain waveform data and performing data interaction with the communication station by means of wireless communication (e.g. Zigbee, NB-IoT), the communication station is used for performing data interaction with the host computer 110, and edge calculation in the vibration monitoring process is independently performed by the wireless vibration sensor, or is independently performed by the communication station, or is shared by the communication station and the wireless vibration sensor.

It should be noted that, the reason why the communication station and the wireless vibration sensor share to perform edge calculation or the wireless vibration sensor independently perform edge calculation is that on one hand, more bandwidth is required to be occupied when data is wirelessly transmitted, if the transmission bandwidth of the wireless vibration sensor is limited, the edge calculation can be performed locally more, only the edge calculation result is transmitted, and the transmission of original time domain waveform data with larger data volume is reduced, so that the bandwidth occupation is reduced; on the other hand, the wireless vibration sensor may be powered by a battery, instead of a power supply line, because the power consumption of wireless transmission is large, edge calculation can be performed locally, only the edge calculation result is transmitted, and the transmission of original time domain waveform data with large data volume is reduced, so that the power consumption is reduced. In summary, in the above scheme, the position of the edge calculation is very flexible, and can be allocated according to the need.

Optionally, the upper computer 110 is configured to perform at least one of the following operations to enable monitoring of a fault condition of the mechanical moving device: and performing data display based on the second frequency domain data, performing signal analysis according to the second frequency domain data, and performing signal confirmation according to the second frequency domain data.

Wherein the data presentation operation comprises presenting at least one of: the frequency band information comprises second frequency domain type data, frequency band information corresponding to the second frequency domain type data, envelope segment information corresponding to the second frequency domain type data and association relation between index type data and the second frequency domain type data. The above frequency band information is, for example, a first low frequency band from W1Hz to W2Hz, a second low frequency band from W3Hz to W4Hz, and the like. Envelope segment information such as sub-band information of the target demodulation frequency band and sub-order segment information of the target demodulation order segment.

The signal analysis may be a manual analysis or an analysis performed using a tool.

The signal confirmation is to judge the error signal, that is, to confirm whether the data received by the host computer 110 originates from a correct signal, and if the data originates from the error signal, it is not meaningful to monitor the fault based on such data. The signal confirmation may be performed by the second frequency domain type data or may be performed based on the second index type data calculated from the second frequency domain type data.

In the above scheme, the upper computer 110 may perform a series of monitoring operations, such as data display, signal analysis, signal confirmation, etc., based on the second frequency domain data, which is beneficial to improving the monitoring function of the edge vibration monitoring system.

The algorithm platform in the related art adopts time domain waveform data and/or index data when carrying out algorithm iteration, which restricts the iteration quality of the algorithm and/or model in terms of density and precision, and is specifically expressed in that:

firstly, in terms of data density, due to the fact that uploading and storage pressure are limited, the time domain waveform data is longer in uploading period due to the fact that the data size is larger, the data size is smaller, and therefore the proportion of high-quality data is lower;

secondly, the low-density time domain waveform data does not accord with the data density environment in which edge calculation alarm is actually performed (the edge calculation alarm is performed once in a few minutes, and the time domain waveform data can be uploaded once in a few hours), so that the algorithm and/or the model generated by iteration has poor adaptability to the actual scene;

finally, the data precision of the existing time domain waveform data is mainly based on meeting the requirement of manual analysis, and the data precision is low.

Based on this, referring to fig. 1, optionally, the system further includes an algorithm platform 130; the upper computer 110 is further configured to send second frequency domain class data to the algorithm platform 130; the algorithm platform 130 is configured to iterate a model and/or algorithm used by the edge computing acquisition system 120 according to the second frequency domain class data; the model used by the edge computing and collecting system 120 comprises a device model of mechanical equipment, and the algorithm used by the edge computing and collecting system 120 comprises at least one of an index type data computing algorithm, a data collecting strategy, an edge end alarm strategy and a data uploading strategy.

Of course, if the host computer 110 also sends other data, such as the time domain waveform data, the first index class data, the second index class data, and the third comprehensive class index in fig. 1, to the algorithm platform 130, the algorithm platform 130 may iterate according to the models and/or algorithms used by the edge computing acquisition system 120. It will be appreciated that the above-described device model of the mechanical device is used to describe the mechanical device, which may include the device type (e.g., synchronous motor, asynchronous motor, blower, pump, etc.), device properties (e.g., variable speed motor, steady state motor, etc.), and some parameters of the device (e.g., rotational speed), among others.

If the host computer 110 also performs index calculation and alarm, the algorithm platform 130 may iterate the algorithm and/or model used by the host computer 110 based on the second frequency domain class data and other necessary data.

It may be appreciated that the above-mentioned second frequency domain data may be high-density and high-resolution data, and include a large amount of original information in time domain waveform data, and the algorithm platform 130 may simulate a real density environment (for example, an edge calculation alarm data density environment) by using the high-density and high-resolution second frequency domain data, so as to perform algorithm and/or model iteration of the device model in the upper computer 110 and the edge calculation acquisition system 120, the index class data calculation algorithm in the upper computer 110 and the edge calculation acquisition system 120, the data acquisition policy in the edge calculation acquisition system 120, the edge alarm policy in the edge calculation acquisition system 120, the data uploading policy in the edge calculation acquisition system 120, the upper computer alarm policy in the upper computer 110, and the like according to the simulated real density environment.

In the above scheme, the algorithm platform 130 iterates the model and/or algorithm according to the second spectrum class data, so as to be beneficial to improving the iteration speed and iteration quality of the model and/or algorithm, and promoting standardization of the index class data calculation algorithm and perfection of the data acquisition strategy, the edge end alarm strategy, the data uploading strategy and the upper computer alarm strategy.

Referring to fig. 2, an embodiment of the present application further provides a method for monitoring edge vibration, where the method includes:

step S210: the edge calculation acquisition system acquires time domain waveform data of vibration signals generated by mechanical dynamic equipment;

step S220: the edge computing and collecting system computes first frequency domain data according to the time domain waveform data; wherein the first frequency domain class data comprises at least one of a frequency spectrum, an envelope spectrum, an order spectrum and an order envelope spectrum of the vibration signal, the frequency spectrum comprising an amplitude spectrum and/or a power spectrum;

step S230: the edge computing and collecting system determines second frequency domain data in a partial frequency band in the first frequency domain data and sends the second frequency domain data to an upper computer;

step S240: and the upper computer monitors the fault condition of the mechanical movable equipment according to the second frequency domain data.

The method for monitoring the edge vibration provided by the embodiment of the present application may be implemented by the edge vibration monitoring system provided by the embodiment of the present application, and its implementation principle and the technical effects are described in the foregoing system embodiments, and for the sake of brevity, reference may be made to the corresponding content in any one of the foregoing system embodiments where the method embodiment portion is not mentioned.

The embodiment of the application also provides a computer readable storage medium, and the computer readable storage medium is stored with computer program instructions, and when the computer program instructions are read and run by a processor of a computer, the edge end vibration monitoring method provided by the embodiment of the application is executed.

The embodiment of the application also provides a computer program product, which comprises computer program instructions, and the computer program instructions are read and run by a processor to execute the edge end vibration monitoring method provided by the embodiment of the application.

The above description is only an example of the present application and is not intended to limit the scope of the present application, and various modifications and variations will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. An edge-side vibration monitoring system, comprising: the upper computer and the edge computing and collecting system;

the edge calculation acquisition system is used for acquiring time domain waveform data of vibration signals generated by mechanical dynamic equipment, calculating first frequency domain data according to the time domain waveform data, determining second frequency domain data in a part of frequency bands in the first frequency domain data, and sending the second frequency domain data to the upper computer; wherein the first frequency domain class data comprises at least one of a frequency spectrum, an envelope spectrum, an order spectrum and an order envelope spectrum of the vibration signal, the frequency spectrum comprising an amplitude spectrum and/or a power spectrum;

the upper computer is used for monitoring the fault condition of the mechanical movable equipment according to the second frequency domain data;

if the first frequency domain data includes the spectrum and the envelope spectrum, and the second frequency domain data corresponding to the spectrum is a portion of the spectrum located in a first low frequency band, and the frequency range of the first low frequency band is from W1Hz to W2Hz, W1 and W2 are configurable values, W1 > 0, and 5000 is greater than or equal to W2 > W1, the edge computing and collecting system is configured to calculate the envelope spectrum by:

Calculating the frequency spectrum according to the time domain waveform data;

determining each sub-band of the target demodulation frequency band; wherein, the frequency range of the target demodulation frequency band is from W2Hz to W5Hz, W5 is a configurable value, W5 is more than W2, if the number of the sub-frequency bands is one, the sub-frequency band is the target demodulation frequency band, if the number of the sub-frequency bands is a plurality of sub-frequency bands, each sub-frequency band is generated by dividing the target demodulation frequency band, and the frequency range of each sub-frequency band is configurable;

and respectively carrying out envelope demodulation on the frequency spectrums in each sub-frequency band to obtain envelope spectrums in each sub-frequency band.

2. The edge-side vibration monitoring system of claim 1, wherein the vibration signal comprises a velocity signal and/or an acceleration signal.

3. The system of claim 1, wherein if the mechanical device is a variable speed device, the first frequency domain class data comprises the order spectrum and/or the order envelope spectrum.

4. The system of claim 1, wherein the edge computing and collecting system is configured to determine, when the first frequency domain class data includes the spectrum, a portion of the spectrum that is located in a first low frequency band as second frequency domain class data corresponding to the spectrum; wherein the frequency range of the first low frequency range is from W1Hz to W2Hz, W1 and W2 are configurable values, W1 is more than 0, and W2 is more than or equal to 5000 and more than W1;

And determining a portion of the envelope spectrum located within a second low frequency band as second frequency domain class data corresponding to the envelope spectrum when the first frequency domain class data includes the envelope spectrum; wherein the frequency range of the second low frequency range is from W3Hz to W4Hz, W3 and W4 are configurable values, W3 is more than 0, and 5000 is more than or equal to W4 and more than W3.

5. The system of claim 4, wherein W2 and W4 are each one of the following values: 100. 500, 1000, 2000 and 5000.

6. The rim-end vibration monitoring system of claim 1, wherein the frequency range of the target demodulation frequency band is from 1000Hz to 20000Hz, and the sub-frequency bands of the target demodulation frequency band include frequency bands from 1000Hz to 5000Hz, frequency bands from 5000Hz to 10000Hz, frequency bands from 10000Hz to 15000Hz, and frequency bands from 15000Hz to 20000 Hz.

7. The system of claim 1, wherein the edge computing and collecting system is configured to determine, when the first frequency domain class data includes the order spectrum, a portion of the order spectrum that is located in a first low-order segment as second frequency domain class data corresponding to the order spectrum; the first low-order section has an order range from M1 order to M2 order, M1 and M2 are configurable values, M1 is more than 0, 5000/reference frequency is more than or equal to M2 and more than M1, and the reference frequency is a frequency corresponding to 1 order in the order spectrum;

And determining, when the first frequency domain class data includes the order envelope spectrum, a portion of the order envelope spectrum within a second lower order segment as second frequency domain class data corresponding to the order envelope spectrum; wherein the order of the second low-order section ranges from M3 order to M4 order, M3 and M4 are configurable values, M3 is more than 0, and 5000/reference frequency is more than or equal to M4 and more than M3.

8. The system of claim 7, wherein M2 and M4 are each one of the following values: 100/reference frequency, 500/reference frequency, 1000/reference frequency, 2000/reference frequency, and 5000/reference frequency.

9. The edge vibration monitoring system of claim 1, wherein if the first frequency domain class data includes the order envelope spectrum, the edge computing acquisition system is configured to compute the order envelope spectrum by:

calculating the frequency spectrum according to the time domain waveform data;

calculating the order spectrum according to the frequency spectrum and the rotating speed time scale data of the mechanical moving equipment;

determining each sub-order of the target demodulation order segment; wherein, the order range of the target demodulation order section is configurable, if the number of the sub-order sections is one, the sub-order section is the target demodulation order section, if the number of the sub-order sections is a plurality of sub-order sections, each sub-order section is generated by dividing the target demodulation order section, and the order range of each sub-order section is configurable;

And respectively carrying out envelope demodulation on the order spectrums in each sub-order section to obtain order envelope spectrums in each sub-order section.

10. The edge-side vibration monitoring system according to claim 9, wherein the order of the target demodulation order section ranges from 1000/fc to 20000/fc, and the sub-order section of the target demodulation order section includes an order section from 1000/fc to 5000/fc, an order section from 5000/fc to 10000/fc, an order section from 10000/fc to 15000/fc, and an order section from 15000/fc to 20000/fc, where fc is a reference frequency, and where fc is a frequency corresponding to 1 st order in the order spectrum.

11. The system of claim 9, wherein if the first frequency domain class data includes the order spectrum and the order envelope spectrum, and the second frequency domain class data corresponding to the order spectrum is a portion of the order spectrum located in a first low-order segment, and the order of the first low-order segment ranges from M1 order to M2 order, M1 and M2 are configurable values, M1 > 0, 5000/reference frequency is equal to or greater than M2 > M1, then the order of the target demodulation order segment ranges from M2 order to M5 order, M5 is a configurable value, M5 > M2, and the reference frequency is a frequency corresponding to 1 order in the order spectrum.

12. The system of claim 1, wherein the edge computing and acquisition system is further configured to send the time domain waveform data to the host computer;

the upper computer is used for monitoring the fault condition of the mechanical equipment according to the time domain waveform data and the second frequency domain data; the period of the edge computing and collecting system for sending the second frequency domain data to the upper computer is smaller than the period of the time domain waveform data to the upper computer.

13. The system of claim 1, wherein the period of the edge computing and collecting system sending the second frequency domain data to the upper computer is m minutes, m is a configurable value, and 120 is greater than or equal to m is greater than or equal to 1.

14. The edge vibration monitoring system of claim 13, wherein m is one of: 5. 10, 15, 20, 25, 30, 60, 120.

15. The system of claim 1, wherein the edge computing and collecting system is further configured to compute first index class data according to the time domain waveform data, and send the first index class data to the upper computer; wherein the first index class data comprises a first single class index and/or a first comprehensive class index;

And the device is used for judging whether the mechanical equipment needs to perform fault alarm according to the first index type data and an edge end alarm strategy, and sending a corresponding alarm result to the upper computer if the mechanical equipment needs to perform fault alarm;

the upper computer is used for monitoring the fault condition of the mechanical movable equipment according to the first index class data and the second frequency domain class data when receiving the first index class data; or the method is used for monitoring the fault condition of the mechanical movable equipment according to the first index class data, the alarm result and the second frequency domain class data when the first index class data and the alarm result are received.

16. The system for monitoring the vibration of the edge end according to claim 1, wherein the upper computer is configured to calculate second index class data according to the second frequency domain class data, and determine whether fault alarm of the mechanical equipment is required according to the second index class data and an alarm policy of the upper computer; wherein the second index class data comprises a second single class index and/or a second comprehensive class index.

17. The system of claim 15, wherein the upper computer is configured to calculate a third comprehensive class index according to the first single class index combination, and determine whether a fault alarm of the mechanical equipment is required according to the third comprehensive class index and an alarm policy of the upper computer.

18. The edge-side vibration monitoring system of any one of claims 1-17, wherein the edge computing acquisition system is a wired acquisition system or a wireless acquisition system.

19. The edge vibration monitoring system of claim 18, wherein if the edge computing acquisition system is a wireless acquisition system, the edge computing acquisition system comprises a communication station and a wireless vibration sensor mounted on the mechanical movement device;

the wireless vibration sensor is used for acquiring the time domain waveform data and carrying out data interaction with the communication station in a wireless communication mode, the communication station is used for carrying out data interaction with the upper computer, and edge calculation in the vibration monitoring process is independently carried out by the wireless vibration sensor, or is independently carried out by the communication station, or is shared and carried out by the communication station and the wireless vibration sensor.

20. The edge vibration monitoring system of any one of claims 1-17, wherein the host computer is configured to perform at least one of: performing data display based on the second frequency domain data, performing signal analysis according to the second frequency domain data, and performing signal confirmation according to the second frequency domain data;

Wherein the data presentation operation comprises presenting at least one of: the second frequency domain data comprises frequency band information corresponding to the second frequency domain data, envelope segment information corresponding to the second frequency domain data and association relation between index data and the second frequency domain data.

21. The edge-end vibration monitoring system of any one of claims 1-17, further comprising an algorithm platform;

the upper computer is also used for sending the second frequency domain data to the algorithm platform;

the algorithm platform is used for iterating a model and/or algorithm used by the edge computing and collecting system according to the second frequency domain data; the model used by the edge computing and collecting system comprises an equipment model of the mechanical equipment, and the algorithm used by the edge computing and collecting system comprises at least one of an index data computing algorithm, a data collecting strategy, an edge end alarming strategy and a data uploading strategy.

22. A method for monitoring edge vibration, comprising:

the upper computer monitors the fault condition of the mechanical movable equipment according to the second frequency domain data;

calculating the frequency spectrum according to the time domain waveform data;

23. A computer readable storage medium, having stored thereon computer program instructions which, when read and executed by a processor, perform the method of claim 22.