Vibration abnormity discrimination threshold updating method and electronic equipment

By constructing a coordinate system of impact and harmonic characteristics and a three-stage sample annotation strategy, combined with the threshold update method of automatic review and manual review, the problem of low fault recognition accuracy in vibration monitoring is solved, and higher accuracy is achieved Sex and robustness.

CN118939665APending Publication Date: 2024-11-12HUAWEI TECH CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202310532768.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-05-11
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing vibration monitoring technology has low accuracy when identifying faults of rotating mechanical equipment, which is prone to abnormal missed or false alarms, and requires expert experience to set thresholds, resulting in poor results.

Method used

By constructing a characteristic coordinate system based on impact and harmonicity, using robustness cliffiness and envelope harmonic rate as characteristic quantization of the cyclic impact characteristics of the vibration signal, a three-stage sample annotation strategy and automatic review manual review are used to combine it with The threshold update method improves the accuracy of vibration monitoring.

Benefits of technology

It significantly improves the accuracy of vibration monitoring, reduces the workload and error of manual labeling, avoids false alarms and missed alarms, and enhances the robustness of threshold optimization.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118939665A_ABST
    Figure CN118939665A_ABST
Patent Text Reader

Abstract

The embodiment of the invention provides a vibration abnormity discrimination threshold updating method and electronic equipment. In the technical scheme provided by the embodiment of the invention, the method comprises the following steps: obtaining an impact characteristic and a harmonic characteristic according to a vibration signal of target equipment; according to the impact characteristic, the harmonic characteristic, a first threshold value and a second threshold value, abnormal judgment and alarm judgment are carried out on a sample in the vibration signal; according to the impact characteristics and the harmonic characteristics, labeling and storing samples in the vibration signals based on a three-section type sample labeling strategy to obtain a first labeled sample sequence; performing labeling verification on the first labeling sample sequence through automatic re-checking and manual re-checking to obtain a second labeling sample sequence, updating the first threshold value and the second threshold value according to the second labeling sample sequence, and establishing indexes from impact and harmonic performance to better express the cyclic impact characteristic of the vibration signal, so as to improve the accuracy of the cyclic impact characteristic of the vibration signal. Therefore, the accuracy of vibration monitoring is improved.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The invention relates to the field of fault monitoring, and in particular to a vibration abnormality discrimination threshold updating method and electronic equipment. [Background technology]

[0002] Vibration monitoring is currently the most commonly used sensing method. The transmission loss of mechanical vibration in a rigid body is small, and the response to equipment abnormalities or failures is more sensitive, especially for early failures, which can be identified through weak early features in the vibration signal. In addition, the failure of most rotating parts will cause an instantaneous decrease in the stiffness of the transmission system, causing an impact response of the mechanical system. The difficulty of vibration monitoring lies in effectively processing the vibration signal to achieve robust identification of faults. Due to the differences in the vibration characteristics of equipment in different scenarios, especially for some equipment with non-constant working conditions, the effect of vibration amplitude monitoring and single impact monitoring is not good, and certain expert experience is required for threshold setting for different equipment and different scenarios, resulting in problems in technical implementation, and a large number of abnormal omissions or false alarms may occur on site. Therefore, the accuracy of vibration monitoring is currently low. [Summary of the invention]

[0003] In view of this, an embodiment of the present invention provides a vibration abnormality discrimination threshold updating method and electronic device, which establishes indicators from impact and harmonic properties to better express the cyclic impact characteristics of the vibration signal, thereby improving the accuracy of vibration monitoring.

[0004] In a first aspect, an embodiment of the present invention provides a method for updating a vibration abnormality discrimination threshold, the method comprising:

[0005] According to the vibration signal of the target equipment, the impact characteristics and harmonic characteristics are obtained;

[0006] According to the impact characteristic, the harmonic characteristic, the first threshold and the second threshold, performing abnormality discrimination and alarm discrimination on the samples in the vibration signal;

[0007] According to the impact characteristics and the harmonic characteristics, samples in the vibration signal are labeled and stored based on a three-segment sample labeling strategy to obtain a first labeled sample sequence;

[0008] The first labeled sample sequence is labeled and verified by automatic and manual verification to obtain a second labeled sample sequence, and the first threshold and the second threshold are updated according to the second labeled sample sequence. The embodiment of the present invention uses robust kurtosis and envelope harmonic rate to quantify the characteristics of the impact abnormal components in the vibration signal from the perspective of impact and harmonicity, respectively, and can use a very small number of indicators to achieve a clear distinction between the distribution of normal samples and abnormal samples in the constructed characteristic coordinate system, so that the cyclic impact characteristics of the vibration signal can be better expressed. In the established characteristic coordinate system, the distribution of normal and abnormal states is more separable. The embodiment of the present invention formulates corresponding sample labeling strategies and threshold updating strategies in combination with the constructed feature coordinate system. The feature coordinate system is divided into three sections to agree on normal, abnormal, and pending labeling areas, and artificial capabilities are used to assist in verifying the pending labeling samples, thereby limiting the scope of manual labeling, reducing the workload and blindness of labeling, improving the reliability of sample labeling, and avoiding negative optimization problems caused by sample labeling errors. By optimizing threshold parameters based on partitioned samples of the feature coordinate system, a single linear threshold estimation can be performed for each parameter, avoiding the indicator correlation problem that needs to be considered in the joint estimation and the overfitting problem of the nonlinear method, making the threshold optimization more robust and improving the effect of threshold updating using labeled samples.

[0009] In combination with the first aspect, in certain implementations of the first aspect, before obtaining the impact characteristic and the harmonic characteristic according to the vibration signal of the target device, the method further includes:

[0010] A vibration signal of the target device is obtained.

[0011] In combination with the first aspect, in some implementations of the first aspect, the impact characteristic is robust kurtosis.

[0012] In combination with the first aspect, in some implementations of the first aspect, the harmonic characteristic is an envelope harmonic ratio.

[0013] The embodiment of the present invention constructs a characteristic coordinate system through robust kurtosis and envelope harmonic rate, which can well quantify the cyclic impact abnormal information of rotating mechanical equipment. The linear threshold discrimination scheme established using the two-dimensional feature distribution can avoid the overfitting problem that may be caused by high-dimensional features and nonlinear methods, and is well adapted to specific scenario data. The accuracy of sample abnormality identification can be greatly improved.

[0014] In combination with the first aspect, in certain implementations of the first aspect, obtaining the impact characteristic according to the vibration signal of the target device includes:

[0015] Segmenting the vibration signal according to the segment signal length to obtain a plurality of vibration signal sequences;

[0016] According to the plurality of vibration signal sequences, obtaining a first kurtosis sequence corresponding to the plurality of vibration signal sequences;

[0017] According to the first kurtosis sequence, obtaining a mean and a variance of the first kurtosis sequence;

[0018] Determine whether there is a kurtosis value in the first kurtosis sequence whose difference from the mean is more than three times the variance;

[0019] If it is determined that there is a kurtosis value of the variance whose difference from the mean is more than three times in the first kurtosis sequence, the kurtosis value of the variance whose difference from the mean is more than three times in the first kurtosis sequence is removed to obtain a second kurtosis sequence;

[0020] According to the second kurtosis sequence, obtaining a mean value of the second kurtosis sequence;

[0021] The mean value of the second kurtosis sequence is used as the impact feature.

[0022] In combination with the first aspect, in some implementations of the first aspect, after determining whether there is a kurtosis value in the first kurtosis sequence whose difference from the mean value exceeds three times of the variance, the method further includes:

[0023] If it is determined that there is no kurtosis value in the first kurtosis sequence whose difference from the mean exceeds three times of the variance, the mean of the first kurtosis sequence is used as the impact feature.

[0024] In combination with the first aspect, in certain implementations of the first aspect, obtaining a harmonic characteristic according to a vibration signal of a target device includes:

[0025] Obtaining an envelope signal of the vibration signal according to Hilbert transform and the vibration signal;

[0026] The harmonic characteristic is obtained based on the autocorrelation sequence of the envelope signal.

[0027] In combination with the first aspect, in some implementations of the first aspect, the performing abnormality discrimination and alarm discrimination on samples in the vibration signal of the target device according to the impact characteristic, the harmonic characteristic, the first threshold and the second threshold includes:

[0028] Determining whether the impulse characteristic is greater than the first threshold and whether the harmonic characteristic is greater than the second threshold;

[0029] If it is determined that the impact characteristic is greater than the first threshold value and the harmonic characteristic is greater than the second threshold value, the abnormality determination result is abnormal;

[0030] Starting from the current sample, trace back the first numerical number of samples in the running state to obtain multiple samples;

[0031] Determining whether the ratio of abnormal samples among the plurality of samples is greater than or equal to a second value;

[0032] If it is determined that the proportion of abnormal samples among the plurality of samples is greater than or equal to a second value, an alarm is issued;

[0033] If it is determined that the proportion of abnormal samples among the multiple samples is less than the second value, no alarm is issued.

[0034] In combination with the first aspect, in some implementations of the first aspect, after determining whether the impulse characteristic is greater than the first threshold and whether the harmonic characteristic is greater than the second threshold, the method further includes:

[0035] If it is determined that the impact characteristic is less than or equal to the first threshold, or the harmonic characteristic is less than or equal to the second threshold, the abnormality determination result is normal.

[0036] In combination with the first aspect, in some implementations of the first aspect, the samples in the vibration signal are labeled and stored based on a three-segment sample labeling strategy according to the impact feature and the harmonic feature to obtain a first labeled sample sequence, including:

[0037] Establish a characteristic coordinate system based on the impact dimension and the harmonic dimension;

[0038] Setting a first interval division point in the harmonic dimension and setting a second interval division point in the impact dimension;

[0039] According to the characteristic coordinate system, the first interval division point, the second interval division point, the impact characteristic, and the harmonic characteristic, samples in the vibration signal are labeled and stored based on the three-segment sample labeling strategy to obtain a first labeled sample sequence.

[0040] In combination with the first aspect, in some implementations of the first aspect, the first interval division points include a first division point and a second division point, and the second interval division points include a third division point and a fourth division point;

[0041] The absolute value of the value of the second dividing point is greater than the absolute value of the value of the first dividing point;

[0042] The absolute value of the value of the fourth division point is greater than the absolute value of the value of the third division point.

[0043] In conjunction with the first aspect, in some implementations of the first aspect, the three-segment sample labeling strategy includes:

[0044] When the impact characteristic and the harmonic characteristic meet the first condition, marking the sample as normal;

[0045] When the impact characteristic and the harmonic characteristic meet the second condition, marking the sample as abnormal;

[0046] When the impact characteristic and the harmonic characteristic meet the third condition, the sample is marked as pending.

[0047] The embodiment of the present invention performs three-segment division and labeling based on the feature coordinate system, and classifies the sample labels into normal, abnormal or pending. It can extract weak feature samples near the time point of degradation mutation of rotating mechanical equipment, and based on the envelope spectrum analysis multiple frequency comparison method, automatically or manually verify the pending sample labels in a layered manner, thereby introducing auxiliary verification. Due to the differences in data distribution in different scenarios or equipment, this method reduces the workload and scope of manual labeling, and can better clarify the boundary samples in the normal state, thereby avoiding the problem of negative optimization of the model caused by incorrect labeling.

[0048] In combination with the first aspect, in some implementations of the first aspect, the first condition includes: the impact characteristic is less than or equal to the value of the third division point, or the harmonic characteristic is less than or equal to the value of the first division point.

[0049] In combination with the first aspect, in some implementations of the first aspect, the second condition includes: the impact characteristic is greater than or equal to the value of the fourth division point, and the harmonic characteristic is greater than or equal to the value of the second division point.

[0050] In combination with the first aspect, in some implementations of the first aspect, the third condition is different from the first condition and the second condition.

[0051] In conjunction with the first aspect, in some implementations of the first aspect, the step of performing label verification on the first labeled sample sequence by automatic review includes:

[0052] Performing impact feature enhancement on the first labeled sample sequence by a minimum entropy deconvolution method to obtain an enhanced signal sequence;

[0053] According to the enhanced signal sequence, obtaining a characteristic index of the enhanced signal sequence;

[0054] Determining whether the characteristic index of the enhanced signal sequence belongs to a normal area;

[0055] If it is determined that the characteristic index belongs to the normal area, the labels of the labeled samples marked as pending in the first labeled sample sequence are corrected to normal, so as to obtain the second labeled sample sequence.

[0056] In combination with the first aspect, in certain implementations of the first aspect, the characteristic indicators of the enhanced signal sequence include the robustness kurtosis and envelope harmonic rate of the enhanced signal sequence.

[0057] In combination with the first aspect, in certain implementations of the first aspect, the normal area includes: the robustness kurtosis of the enhanced signal is less than or equal to the value of the fourth division point, or the envelope harmonic rate of the enhanced signal is less than or equal to the value of the second division point.

[0058] In conjunction with the first aspect, in some implementations of the first aspect, the step of manually verifying the first labeled sample sequence includes:

[0059] If it is determined that the characteristic indicator does not belong to a normal region, obtaining a first envelope spectrum according to the enhanced signal sequence;

[0060] Obtaining a second envelope spectrum according to the first labeled sample sequence;

[0061] Determining whether the frequency-doubled characteristics of the first envelope spectrum and the second envelope spectrum in the low-frequency domain are consistent;

[0062] If it is determined that the first envelope spectrum and the second envelope spectrum have the same frequency doubling characteristics in the low frequency domain, in response to the user's operation, the labeling of the labeled sample marked as pending in the first labeled sample sequence is corrected to abnormal, so as to obtain the second labeled sample sequence;

[0063] If it is determined that the frequency doubling characteristics of the first envelope spectrum and the second envelope spectrum in the low frequency domain are inconsistent, in response to the user's operation, the labeling of the labeled samples marked as pending in the second labeled sample sequence is corrected to normal, so as to obtain the second labeled sample sequence.

[0064] With reference to the first aspect, in some implementations of the first aspect, updating the first threshold and the second threshold according to the second labeled sample sequence includes:

[0065] Acquire a plurality of annotated samples and their characteristic indicators within a preset interval from the second annotated sample sequence;

[0066] According to the characteristic indicators of the plurality of labeled samples and the preset interval, a fourth condition and a fifth condition are obtained;

[0067] According to the fourth condition, obtaining a plurality of first normal labeled samples from the plurality of labeled samples;

[0068] Obtaining a harmonic rate sequence by extracting the harmonic rate values ​​of the plurality of first normal labeled samples;

[0069] updating the second threshold value according to the harmonic rate sequence;

[0070] According to the fifth condition, a plurality of second normal labeled samples are obtained from the plurality of labeled samples;

[0071] A third kurtosis sequence is obtained by extracting the robust kurtosis values ​​of the plurality of second normal labeled samples;

[0072] The first threshold is updated according to the third kurtosis sequence.

[0073] The embodiment of the present invention optimizes the threshold parameters by partitioning samples based on the feature coordinate system, and can perform a single linear threshold estimation for each parameter respectively, avoiding the indicator correlation problem that needs to be considered in the joint estimation and the overfitting problem of the nonlinear method, making the threshold optimization more robust and improving the effect of using labeled samples for threshold updating.

[0074] The embodiment of the present invention uses a threshold updating method using normal state samples based on feature coordinate system partitioning. Since equipment failure is a low-probability event, using only normal samples for threshold updating avoids the requirement for failure data, is more in line with actual needs, and has better recognition capabilities for unknown samples that deviate from the overall distribution of state data.

[0075] In combination with the first aspect, in certain implementations of the first aspect, the interval division point of the preset interval in the harmonic dimension is a first interval division point.

[0076] In combination with the first aspect, in some implementations of the first aspect, the interval division point of the preset interval in the impact dimension is a second interval division point.

[0077] In combination with the first aspect, in certain implementations of the first aspect, the target device includes a reducer, a motor, a gearbox, a fan, a pump, or a turbine rotor.

[0078] In combination with the first aspect, in some implementations of the first aspect, obtaining the vibration signal of the target device includes:

[0079] A vibration signal of the target device is obtained through a vibration sensor.

[0080] In combination with the first aspect, in certain implementations of the first aspect, the installation position of the vibration sensor includes a device driving end, a device non-driving end, a device top, a device bottom, or other parts connected to the device body.

[0081] In a second aspect, an embodiment of the present invention provides an electronic device, comprising a processor and a memory, wherein the memory is used to store a computer program, and the computer program comprises program instructions, and when the processor runs the program instructions, the electronic device executes the steps of the method described above.

[0082] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program request is executed by a computer, the computer is caused to execute the method as described above.

[0083] In a fourth aspect, an embodiment of the present invention provides a computer program product, which includes instructions, and when the computer program product is run on a computer or any at least one processor, the computer executes the functions / steps in the above method.

[0084] In the technical solution of the vibration abnormality discrimination threshold updating method and electronic device provided by the embodiment of the present invention, the method includes: obtaining the impact characteristic and the harmonic characteristic according to the vibration signal of the target device; performing abnormality discrimination and alarm discrimination on the samples in the vibration signal according to the impact characteristic, the harmonic characteristic, the first threshold and the second threshold; labeling and storing the samples in the vibration signal based on the three-segment sample labeling strategy according to the impact characteristic and the harmonic characteristic to obtain a first labeled sample sequence; labeling and verifying the first labeled sample sequence through automatic review and manual review to obtain a second labeled sample sequence, and updating the first threshold and the second threshold according to the second labeled sample sequence, establishing indicators from the impact characteristic and the harmonic characteristic to better express the cyclic impact characteristics of the vibration signal, thereby improving the accuracy of vibration monitoring.

[0085] The embodiment of the present invention provides a vibration anomaly discrimination threshold updating method for rotating mechanical equipment, which can be used to optimize the parameters of the sample anomaly discriminator in combination with field data. Specifically, an indicator is constructed from the two characteristic dimensions of impact and harmonicity to characterize the cyclic impact information of the vibration signal, forming a characteristic coordinate system; sample anomalies are identified by constructing a threshold discriminator, and equipment anomalies are discriminated based on the continuity criterion; in addition, in order to meet the generalization ability of the model, the three-segment annotation area is divided based on the characteristic coordinate system for annotation verification, avoiding the model optimization deviation caused by the annotation error, and effectively reducing the number of samples to be labeled. Finally, the threshold discriminator parameters are optimized through sample partitioning to achieve model tuning.

Brief Description of the Drawings

[0086] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0087] Figure 1 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention;

[0088] Figure 2 is a software structure block diagram of the electronic device 100 according to an embodiment of the present invention;

[0089] Figure 3 An architectural diagram of a vibration abnormality discrimination threshold updating system provided by an embodiment of the present invention;

[0090] Figure 4 for Figure 3 Schematic diagram of the working process of the vibration abnormality discrimination threshold updating system;

[0091] Figure 5 is a schematic diagram of an updating process of a first threshold value and a second threshold value;

[0092] Figure 6 is a schematic diagram of a segmented signal and its kurtosis curve of a normal state sample;

[0093] Figure 7 Schematic diagram of the envelope signal and its autocorrelation sequence of a normal sample;

[0094] Figure 8 It is a schematic diagram of three-segment annotation based on the impact characteristic coordinate system;

[0095] Fig. 9 A schematic diagram of the sample with the verification results;

[0096] Fig.10 Schematic diagram of the recognition results for the threshold-optimized dataset;

[0097] Fig.11 is a schematic diagram of sample anomaly recognition results based on the updated threshold;

[0098] Fig.12 A flowchart of a vibration abnormality discrimination threshold updating method provided by an embodiment of the present invention;

[0099] Fig.13 A flowchart of another method for updating a vibration abnormality discrimination threshold value provided by an embodiment of the present invention;

[0100] Fig.14 for Fig.13A specific flow chart of obtaining the impact characteristics according to the vibration signal of the target device;

[0101] Fig.15 for Fig.13 A specific flow chart of obtaining harmonic characteristics according to the vibration signal of the target equipment;

[0102] Fig.16 for Fig.13 A specific flow chart of performing abnormality discrimination and alarm discrimination on samples in a vibration signal according to the impact characteristics, harmonic characteristics, first threshold and second threshold;

[0103] Fig.17 for Fig.13 In the step, according to the impact characteristics and harmonic characteristics, samples in the vibration signal are labeled and stored based on a three-segment sample labeling strategy to obtain a specific flow chart of the first labeled sample sequence;

[0104] Fig.18 for Fig.13 A specific flow chart of performing annotation verification on a first annotated sample sequence by automatic review and manual review to obtain a second annotated sample sequence, and updating a first threshold and a second threshold according to the second annotated sample sequence;

[0105] Fig.19 for Fig.13 A specific flow chart of updating the first threshold and the second threshold according to the second labeled sample sequence;

[0106] Fig. 20 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. [Specific implementation method]

[0107] In order to better understand the technical solution of the present invention, the embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0108] It should be clear that the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0109] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings.

[0110] It should be understood that the term "and / or" used in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0111] Figure 1 A schematic structural diagram of an electronic device 100 is shown.

[0112] The electronic device 100 may include a processor 110, an external memory interface 120, an internal memory 121, a universal serial bus (USB) interface 130, a charging management module 140, a power management module 141, a battery 142, an antenna 1, an antenna 2, a mobile communication module 150, a wireless communication module 160, an audio module 170, a speaker 170A, a receiver 170B, a microphone 170C, an earphone interface 170D, a sensor module 180, a button 190, a motor 191, an indicator 192, a camera 193, a display screen 194, and a subscriber identification module (SIM) card interface 195, etc. The sensor module 180 may include a pressure sensor 180A, a gyroscope sensor 180B, an air pressure sensor 180C, a magnetic sensor 180D, an acceleration sensor 180E, a distance sensor 180F, a proximity light sensor 180G, a fingerprint sensor 180H, a temperature sensor 180J, a touch sensor 180K, an ambient light sensor 180L, a bone conduction sensor 180M, etc.

[0113] It is to be understood that the structure illustrated in the embodiment of the present invention does not constitute a specific limitation on the electronic device 100. In other embodiments of the present application, the electronic device 100 may include more or fewer components than shown in the figure, or combine some components, or separate some components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.

[0114] The processor 110 may include one or more processing units, for example, the processor 110 may include an application processor (AP), a modem processor, a graphics processor (GPU), an image signal processor (ISP), a controller, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Different processing units may be independent devices or integrated into one or more processors.

[0115] The controller can generate operation control signals according to the instruction operation code and timing signal to complete the control of instruction fetching and execution.

[0116] The processor 110 may also be provided with a memory for storing instructions and data. In some embodiments, the memory in the processor 110 is a cache memory. The memory may store instructions or data that the processor 110 has just used or cyclically used. If the processor 110 needs to use the instruction or data again, it may be directly called from the memory. This avoids repeated access, reduces the waiting time of the processor 110, and thus improves the efficiency of the system.

[0117] In some embodiments, the processor 110 may include one or more interfaces. The interface may include an inter-integrated circuit (I2C) interface, an inter-integrated circuit sound (I2S) interface, a pulse code modulation (PCM) interface, a universal asynchronous receiver / transmitter (UART) interface, a mobile industry processor interface (MIPI), a general-purpose input / output (GPIO) interface, a subscriber identity module (SIM) interface, and / or a universal serial bus (USB) interface, etc.

[0118] The USB interface 130 is an interface that complies with the USB standard specification, and specifically can be a Mini USB interface, a Micro USB interface, a USB Type C interface, etc. The USB interface 130 can be used to connect a charger to charge the electronic device 100, and can also be used to transfer data between the electronic device 100 and a peripheral device. It can also be used to connect headphones to play audio through the headphones. The interface can also be used to connect other electronic devices, such as AR devices, etc.

[0119] It is understandable that the interface connection relationship between the modules illustrated in the embodiment of the present invention is only a schematic illustration and does not constitute a structural limitation on the electronic device 100. In other embodiments of the present application, the electronic device 100 may also adopt different interface connection methods in the above embodiments, or a combination of multiple interface connection methods.

[0120] The charging management module 140 is used to receive charging input from a charger. The charger may be a wireless charger or a wired charger. In some wired charging embodiments, the charging management module 140 may receive charging input from a wired charger through the USB interface 130. In some wireless charging embodiments, the charging management module 140 may receive wireless charging input through a wireless charging coil of the electronic device 100. While the charging management module 140 is charging the battery 142, it may also power the electronic device through the power management module 141.

[0121] The power management module 141 is used to connect the battery 142, the charging management module 140 and the processor 110. The power management module 141 receives input from the battery 142 and / or the charging management module 140, and supplies power to the processor 110, the internal memory 121, the display screen 194, the camera 193, and the wireless communication module 160. The power management module 141 can also be used to monitor parameters such as battery capacity, battery cycle number, battery health status (leakage, impedance), etc. In some other embodiments, the power management module 141 can also be set in the processor 110. In other embodiments, the power management module 141 and the charging management module 140 can also be set in the same device.

[0122] The wireless communication function of the electronic device 100 can be implemented through the antenna 1, the antenna 2, the mobile communication module 150, the wireless communication module 160, the modem processor and the baseband processor.

[0123] Antenna 1 and antenna 2 are used to transmit and receive electromagnetic wave signals. Each antenna in electronic device 100 can be used to cover a single or multiple communication frequency bands. Different antennas can also be reused to improve the utilization of antennas. For example, antenna 1 can be reused as a diversity antenna for a wireless local area network. In some other embodiments, the antenna can be used in combination with a tuning switch.

[0124] The mobile communication module 150 can provide solutions for wireless communications including 2G / 3G / 4G / 5G, etc., applied to the electronic device 100. The mobile communication module 150 may include at least one filter, a switch, a power amplifier, a low noise amplifier (LNA), etc. The mobile communication module 150 can receive electromagnetic waves from the antenna 1, and filter, amplify, and process the received electromagnetic waves, and transmit them to the modulation and demodulation processor for demodulation. The mobile communication module 150 can also amplify the signal modulated by the modulation and demodulation processor, and convert it into electromagnetic waves for radiation through the antenna 1. In some embodiments, at least some of the functional modules of the mobile communication module 150 can be set in the processor 110. In some embodiments, at least some of the functional modules of the mobile communication module 150 can be set in the same device as at least some of the modules of the processor 110.

[0125] The modem processor may include a modulator and a demodulator. Among them, the modulator is used to modulate the low-frequency baseband signal to be sent into a medium-high frequency signal. The demodulator is used to demodulate the received electromagnetic wave signal into a low-frequency baseband signal. The demodulator then transmits the demodulated low-frequency baseband signal to the baseband processor for processing. After the low-frequency baseband signal is processed by the baseband processor, it is passed to the application processor. The application processor outputs a sound signal through an audio device (not limited to a speaker 170A, a receiver 170B, etc.), or displays an image or video through a display screen 194. In some embodiments, the modem processor may be an independent device. In other embodiments, the modem processor may be independent of the processor 110 and be set in the same device as the mobile communication module 150 or other functional modules.

[0126] The wireless communication module 160 can provide wireless communication solutions including wireless local area networks (WLAN) (such as wireless fidelity (Wi-Fi) networks), bluetooth (BT), global navigation satellite system (GNSS), frequency modulation (FM), near field communication (NFC), infrared (IR), etc., which are applied to the electronic device 100. The wireless communication module 160 can be one or more devices integrating at least one communication processing module. The wireless communication module 160 receives electromagnetic waves via the antenna 2, modulates the frequency of the electromagnetic wave signal and performs filtering, and sends the processed signal to the processor 110. The wireless communication module 160 can also receive the signal to be sent from the processor 110, modulate the frequency of it, amplify it, and convert it into electromagnetic waves for radiation through the antenna 2.

[0127] In some embodiments, the antenna 1 of the electronic device 100 is coupled to the mobile communication module 150, and the antenna 2 is coupled to the wireless communication module 160, so that the electronic device 100 can communicate with the network and other devices through wireless communication technology. The wireless communication technology may include global system for mobile communications (GSM), general packet radio service (GPRS), code division multiple access (CDMA), wideband code division multiple access (WCDMA), time-division code division multiple access (TD-SCDMA), long term evolution (LTE), BT, GNSS, WLAN, NFC, FM, and / or IR technology. The GNSS may include a global positioning system (GPS), a global navigation satellite system (GLONASS), a Beidou navigation satellite system (BDS), a quasi-zenith satellite system (QZSS) and / or a satellite based augmentation system (SBAS).

[0128] The electronic device 100 implements the display function through a GPU, a display screen 194, and an application processor. The GPU is a microprocessor for image processing, which connects the display screen 194 and the application processor. The GPU is used to perform mathematical and geometric calculations for graphics rendering. The processor 110 may include one or more GPUs that execute program instructions to generate or change display information.

[0129] The display screen 194 is used to display images, videos, etc. The display screen 194 includes a display panel. The display panel can be a liquid crystal display (LCD), an organic light-emitting diode (OLED), an active-matrix organic light-emitting diode or an active-matrix organic light-emitting diode (AMOLED), a flexible light-emitting diode (FLED), Miniled, MicroLed, Micro-oLed, a quantum dot light-emitting diode (QLED), etc. In some embodiments, the electronic device 100 may include 1 or N display screens 194, where N is a positive integer greater than 1.

[0130] The electronic device 100 can realize the shooting function through ISP, camera 193, video codec, GPU, display screen 194 and application processor.

[0131] ISP is used to process the data fed back by camera 193. For example, when taking a photo, the shutter is opened, and the light is transmitted to the camera photosensitive element through the lens. The light signal is converted into an electrical signal, and the camera photosensitive element transmits the electrical signal to ISP for processing and converts it into an image visible to the naked eye. ISP can also perform algorithm optimization on the noise, brightness, and skin color of the image. ISP can also optimize the exposure, color temperature and other parameters of the shooting scene. In some embodiments, ISP can be set in camera 193.

[0132] The camera 193 is used to capture still images or videos. The object generates an optical image through the lens and projects it onto the photosensitive element. The photosensitive element can be a charge coupled device (CCD) or a complementary metal-oxide-semiconductor (CMOS) phototransistor. The photosensitive element converts the optical signal into an electrical signal, and then passes the electrical signal to the ISP to be converted into a digital image signal. The ISP outputs the digital image signal to the DSP for processing. The DSP converts the digital image signal into an image signal in a standard RGB, YUV or other format. In some embodiments, the electronic device 100 may include 1 or N cameras 193, where N is a positive integer greater than 1.

[0133] The digital signal processor is used to process digital signals, and can process not only digital image signals but also other digital signals. For example, when the electronic device 100 is selecting a frequency point, the digital signal processor is used to perform Fourier transform on the frequency point energy.

[0134] Video codecs are used to compress or decompress digital videos. The electronic device 100 may support one or more video codecs. Thus, the electronic device 100 may play or record videos in a variety of coding formats, such as Moving Picture Experts Group (MPEG) 1, MPEG2, MPEG3, MPEG4, etc.

[0135] NPU is a neural network (NN) computing processor. By drawing on the structure of biological neural networks, such as the transmission mode between neurons in the human brain, it can quickly process input information and can also continuously self-learn. Through NPU, applications such as intelligent cognition of electronic device 100 can be realized, such as image recognition, face recognition, voice recognition, text understanding, etc.

[0136] The external memory interface 120 can be used to connect an external memory card, such as a Micro SD card, to expand the storage capacity of the electronic device 100. The external memory card communicates with the processor 110 through the external memory interface 120 to implement a data storage function, such as storing music, video and other files in the external memory card.

[0137] The internal memory 121 can be used to store computer executable program codes, which include instructions. The internal memory 121 may include a program storage area and a data storage area. Among them, the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc. The data storage area may store data created during the use of the electronic device 100 (such as audio data, a phone book, etc.), etc. In addition, the internal memory 121 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, a universal flash storage (UFS), etc. The processor 110 executes various functional applications and data processing of the electronic device 100 by running instructions stored in the internal memory 121, and / or instructions stored in a memory provided in the processor.

[0138] The electronic device 100 can implement audio functions such as music playing and recording through the audio module 170, the speaker 170A, the receiver 170B, the microphone 170C, the headphone jack 170D, and the application processor.

[0139] The audio module 170 is used to convert digital audio information into analog audio signal output, and is also used to convert analog audio input into digital audio signals. The audio module 170 can also be used to encode and decode audio signals. In some embodiments, the audio module 170 can be arranged in the processor 110, or some functional modules of the audio module 170 can be arranged in the processor 110.

[0140] The speaker 170A, also called a "speaker", is used to convert an audio electrical signal into a sound signal. The electronic device 100 can listen to music or listen to a hands-free call through the speaker 170A.

[0141] The receiver 170B, also called a "earpiece", is used to convert audio electrical signals into sound signals. When the electronic device 100 receives a call or voice message, the voice can be received by placing the receiver 170B close to the human ear.

[0142] Microphone 170C, also called "microphone" or "microphone", is used to convert sound signals into electrical signals. When making a call or sending a voice message, the user can speak by putting their mouth close to microphone 170C to input the sound signal into microphone 170C. The electronic device 100 can be provided with at least one microphone 170C. In other embodiments, the electronic device 100 can be provided with two microphones 170C, which can not only collect sound signals but also realize noise reduction function. In other embodiments, the electronic device 100 can also be provided with three, four or more microphones 170C to collect sound signals, reduce noise, identify the sound source, realize directional recording function, etc.

[0143] The earphone interface 170D is used to connect a wired earphone and can be a USB interface 130 or a 3.5 mm open mobile terminal platform (OMTP) standard interface or a cellular telecommunications industry association of the USA (CTIA) standard interface.

[0144] The key 190 includes a power key, a volume key, etc. The key 190 may be a mechanical key or a touch key. The electronic device 100 may receive key input and generate key signal input related to user settings and function control of the electronic device 100.

[0145] Motor 191 can generate vibration prompts. Motor 191 can be used for incoming call vibration prompts, and can also be used for touch vibration feedback. For example, touch operations acting on different applications (such as taking pictures, audio playback, etc.) can correspond to different vibration feedback effects. For touch operations acting on different areas of the display screen 194, motor 191 can also correspond to different vibration feedback effects. Different application scenarios (for example: time reminders, receiving messages, alarm clocks, games, etc.) can also correspond to different vibration feedback effects. The touch vibration feedback effect can also support customization.

[0146] Indicator 192 may be an indicator light, which may be used to indicate charging status, power changes, messages, missed calls, notifications, etc.

[0147] The SIM card interface 195 is used to connect a SIM card. The SIM card can be connected to and separated from the electronic device 100 by inserting it into the SIM card interface 195 or pulling it out from the SIM card interface 195. The electronic device 100 can support 1 or N SIM card interfaces, where N is a positive integer greater than 1. The SIM card interface 195 can support Nano SIM cards, Micro SIM cards, SIM cards, and the like. Multiple cards can be inserted into the same SIM card interface 195 at the same time. The types of the multiple cards can be the same or different. The SIM card interface 195 can also be compatible with different types of SIM cards. The SIM card interface 195 can also be compatible with external memory cards. The electronic device 100 interacts with the network through the SIM card to implement functions such as calls and data communications. In some embodiments, the electronic device 100 uses an eSIM, i.e., an embedded SIM card. The eSIM card can be embedded in the electronic device 100 and cannot be separated from the electronic device 100.

[0148] The software system of the electronic device 100 may adopt a layered architecture, an event-driven architecture, a micro-core architecture, a micro-service architecture, or a cloud architecture. In the embodiment of the present invention, the Android system of the layered architecture is taken as an example to exemplify the software structure of the electronic device 100.

[0149] Figure 2 It is a software structure block diagram of the electronic device 100 according to an embodiment of the present invention.

[0150] The layered architecture divides the software into several layers, each with clear roles and division of labor. The layers communicate with each other through software interfaces. In some embodiments, the Android system is divided into four layers, from top to bottom: the application layer, the application framework layer, the Android runtime and system library, and the kernel layer.

[0151] The application layer can include a series of application packages.

[0152] like Figure 2 As shown, the application package may include camera, gallery, calendar, call, map, navigation, WLAN, Bluetooth, music, video, short message and other applications.

[0153] The application framework layer provides an application programming interface (API) and a programming framework for the applications in the application layer. The application framework layer includes some predefined functions.

[0154] like Figure 2 As shown, the application framework layer may include a window manager, a content provider, a view system, a phone manager, a resource manager, a notification manager, and the like.

[0155] The window manager is used to manage window programs. The window manager can obtain the display screen size, determine whether there is a status bar, lock the screen, capture the screen, etc.

[0156] Content providers are used to store and retrieve data and make it accessible to applications. The data may include videos, images, audio, calls made and received, browsing history and bookmarks, phone books, etc.

[0157] The view system includes visual controls, such as controls for displaying text, controls for displaying images, etc. The view system can be used to build applications. A display interface can be composed of one or more views. For example, a display interface including a text notification icon can include a view for displaying text and a view for displaying images.

[0158] The phone manager is used to provide communication functions of the electronic device 100, such as management of call status (including connecting, hanging up, etc.).

[0159] The resource manager provides various resources for applications, such as localized strings, icons, images, layout files, video files, and so on.

[0160] The notification manager enables applications to display notification information in the status bar. It can be used to convey notification-type messages and can disappear automatically after a short stay without user interaction. For example, the notification manager is used to notify download completion, message reminders, etc. The notification manager can also be a notification that appears in the system top status bar in the form of a chart or scroll bar text, such as notifications of applications running in the background, or a notification that appears on the screen in the form of a dialog window. For example, a text message is displayed in the status bar, a prompt sound is emitted, an electronic device vibrates, an indicator light flashes, etc.

[0161] Android Runtime includes core libraries and virtual machines. Android runtime is responsible for scheduling and management of the Android system.

[0162] The core library consists of two parts: one part is the function that needs to be called by the Java language, and the other part is the Android core library.

[0163] The application layer and the application framework layer run in a virtual machine. The virtual machine executes the Java files of the application layer and the application framework layer as binary files. The virtual machine is used to perform functions such as object life cycle management, stack management, thread management, security and exception management, and garbage collection.

[0164] The system library may include multiple functional modules, such as surface manager, media library, 3D graphics processing library (such as OpenGL ES), 2D graphics engine (such as SGL), etc.

[0165] The surface manager is used to manage the display subsystem and provide the fusion of 2D and 3D layers for multiple applications.

[0166] The media library supports playback and recording of a variety of commonly used audio and video formats, as well as static image files, etc. The media library can support a variety of audio and video encoding formats, such as: MPEG4, H.264, MP3, AAC, AMR, JPG, PNG, etc.

[0167] The 3D graphics processing library is used to implement 3D graphics drawing, image rendering, compositing, and layer processing.

[0168] A 2D graphics engine is a drawing engine for 2D drawings.

[0169] The kernel layer is the layer between hardware and software. The kernel layer contains at least display driver, camera driver, audio driver, and sensor driver.

[0170] The following is an illustrative description of the workflow of the software and hardware of the electronic device 100 in conjunction with capturing a photo scene.

[0171] When the touch sensor 180K receives a touch operation, the corresponding hardware interrupt is sent to the kernel layer. The kernel layer processes the touch operation into a raw input event (including touch coordinates, timestamp of the touch operation, and other information). The raw input event is stored in the kernel layer. The application framework layer obtains the raw input event from the kernel layer and identifies the control corresponding to the input event. For example, if the touch operation is a touch single-click operation and the control corresponding to the single-click operation is the control of the camera application icon, the camera application calls the interface of the application framework layer to start the camera application, and then starts the camera driver by calling the kernel layer to capture static images or videos through the camera 193.

[0172] In the industrial field, rotating machinery and equipment are the most widely used type of equipment to meet the needs of power output and transmission, such as motors, gearboxes, centrifugal pumps, turbine rotors, etc. Since rotating machinery and equipment are affected by various factors such as insufficient lubrication and maintenance, load impact, contaminant mixing, manufacturing and installation errors, bumps and scratches, etc., under long-term continuous load operation, key components will accelerate wear and aging, leading to frequent failures; if these failures are not discovered in time, temporary shutdowns will occur, increasing production costs, and even causing a sharp increase in the operating risk factor of rotating machinery and equipment. Therefore, it is of great significance to detect abnormalities in rotating machinery and equipment through online monitoring and formulate maintenance plans.

[0173] The current means of monitoring rotating machinery faults are mainly temperature monitoring and vibration monitoring. When equipment is abnormal, such as wear or poor lubrication of bearings and gears, changes in the contact friction characteristics between components will generate additional heat and cause the temperature to rise. By monitoring the temperature, some late-stage faults can be identified, and this method is often used for on-site point inspections. However, temperature changes are related to ambient temperature and load, and there is temperature dissipation, so it can only be used as an auxiliary means.

[0174] Vibration monitoring is currently the most commonly used sensing method. The transmission loss of mechanical vibration in a rigid body is small, and the response to equipment abnormalities or failures is more sensitive, especially for early failures, which can be identified through weak early features in the vibration signal. In addition, the failure of most rotating parts will cause an instantaneous decrease in the stiffness of the transmission system, causing an impact response of the mechanical system; and the difficulty of vibration monitoring lies in effectively processing the vibration signal to achieve robust identification of faults. Due to the differences in the vibration characteristics of equipment in different scenarios, especially for some equipment with non-constant working conditions, monitoring based on vibration amplitude and single impact monitoring are not effective; and for different equipment and different scenarios, certain expert experience is required to set thresholds, resulting in problems in technical implementation, and a large number of abnormal omissions or false alarms may occur on site.

[0175] A method for detecting rotational impact faults of mechanical equipment based on time domain analysis includes: firstly obtaining original vibration time domain waveform data of the mechanical equipment, and performing signal enhancement processing to obtain enhanced vibration time domain waveform data; performing preset normalization processing on the enhanced vibration time domain waveform data to obtain vibration time domain waveform data; dividing the vibration time domain waveform data according to preset time intervals to obtain a number of sub-vibration time domain waveform data, obtaining the maximum value of each sub-vibration time domain waveform data, and combining them to obtain a peak sequence; obtaining the mode and maximum value of the peak sequence, dividing the maximum value of the peak sequence by the mode of the peak sequence to obtain a comparison value, and when the comparison value is greater than a preset first judgment value, When the discrimination threshold is reached, the subscripts corresponding to the elements in the peak sequence whose result of dividing by the mode of the peak sequence is greater than or equal to the second discrimination threshold are obtained to obtain the suspected impact position sequence; otherwise, the detection result is no fault; the number of elements in the suspected impact position sequence is obtained, and when the number of elements in the suspected impact position sequence is less than or equal to 1, the detection result is no fault; otherwise, according to the suspected impact position sequence, the periodic impact interval is obtained, and the operating period of the rotating shaft of the mechanical equipment is obtained, and when the error between the periodic impact interval and the operating period of the rotating shaft of the mechanical equipment is within the preset error, the detection result is that there is a fault; otherwise, the detection result is no fault.

[0176] However, the above-mentioned mechanical equipment rotation impact fault detection method based on time domain analysis has the following disadvantages:

[0177] 1) Using the segmented signal peak sequence as a feature, since the local peak is easily affected by interference and has large random changes, it is impossible to ensure the continuous correlation between the indicator and the fault characteristics;

[0178] 2) A two-layer threshold discrimination method is used for fault determination. The threshold is configured based on experience, lacks a guiding method, and has no corresponding adaptation method for scenarios with historical data;

[0179] 3) When judging the fault, it is necessary to input the operation cycle of the equipment's rotating shaft as a basis for judgment. This condition cannot be met in a large number of scenarios. In addition, since the peak value is easily disturbed, the cycle matching fails and the fault identification accuracy is not high.

[0180] A shock detection method includes: obtaining a vibration signal of a device to be tested within a predetermined time, determining a vibration time domain waveform according to the vibration signal; determining a positive envelope waveform of the vibration time domain waveform; dividing multiple peaks on the positive envelope waveform into two categories according to the amplitude, and taking the average value of the amplitude of the first category of peaks as the shock energy, wherein the amplitude of the first category of peaks is greater than the amplitude of the second category of peaks; dividing the peaks on the positive envelope waveform into multiple peak groups according to the time interval between adjacent peaks, taking the total number of peak groups as the shock density, and taking the average value of the duration of each peak group as the shock duration; determining the main frequency component of the positive envelope waveform, and taking the frequency of the main frequency component as the shock frequency; determining the shock type and the severity of the shock of the device to be tested according to the shock frequency, shock energy, shock density and shock duration.

[0181] However, the above impact detection method has the following disadvantages:

[0182] 1) Impact energy, impact density, and impact duration are used as quantitative measurement criteria for impact failures. Impact energy is a dimensionless indicator, and differences in vibration intensity caused by changes in working conditions will also cause significant fluctuations in this indicator. Impact density is related to the type of failure, and as a measure of the degree of failure, it will affect the judgment of the results.

[0183] 2) Characteristic indicators that characterize the impact type and impact severity are given, while the evaluation criteria for impact anomalies are determined by engineers. When used on site, high data experience is required and usability is poor.

[0184] Based on the above technical problem, an embodiment of the present invention provides a vibration abnormality discrimination threshold updating system. Figure 3 An architectural diagram of a vibration abnormality discrimination threshold updating system provided in an embodiment of the present invention.

[0185] like Figure 3 As shown, the vibration abnormality discrimination threshold update system 200 includes: a target device 201, a vibration data acquisition module 202, a data storage module 203, an online monitoring module 204, an offline update module 205, a parameter configuration module 206, a result output module 207 and a label verification module 208.

[0186] The target device 201 includes a rotating mechanical device, such as a reducer, a motor, a pump, a fan or a turbine.

[0187] A single-point or multi-point vibration sensor is installed on the target device 201 , and the vibration sensor is electrically connected to the vibration data acquisition module 202 .

[0188] The vibration data acquisition module 202 is used to obtain a vibration signal of the target device 201 based on set sampling parameters, where the vibration signal is a vibration acceleration signal or a vibration velocity signal.

[0189] The online monitoring module 204 is used to calculate the impact characteristics and harmonic characteristics based on the real-time collected vibration signals, use the obtained impact characteristics and harmonic characteristics to perform online sample labeling and sample anomaly identification, and further realize equipment alarm identification through the sample anomaly identification results.

[0190] The offline update module 205 is used to perform offline verification of sample annotation and optimization of sample anomaly discriminator parameters based on historical data.

[0191] The data storage module 203 is used to store information such as vibration signals acquired in real time, characteristic indicators calculated based on the vibration signals, automatic or manual sample labeling results, etc.

[0192] The parameter configuration module 206 is used to configure and initialize acquisition parameters and model parameters.

[0193] The result output module 207 is used for outputting and displaying alarm results, displaying sample abnormality identification results, etc.

[0194] The annotation verification module 208 is used for displaying sample annotations, displaying signal analysis results, inputting annotation verification results, etc.

[0195] The electronic device includes a vibration data acquisition module 202, a data storage module 203, an online monitoring module 204, an offline update module 205, a parameter configuration module 206, a result output module 207 and a label verification module 208. The hardware structure and software structure of the electronic device provided by the embodiment of the present invention can be found in Figure 1 and Figure 2 Related description about electronic device 100.

[0196] Figure 4 for Figure 3 The schematic diagram of the work flow of the vibration abnormality discrimination threshold update system is as follows: Figure 4 As shown in FIG. 1 , the workflow of the vibration abnormality discrimination threshold updating system mainly includes the following four steps S1-S4:

[0197] S1: Data acquisition parameter setting and vibration signal acquisition.

[0198] Install vibration sensors on target equipment, which is rotating mechanical equipment. Target equipment includes but is not limited to reducers, motors, gear boxes, fans, pumps or turbine rotors, etc.

[0199] The installation locations of the vibration sensor include but are not limited to the device driving end, the device non-driving end, the device top, the device bottom and other parts connected to the device body.

[0200] The sensor can be a wired sensor or a wireless sensor.

[0201] The acquisition parameters are set according to actual needs. The embodiment of the present invention does not limit the parameters, but it is recommended that the sampling frequency of the vibration sensor is greater than or equal to 2000 Hz, the sampling time is greater than or equal to 0.5 s, and the sampling interval is greater than or equal to 5 h.

[0202] The electronic device obtains the vibration signal collected by the vibration sensor and stores the historical data, that is, the historical vibration signal.

[0203] S2: Calculate the impact characteristics and harmonic characteristics.

[0204] Among them, the impact characteristic is robust kurtosis.

[0205] Among them, the harmonic characteristic is the envelope harmonic ratio.

[0206] Exemplarily, the segment signal length L is set, and the vibration signal x=(x1, x2, ..., x N ) is segmented with L as the step size to obtain P segments of vibration signal x 1 ,x 2 ,…,x P , P represents the integer of N / L; calculate the kurtosis of each segment signal respectively, and obtain the first kurtosis sequence K=(k1,k2,…,k P );According to the first kurtosis sequence, calculate the mean m of the first kurtosis sequence K and variance σ K ; Determine whether there is k in the first kurtosis sequence i ∈K,i=1,2,...,P and m K The deviation exceeds 3σ K If it does not exist, the mean of the first kurtosis sequence is used as the robust kurtosis; if it exists, all the sequences in the first kurtosis sequence K that are related to m are removed. K Deviation exceeds 3σ K The kurtosis value of the range is obtained to obtain the second kurtosis sequence. All kurtosis values ​​in the second kurtosis sequence are equal to m K The deviation is within 3σ K In the range, the robust kurtosis RK is obtained x =m K .

[0207] Exemplarily, the envelope signal x of the vibration signal x is calculated using Hilbert transform env :

[0208]

[0209] in, represents the vibration signal after removing the mean value, and hilbert(·) represents the Hilbert transform. Then, based on the envelope signal x envThe autocorrelation sequence C i Calculate HR x :

[0210]

[0211] HR x =max(C i ),i0 <i<N

[0212] in,

[0213] S3: Sample anomaly identification, alarm identification and sample labeling.

[0214] Exemplarily, the rules for sample abnormality identification are as follows:

[0215]

[0216] Among them, RK x and HR x They represent the robustness kurtosis and envelope harmonic rate of S2 output, RK th and HR th Respectively represent the first threshold and the second threshold. The first threshold is the robust kurtosis abnormality discrimination threshold; the second threshold is the envelope harmonic rate abnormality discrimination threshold.

[0217] Exemplarily, if the abnormality determination result of the sample is normal, no alarm is issued; if the abnormality determination result of the sample is abnormal, an alarm is issued.

[0218] Exemplarily, the alarm judgment conditions are as follows: starting from the current sample, trace back to the samples under the first value of the operating state to obtain multiple samples. If the proportion of abnormal samples in the multiple samples is greater than or equal to the second value, an alarm is issued, otherwise no alarm is issued.

[0219] Exemplarily, sample labeling refers to labeling samples in the vibration signal based on a three-segment sample labeling strategy according to the impact characteristics and the harmonic characteristics. Figure 5 is a schematic diagram of the updating process of the first threshold and the second threshold, as shown in Figure 5 As shown, a characteristic coordinate system is established from two dimensions of impulse and harmonicity, a first interval (also called envelope harmonic rate interval) is set in the harmonic dimension, and the first interval division points include a first division point and a second division point.

[0220] For example, the first segmentation point is HR1, HR1=0.1(1+ε HR ); The second dividing point is HR2, HR2 = 0.3 (1 + ε HR ). Among them, the margin coefficient ε HR ∈[-0.2,0.2], ε HRFine-tune according to the target equipment operating conditions; the more stable the operating conditions, the smaller the value.

[0221] Exemplarily, the absolute value of the numerical value of the second division point is greater than the absolute value of the numerical value of the first division point.

[0222] Similarly, if Figure 5 As shown, a second interval (also called robust kurtosis interval) is set in the impact dimension, and the second interval segmentation points include a third segmentation point and a fourth segmentation point.

[0223] For example, the third segmentation point is RK1, RK1=3(1+ε RK ); the fourth dividing point is RK2, RK2 = 3 (1.5 + ε R x). Among them, the margin coefficient ε RK ∈[0,0.5],ε RK Fine-tune according to the target equipment operating conditions; the more stable the operating conditions, the smaller the value.

[0224] According to the above-mentioned S2 output impact characteristic calculation and harmonic characteristic, the samples in the vibration signal are labeled based on the following three-stage sample labeling strategy:

[0225]

[0226] Therefore, the three-stage sample labeling strategy includes: when the impact characteristics and harmonic characteristics meet the first condition, the sample is labeled as normal; when the impact characteristics and harmonic characteristics meet the second condition, the sample is labeled as abnormal; when the impact characteristics and harmonic characteristics meet the third condition, the sample is labeled as pending.

[0227] Exemplarily, the first condition includes: the impact characteristic is less than or equal to the value of the third division point, or the harmonic characteristic is less than or equal to the value of the first division point.

[0228] Exemplarily, the second condition includes: the impact characteristic is greater than or equal to the value of the fourth division point, and the harmonic characteristic is greater than or equal to the value of the second division point.

[0229] Exemplarily, the third condition is different from the first condition and the second condition, that is, the value ranges of the impulse characteristic and the harmonic characteristic satisfy neither the first condition nor the second condition.

[0230] S4: Annotation verification and threshold update.

[0231] Optionally, a threshold update cycle or an upper limit on the number of pending labeled samples is set to trigger labeling verification and threshold optimization. When the corresponding conditions are met, labeling verification and threshold update are performed.

[0232] Specifically, the electronic device performs label verification on the first labeled sample sequence through automatic review and manual review to obtain a second labeled sample sequence, and updates the first threshold and the second threshold according to the second labeled sample sequence.

[0233] Exemplarily, the minimum entropy deconvolution method is first used to perform impact feature enhancement on the first labeled sample sequence to obtain an enhanced signal sequence; based on the enhanced signal sequence, characteristic indicators of the enhanced signal sequence are obtained; if the characteristic indicators of the enhanced signal belong to the normal area, the labels of the labeled samples marked as pending in the first labeled sample sequence are corrected to normal to obtain a second labeled sample sequence, otherwise the first labeled sample sequence is manually verified to obtain a second labeled sample sequence.

[0234] Exemplarily, the characteristic indicators of the enhanced signal sequence include the robustness kurtosis and envelope harmonic rate of the enhanced signal sequence.

[0235] Exemplarily, the normal region includes: the robustness kurtosis of the enhanced signal is less than or equal to the value of the fourth segmentation point, or the envelope harmonic rate of the enhanced signal is less than or equal to the value of the second segmentation point.

[0236] Exemplarily, the process of manually reviewing and verifying the first labeled sample sequence to obtain the second labeled sample sequence includes: obtaining a first envelope spectrum based on an enhanced signal sequence; obtaining a second envelope spectrum based on the first labeled sample sequence; determining whether the double frequency characteristics of the first envelope spectrum and the second envelope spectrum in the low-frequency domain are consistent; if it is determined that the double frequency characteristics of the first envelope spectrum and the second envelope spectrum in the low-frequency domain are consistent, in response to the user's operation, the annotation of the labeled samples marked as pending in the first labeled sample sequence is corrected to abnormal, and the second labeled sample sequence is obtained; if it is determined that the double frequency characteristics of the first envelope spectrum and the second envelope spectrum in the low-frequency domain are inconsistent, in response to the user's operation, the annotation of the labeled samples marked as pending in the second labeled sample sequence is corrected to normal, and the second labeled sample sequence is obtained.

[0237] like Figure 5 As shown, multiple labeled samples and their characteristic indicators within a preset interval are obtained from the second labeled sample sequence; the fourth condition and the fifth condition are obtained according to the characteristic indicators and the preset interval of the multiple labeled samples; according to the fourth condition, multiple first normal labeled samples are obtained from the multiple labeled samples; a harmonic rate sequence is obtained by extracting the harmonic rate values ​​of the multiple first normal labeled samples; the second threshold is updated according to the harmonic rate sequence; according to the fifth condition, multiple second normal labeled samples are obtained from the multiple labeled samples; a third kurtosis sequence is obtained by extracting the robust kurtosis values ​​of the multiple second normal labeled samples; and the first threshold is updated according to the third kurtosis sequence.

[0238] Exemplarily, the interval division point of the preset interval in the harmonic dimension is the first interval division point.

[0239] Exemplarily, the interval division point of the preset interval in the impact dimension is the second interval division point.

[0240] Exemplarily, multiple labeled samples are labeled as normal.

[0241] Exemplarily, the fourth condition is: RK x >RK1 and RK x >RK1+(HR x -HR1)×(RK2-RK1) / (HR2-HR1), that is Figure 5 Normally labeled samples within the dotted area on the left.

[0242] For example, the fifth condition is: HR x >HR1 and RK x <RK1+(HR x -HR1)×(RK2-RK1) / (HR2-HR1), that is Figure 5 Normally labeled samples within the dotted area on the right.

[0243] Exemplarily, the updated second threshold is obtained according to the harmonic rate sequence, and the updated first threshold is obtained according to the third kurtosis sequence, as follows:

[0244] HR th =mean(HR normal )+3σ(HR normal )

[0245] R th =mean(RK normal )+3σ(RK normal )

[0246] Among them, HR normal Represents the harmonic rate sequence, HR th represents the updated second threshold, RK normal represents the third kurtosis sequence, RK th represents the first threshold after update, mean(·) represents the calculated mean, and σ(·) represents the calculated variance.

[0247] Exemplarily, the above-mentioned vibration abnormality discrimination threshold updating system and its workflow are applied to the centrifugal pump shock abnormality identification solution, specifically as follows:

[0248] M1. Select the vibration acceleration data of the centrifugal pump drive end for 5 consecutive months and import it into the vibration abnormality discrimination threshold update system for analysis. There are a total of 10671 samples, the single sample signal length is 16384 sampling points, and the sampling frequency is 25600Hz.

[0249] M2. Calculate the robust kurtosis and envelope harmonic rate of each sample in turn to express the shock anomaly information of the centrifugal pump.

[0250] The robust kurtosis calculation process includes: the centrifugal pump runs at a speed of about 600 rpm, i.e., 10 Hz, and the robust kurtosis segment length is set to 0.2×sampling frequency / rotation frequency, i.e., L=512, and each sample can be divided into 32 segments. Figure 6 It is a schematic diagram of the segmented signal and its kurtosis curve of a normal state sample, such as Figure 6 As shown, the signal has obvious impact interference at the 36th segment sample position, the overall kurtosis value of the signal is 11.81, the segment kurtosis mean is 4.32, and the robust kurtosis is 2.99. Therefore, the robust kurtosis is closer to the kurtosis range of the normal signal (the kurtosis of white noise is 3).

[0251] The envelope harmonic rate calculation process includes: Figure 7 is a schematic diagram of the envelope signal and its autocorrelation sequence of a normal sample, such as Figure 7 As shown, in the obtained autocorrelation coefficient sequence, the time delay of the first negative value is 634, and the maximum value of the autocorrelation coefficient sequence with a time delay greater than 634 is 0.068, so the envelope harmonic rate value is 0.068.

[0252] M3. Samples are labeled and stored based on a three-stage sample labeling strategy, and sample anomaly and alarm discrimination are performed by constructing a threshold discriminator based on robust kurtosis and envelope harmonic rate.

[0253] The sample labeling process includes: due to the stable working condition of the centrifugal pump, HR1 and HR2 in the characteristic coordinate system are taken as 0.09 and 0.27 respectively, and RK1 and RK2 are taken as 3 and 4.5 respectively; according to the three-stage sample labeling strategy, if the robustness kurtosis of the sample is greater than or equal to 4.5 and the envelope harmonic rate is greater than or equal to 0.27, it is marked as abnormal; if the robustness kurtosis of the sample is less than or equal to 3 or the envelope harmonic rate is less than or equal to 0.09, it is marked as normal; the remaining samples are marked as pending; Figure 8 It is a schematic diagram of the three-segment annotation based on the impact characteristic coordinate system, such as Figure 8 As shown, the number of samples marked as normal is 9369, the number of samples marked as abnormal is 794, and the number of samples marked as pending is 508.

[0254] Among them, the sample anomaly discrimination process includes: when the sample anomaly discriminator is configured for the first time, in order to avoid frequent false alarms, the robust kurtosis anomaly discrimination threshold, that is, the first threshold is taken as RK2, and the envelope harmonic rate anomaly discrimination threshold, that is, the second threshold is taken as HR2, that is, the first threshold is 4.5, and the second threshold is 0.27, then the number of identified abnormal samples is 794;

[0255] The alarm identification process includes: setting the alarm condition to trace back 5 samples in the running state from the current sample. If all 5 samples are abnormal, an alarm is triggered. The total number of alarms is 406 times, the false alarm rate is 0 times, and the first alarm point is the 229th sample.

[0256] M4, sample labeling verification and sample anomaly discriminator update.

[0257] First, all historical data samples for 5 months were labeled and verified, with a total of 508 samples to be verified, of which 370 samples completed label correction through automatic review, and 138 samples completed label correction through manual review. Fig. 9 A schematic diagram of the sample verification results, such as Fig. 9 As shown, the number of samples marked as abnormal after verification is 1064, and the number of samples marked as normal is 9607;

[0258] Secondly, the sample abnormality discriminator is updated, that is, the first threshold and the second threshold are updated. Specifically, the samples that meet RK are extracted from the 9607 samples marked as normal. x >RK1 and RK x >RK1+(HR x -HR1)×(RK2-RK1) / (HR2-HR1) samples are used to update the envelope harmonic rate threshold and extract the samples that meet HR x >HR1 and RK x <RK1+(HR x -HR1)×(RK2-RK1) / (HR2-HR1) samples are used for robust kurtosis threshold update. Fig.10 A schematic diagram of the recognition results for the threshold-optimized dataset, as shown in Fig.10 As shown, there are 3308 samples, namely, the HR data set, which can be used for the second threshold update; there are 1412 samples, namely, the RK data set, which can be used for the first threshold update; the threshold update result is RK th =3.23, HR th =0.07, and use this result to update the parameters of the sample abnormality discriminator. Fig.11 is a schematic diagram of the sample anomaly recognition results based on the updated threshold, as shown in Fig.11 As shown, the abnormal sample recognition rate is 100%, an increase of 25.38%, and the abnormal recognition error rate is only 0.066%. If 5 consecutive sample abnormalities are set to trigger an alarm, the number of alarms is 1050, the false alarm rate is 0, and the first alarm point is sample 191.

[0259] The embodiment of the present invention constructs a characteristic coordinate system through robust kurtosis and envelope harmonic rate, which can well quantify the cyclic impact abnormal information of rotating mechanical equipment. The linear threshold discrimination scheme established using the two-dimensional feature distribution can avoid the overfitting problem that may be caused by high-dimensional features and nonlinear methods, and is well adapted to specific scenario data. The accuracy of sample abnormality identification can be greatly improved.

[0260] The embodiment of the present invention performs three-segment division and labeling based on the feature coordinate system, and classifies the sample labels into normal, abnormal or pending. It can extract weak feature samples near the time point of degradation mutation of rotating mechanical equipment, and based on the envelope spectrum analysis multiple frequency comparison method, automatically or manually verify the pending sample labels in a layered manner, thereby introducing auxiliary verification. Due to the differences in data distribution in different scenarios or equipment, this method reduces the workload and scope of manual labeling, and can better clarify the boundary samples in the normal state, thereby avoiding the problem of negative optimization of the model caused by incorrect labeling.

[0261] The embodiment of the present invention uses a threshold updating method using normal state samples based on feature coordinate system partitioning. Since equipment failure is a low-probability event, using only normal samples for threshold updating avoids the requirement for failure data, is more in line with actual needs, and has better recognition capabilities for unknown samples that deviate from the overall distribution of state data.

[0262] In summary, the embodiment of the present invention uses robust kurtosis and envelope harmonic rate to quantify the impact abnormal components in the vibration signal from the perspective of impact and harmonicity, and can use a very small number of indicators to achieve a clear distinction between the distribution of normal samples and abnormal samples in the constructed characteristic coordinate system, so that the cyclic impact characteristics of the vibration signal can be better expressed. In the established characteristic coordinate system, the distribution of normal and abnormal states is more separable. In combination with the constructed characteristic coordinate system, the embodiment of the present invention formulates a corresponding sample labeling strategy and a threshold update strategy. By dividing the characteristic coordinate system into three sections to agree on normal, abnormal, and pending labeling areas, and using artificial capabilities to assist in verifying pending labeling samples, the scope of manual labeling is limited, the workload and blindness of labeling are reduced, the reliability of sample labeling is improved, and the negative optimization problem caused by sample labeling errors is avoided; by optimizing the threshold parameters based on the partitioned samples of the characteristic coordinate system, a single linear threshold estimation can be performed for each parameter, avoiding the indicator correlation problem that needs to be considered in the joint estimation and the overfitting problem of the nonlinear method, making the robustness of the threshold optimization better, and improving the effect of threshold updating using labeled samples.

[0263] In summary, the embodiment of the present invention provides a vibration anomaly discrimination threshold update system for rotating mechanical equipment, which can be used to optimize the parameters of the sample anomaly discriminator in combination with field data. Specifically, an indicator is constructed from the two characteristic dimensions of impact and harmonicity to characterize the cyclic impact information of the vibration signal, forming a characteristic coordinate system; sample anomalies are identified by constructing a threshold discriminator, and equipment anomalies are discriminated based on the continuity criterion; in addition, in order to meet the generalization ability of the model, the three-segment annotation area is divided based on the characteristic coordinate system for annotation verification, avoiding the model optimization deviation caused by the annotation error, and effectively reducing the number of samples to be labeled. Finally, the threshold discriminator parameters are optimized through sample partitioning to achieve model tuning.

[0264] Based on the above-mentioned vibration abnormality discrimination threshold updating system, an embodiment of the present invention provides a vibration abnormality discrimination threshold updating method. Fig.12 A flow chart of a vibration abnormality discrimination threshold updating method provided by an embodiment of the present invention. Fig.12 As shown, the method includes:

[0265] Step 302: Obtain impact characteristics and harmonic characteristics according to the vibration signal of the target device.

[0266] Step 304: perform abnormality discrimination and alarm discrimination on samples in the vibration signal according to the impact characteristics, harmonic characteristics, the first threshold value, and the second threshold value.

[0267] Step 306: Label and store samples in the vibration signal according to the impulse characteristics and the harmonic characteristics based on a three-segment sample labeling strategy to obtain a first labeled sample sequence.

[0268] Step 308: Perform annotation verification on the first annotated sample sequence through automatic review and manual review to obtain a second annotated sample sequence, and update the first threshold and the second threshold according to the second annotated sample sequence.

[0269] In the technical solution of the vibration abnormality discrimination threshold updating method provided by the embodiment of the present invention, the method includes: obtaining impact characteristics and harmonic characteristics according to the vibration signal of the target equipment; performing abnormality discrimination and alarm discrimination on samples in the vibration signal according to the impact characteristics, harmonic characteristics, a first threshold and a second threshold; labeling and storing samples in the vibration signal based on a three-segment sample labeling strategy according to the impact characteristics and harmonic characteristics to obtain a first labeled sample sequence; labeling and verifying the first labeled sample sequence through automatic review and manual review to obtain a second labeled sample sequence, and updating the first threshold and the second threshold according to the second labeled sample sequence, establishing indicators from the impact and harmonic characteristics to better express the cyclic impact characteristics of the vibration signal, thereby improving the accuracy of vibration monitoring.

[0270] Fig.13A flowchart of another vibration abnormality discrimination threshold updating method provided by an embodiment of the present invention. Fig.13 As shown, the method includes:

[0271] Step 402: Acquire a vibration signal of the target device.

[0272] In the embodiment of the present invention, each step is performed by an electronic device. The hardware structure and software structure of the electronic device provided by the embodiment of the present invention can be found in Figure 1 and Figure 2 Related description about electronic device 100.

[0273] Exemplarily, the target device is a rotating mechanical device.

[0274] Exemplarily, the target equipment includes a reducer, a motor, a gearbox, a fan, a pump or a turbine rotor.

[0275] In some possible embodiments, step 402 includes: acquiring a vibration signal of the target device through a vibration sensor.

[0276] Exemplarily, a single-point or multi-point vibration sensor is installed on the target device; the installation position of the vibration sensor includes the device driving end, the device non-driving end, the device top, the device bottom or other parts connected to the device body.

[0277] Exemplarily, the vibration sensor includes a wired sensor or a wireless sensor.

[0278] Exemplarily, the sampling frequency of the vibration sensor is greater than or equal to 2000 Hz.

[0279] Exemplarily, the sampling duration of the vibration sensor is greater than or equal to 0.5 s.

[0280] Exemplarily, the sampling interval of the vibration sensor is less than or equal to 5 hours.

[0281] Step 404: Obtain impact characteristics and harmonic characteristics according to the vibration signal of the target device.

[0282] Exemplarily, the impact characteristic is robust kurtosis.

[0283] Exemplarily, the harmonic characteristic is an envelope harmonic ratio.

[0284] In some possible embodiments, Fig.14 As shown, according to the vibration signal of the target device, the impact characteristics are obtained, including:

[0285] Step A02: segment the vibration signal according to the segment signal length to obtain multiple vibration signal sequences.

[0286] Exemplarily, the segment signal length L is set, and the vibration signal x=(x1, x2, ..., x N ) is segmented with L as the step size to obtain P segments of vibration signal x 1 ,x 2 ,…,x P , P means N / L rounded.

[0287] Step A04: Obtain a first kurtosis sequence corresponding to the multiple vibration signal sequences according to the multiple vibration signal sequences.

[0288] Exemplarily, the kurtosis of each signal in the multiple vibration signal sequences is calculated respectively to obtain the first kurtosis sequence K=(k1, k2, ..., k P ).

[0289] Step A06: According to the first kurtosis sequence, obtain the mean and variance of the first kurtosis sequence.

[0290] Exemplarily, according to the first kurtosis sequence, the mean m of the first kurtosis sequence is calculated K and variance σ K .

[0291] Step A08, determine whether there is a kurtosis value in the first kurtosis sequence whose difference from the mean value exceeds three times the variance. If so, execute step A10; if not, execute step A16.

[0292] Exemplarily, it is determined whether there is k in the first kurtosis sequence i ∈K,i=1,2,...,P and m K The deviation exceeds 3σ K .

[0293] Step A10: Remove the kurtosis values ​​whose difference from the mean exceeds three times of the variance in the first kurtosis sequence to obtain a second kurtosis sequence.

[0294] For example, if it is determined that there are kurtosis values ​​in the first kurtosis sequence whose difference from the mean exceeds three times the variance, then all kurtosis values ​​in the first kurtosis sequence K that are different from m K Deviation exceeds 3σ K The kurtosis value of the range is obtained to obtain the second kurtosis sequence.

[0295] Step A12: Obtain a mean value of the second kurtosis sequence according to the second kurtosis sequence.

[0296] Exemplarily, the deviations of all kurtosis values ​​in the second kurtosis sequence from the mean of the second kurtosis sequence are all within a range of three times the variance of the second kurtosis sequence, and the mean of the second kurtosis sequence is obtained.

[0297] Step A14: Use the mean of the second kurtosis sequence as the impact feature. Continue to step B02.

[0298] Exemplarily, the mean of the second kurtosis sequence is used as the impact feature.

[0299] Step A16: Use the mean of the first kurtosis sequence as the impact feature. Continue to step B02.

[0300] Exemplarily, if it is determined that there is no kurtosis value in the first kurtosis sequence whose difference from the mean exceeds three times of the variance, the mean of the first kurtosis sequence is used as the impact feature.

[0301] In some possible embodiments, Fig.15 As shown, based on the vibration signal of the target device, the harmonic characteristics are obtained, including:

[0302] Step B02: Obtain an envelope signal of the vibration signal according to the Hilbert transform and the vibration signal.

[0303] Exemplarily, the envelope signal x of the vibration signal x is calculated using Hilbert transform env :

[0304]

[0305] in, represents the vibration signal after removing the mean value, and hilbert(·) represents the Hilbert transform.

[0306] Step B04: Obtain harmonic characteristics based on the autocorrelation sequence of the envelope signal.

[0307] For example, based on the envelope signal x env The autocorrelation sequence C i Calculate HR x :

[0308]

[0309] HR x =max(C i ),i0 <i<N

[0310] in,

[0311] Step 406: According to the impact characteristics, harmonic characteristics, the first threshold value and the second threshold value, perform abnormality judgment and alarm judgment on the samples in the vibration signal.

[0312] Exemplarily, a sample abnormality discriminator is constructed based on the impact characteristics, the harmonic characteristics, the first threshold value and the second threshold value, and the sample abnormality discriminator is used to perform abnormality discrimination and alarm discrimination on samples in the vibration signal.

[0313] In some possible embodiments, Fig.16 As shown, step 406 specifically includes:

[0314] Step 406a, determine whether the impulse characteristic is greater than the first threshold and whether the harmonic characteristic is greater than the second threshold. If so, execute step 406b; if not, execute step 406g.

[0315] Exemplarily, the rules for sample abnormality identification are as follows:

[0316]

[0317] Among them, RK x and HR x They represent the robustness kurtosis and envelope harmonic rate, RK th and HR th Respectively represent the first threshold and the second threshold. The first threshold is the robust kurtosis abnormality discrimination threshold; the second threshold is the envelope harmonic rate abnormality discrimination threshold.

[0318] Step 406b: The abnormality determination result is abnormal.

[0319] Step 406c: trace back the first number of samples in the running state from the current sample to obtain multiple samples.

[0320] Exemplarily, if the abnormality determination result of the sample is normal, no alarm is issued; if the abnormality determination result of the sample is abnormal, an alarm is issued.

[0321] Exemplarily, the alarm judgment conditions are as follows: starting from the current sample, trace back to the samples under the first value of the operating state to obtain multiple samples. If the proportion of abnormal samples in the multiple samples is greater than or equal to the second value, an alarm is issued, otherwise no alarm is issued.

[0322] Step 406d, determine whether the abnormal sample ratio of the multiple samples is greater than or equal to the second value, if so, execute step 406e; if not, execute step 406f.

[0323] Step 406e: issue an alarm and continue to execute step 408.

[0324] Step 406f: No alarm is issued, and the process continues with step 408.

[0325] Step 406g: The abnormality determination result is normal, and the process continues with step 408.

[0326] Step 408: Label and store samples in the vibration signal according to the impulse characteristics and the harmonic characteristics based on a three-segment sample labeling strategy to obtain a first labeled sample sequence.

[0327] In some possible embodiments, Fig.17 As shown, step 408 specifically includes:

[0328] Step 4082: Establish a characteristic coordinate system based on the impact dimension and the harmonic dimension.

[0329] Step 4084: Set a first interval division point in the harmonic dimension, and set a second interval division point in the impact dimension.

[0330] Exemplarily, the first interval division points include a first division point and a second division point, and the second interval division points include a third division point and a fourth division point.

[0331] like Figure 5 As shown, a characteristic coordinate system is established from two dimensions of impulse and harmonicity, a first interval (also called envelope harmonic rate interval) is set in the harmonic dimension, and the first interval division points include a first division point and a second division point.

[0332] For example, the first segmentation point is HR1, HR1=0.1(1+ε HR ); The second dividing point is HR2, HR2 = 0.3 (1 + ε HR ). Among them, the margin coefficient ε HR ∈[-0.2,0.2], ε HR Fine-tune according to the target equipment operating conditions; the more stable the operating conditions, the smaller the value.

[0333] Exemplarily, the absolute value of the numerical value of the second division point is greater than the absolute value of the numerical value of the first division point.

[0334] Similarly, if Figure 5 As shown, a second interval (also called robust kurtosis interval) is set in the impact dimension, and the second interval segmentation points include a third segmentation point and a fourth segmentation point.

[0335] For example, the third segmentation point is RK1, RK1=3(1+ε RK ); the fourth dividing point is RK2, RK2 = 3 (1.5 + ε RK ). Among them, the margin coefficient ε RK ∈[0,0.5],ε RK Fine-tune according to the target equipment operating conditions; the more stable the operating conditions, the smaller the value.

[0336] Step 4086: Label and store samples in the vibration signal based on a three-segment sample labeling strategy according to the feature coordinate system, the first interval segmentation point, the second interval segmentation point, the impact feature, and the harmonic feature to obtain a first labeled sample sequence.

[0337] Exemplarily, samples in the vibration signal are labeled based on the following three-segment sample labeling strategy:

[0338]

[0339] Therefore, the three-stage sample labeling strategy includes: when the impact characteristics and harmonic characteristics meet the first condition, the sample is labeled as normal; when the impact characteristics and harmonic characteristics meet the second condition, the sample is labeled as abnormal; when the impact characteristics and harmonic characteristics meet the third condition, the sample is labeled as pending.

[0340] Exemplarily, the first condition includes: the impact characteristic is less than or equal to the value of the third division point, or the harmonic characteristic is less than or equal to the value of the first division point.

[0341] Exemplarily, the second condition includes: the impact characteristic is greater than or equal to the value of the fourth division point, and the harmonic characteristic is greater than or equal to the value of the second division point.

[0342] Exemplarily, the third condition is different from the first condition and the second condition, that is, the value ranges of the impulse characteristic and the harmonic characteristic satisfy neither the first condition nor the second condition.

[0343] Step 410: Perform annotation verification on the first annotated sample sequence through automatic review and manual review to obtain a second annotated sample sequence, and update the first threshold and the second threshold according to the second annotated sample sequence.

[0344] Optionally, a threshold update period or an upper limit on the number of to-be-determined annotated samples is set to trigger annotation verification and threshold optimization, and when the corresponding conditions are met, step 410 is executed.

[0345] In some possible embodiments, Fig.18 As shown, the second labeled sample sequence is obtained by automatically checking the first labeled sample sequence, which specifically includes the following steps:

[0346] Step 410a: perform impulse feature enhancement on the first labeled sample sequence by using a minimum entropy deconvolution method to obtain an enhanced signal sequence.

[0347] Step 410b: Obtain a characteristic index of the enhanced signal sequence according to the enhanced signal sequence.

[0348] Exemplarily, the characteristic indicators of the enhanced signal sequence include the robustness kurtosis and envelope harmonic rate of the enhanced signal sequence.

[0349] Step 410c: determine whether the characteristic index of the enhanced signal sequence belongs to the normal area. If so, execute step 410d; if not, continue to execute step 410e.

[0350] Exemplarily, the normal region includes: the robustness kurtosis of the enhanced signal is less than or equal to the value of the fourth segmentation point, or the envelope harmonic rate of the enhanced signal is less than or equal to the value of the second segmentation point.

[0351] Step 410d: Correct the labels of the labeled samples marked as pending in the first labeled sample sequence to normal, to obtain a second labeled sample sequence.

[0352] If it is determined that the characteristic index of the enhanced signal sequence belongs to the normal region, the marking of the marked samples marked as pending in the first marked sample sequence is corrected to normal to obtain a second marked sample sequence. The step of updating the first threshold and the second threshold according to the second marked sample sequence is continued.

[0353] In some possible embodiments, Fig.18 As shown, the first labeled sample sequence is manually verified to obtain a second labeled sample sequence, which specifically includes the following steps:

[0354] Step 410e: Obtain a first envelope spectrum according to the enhanced signal sequence.

[0355] If it is determined that the characteristic indicator of the enhanced signal sequence does not belong to the normal region, a first envelope spectrum is obtained according to the enhanced signal sequence.

[0356] Step 410f, obtaining a second envelope spectrum according to the first labeled sample sequence.

[0357] Step 410g, determine whether the frequency-doubled characteristics of the first envelope spectrum and the second envelope spectrum in the low-frequency domain are consistent. If so, execute step 410h; if not, execute step 410i.

[0358] Step 410h: In response to the user's operation, the annotations of the annotation samples marked as pending in the first annotation sample sequence are corrected as abnormal, thereby obtaining a second annotation sample sequence. The steps of updating the first threshold and the second threshold according to the second annotation sample sequence are continued.

[0359] If it is determined that the first envelope spectrum and the second envelope spectrum have consistent frequency doubling characteristics in the low frequency domain, in response to the user's operation, the annotation of the labeled sample marked as pending in the first labeled sample sequence is corrected to be abnormal, thereby obtaining a second labeled sample sequence.

[0360] Step 410i: In response to the user's operation, the labels of the labeled samples marked as pending in the second labeled sample sequence are corrected to normal, thereby obtaining a second labeled sample sequence. The steps of updating the first threshold and the second threshold according to the second labeled sample sequence are continued.

[0361] If it is determined that the frequency doubling characteristics of the first envelope spectrum and the second envelope spectrum in the low frequency domain are inconsistent, in response to the user's operation, the labeling of the labeled samples marked as pending in the second labeled sample sequence is corrected to normal to obtain the second labeled sample sequence.

[0362] In some possible embodiments, Fig.19 As shown, updating the first threshold and the second threshold according to the second labeled sample sequence specifically includes the following steps:

[0363] Step C02: Acquire multiple annotated samples and their characteristic indicators within a preset interval from the second annotated sample sequence.

[0364] Exemplarily, the interval division point of the preset interval in the harmonic dimension is the first interval division point.

[0365] Exemplarily, the interval division point of the preset interval in the impact dimension is the second interval division point.

[0366] Exemplarily, multiple labeled samples are labeled as normal.

[0367] Step C04: Obtain the fourth condition and the fifth condition according to the characteristic indicators and preset intervals of the plurality of labeled samples.

[0368] Exemplarily, the fourth condition is: RK x >RK1 and RK x >RK1+(HR x -HR1)×(RK2-RK1) / (HR2-HR1), that is Figure 5 Normally labeled samples within the dotted area on the left.

[0369] For example, the fifth condition is: HR x >HR1 and RK x <RK1+(HR x -HR1)×(RK2-RK1) / (HR2-HR1), that is Figure 5 Normally labeled samples within the dotted area on the right.

[0370] Step C06: According to the fourth condition, a plurality of first normal labeled samples are obtained from the plurality of labeled samples.

[0371] Step C08: extracting the harmonic rate values ​​of a plurality of first normal labeled samples to obtain a harmonic rate sequence.

[0372] Step C10: Update the second threshold according to the harmonic rate sequence.

[0373] Step C12: According to the fifth condition, a plurality of second normal labeled samples are obtained from the plurality of labeled samples.

[0374] Step C14: extracting robust kurtosis values ​​of multiple second normal labeled samples to obtain a third kurtosis sequence.

[0375] Step C16: Update the first threshold according to the third kurtosis sequence.

[0376] Exemplarily, the updated second threshold is obtained according to the harmonic rate sequence, and the updated first threshold is obtained according to the third kurtosis sequence, as follows:

[0377] HR th =mean(HR normal )+3σ(HR normal )

[0378] R th =mean(RK normal )+3σ(RK normal )

[0379] Among them, HR normal Represents the harmonic rate sequence, HR th represents the updated second threshold, RK normal represents the third kurtosis sequence, RK th represents the first threshold after update, mean(·) represents the calculated mean, and σ(·) represents the calculated variance.

[0380] Among them, steps C06-C10 and steps C12-C16 can be executed successively or simultaneously; steps C12-C16 can also be executed first, and then steps C06-C10.

[0381] The embodiment of the present invention constructs a characteristic coordinate system through robust kurtosis and envelope harmonic rate, which can well quantify the cyclic impact abnormal information of rotating mechanical equipment. The linear threshold discrimination scheme established using the two-dimensional feature distribution can avoid the overfitting problem that may be caused by high-dimensional features and nonlinear methods, and is well adapted to specific scenario data. The accuracy of sample abnormality identification can be greatly improved.

[0382] The embodiment of the present invention performs three-segment division and labeling based on the feature coordinate system, and classifies the sample labels into normal, abnormal or pending. It can extract weak feature samples near the time point of degradation mutation of rotating mechanical equipment, and based on the envelope spectrum analysis multiple frequency comparison method, automatically or manually verify the pending sample labels in a layered manner, thereby introducing auxiliary verification. Due to the differences in data distribution in different scenarios or equipment, this method reduces the workload and scope of manual labeling, and can better clarify the boundary samples in the normal state, thereby avoiding the problem of negative optimization of the model caused by incorrect labeling.

[0383] The embodiment of the present invention uses a threshold updating method using normal state samples based on feature coordinate system partitioning. Since equipment failure is a low-probability event, using only normal samples for threshold updating avoids the requirement for failure data, is more in line with actual needs, and has better recognition capabilities for unknown samples that deviate from the overall distribution of state data.

[0384] In summary, the embodiment of the present invention uses robust kurtosis and envelope harmonic rate to quantify the impact abnormal components in the vibration signal from the perspective of impact and harmonicity, and can use a very small number of indicators to achieve a clear distinction between the distribution of normal samples and abnormal samples in the constructed characteristic coordinate system, so that the cyclic impact characteristics of the vibration signal can be better expressed. In the established characteristic coordinate system, the distribution of normal and abnormal states is more separable. In combination with the constructed characteristic coordinate system, the embodiment of the present invention formulates a corresponding sample labeling strategy and a threshold update strategy. By dividing the characteristic coordinate system into three sections to agree on normal, abnormal, and pending labeling areas, and using artificial capabilities to assist in verifying pending labeling samples, the scope of manual labeling is limited, the workload and blindness of labeling are reduced, the reliability of sample labeling is improved, and the negative optimization problem caused by sample labeling errors is avoided; by optimizing the threshold parameters based on the partitioned samples of the characteristic coordinate system, a single linear threshold estimation can be performed for each parameter, avoiding the indicator correlation problem that needs to be considered in the joint estimation and the overfitting problem of the nonlinear method, making the robustness of the threshold optimization better, and improving the effect of threshold updating using labeled samples.

[0385] In summary, the embodiment of the present invention provides a vibration anomaly discrimination threshold updating method for rotating mechanical equipment, which can be used to optimize the parameters of the sample anomaly discriminator in combination with field data. Specifically, an indicator is constructed from the two characteristic dimensions of impact and harmonicity to characterize the cyclic impact information of the vibration signal, forming a characteristic coordinate system; sample anomalies are identified by constructing a threshold discriminator, and equipment anomalies are discriminated based on the continuity criterion; in addition, in order to meet the generalization ability of the model, the three-segment annotation area is divided based on the characteristic coordinate system for annotation verification, avoiding the model optimization deviation caused by the annotation error, and effectively reducing the number of samples to be labeled. Finally, the threshold discriminator parameters are optimized through sample partitioning to achieve model tuning.

[0386] In the technical solution of the vibration abnormality discrimination threshold updating method provided by the embodiment of the present invention, the impact characteristic and the harmonic characteristic are obtained according to the vibration signal of the target equipment; the samples in the vibration signal are subjected to abnormality discrimination and alarm discrimination according to the impact characteristic, the harmonic characteristic, the first threshold and the second threshold; according to the impact characteristic and the harmonic characteristic, the samples in the vibration signal are labeled and stored based on the three-segment sample labeling strategy to obtain a first labeled sample sequence; the first labeled sample sequence is labeled and verified by automatic review and manual review to obtain a second labeled sample sequence, and according to the second labeled sample sequence, the first threshold and the second threshold are updated, and indicators are established from the impact characteristic and the harmonic characteristic to better express the cyclic impact characteristics of the vibration signal, thereby improving the accuracy of vibration monitoring.

[0387] Fig. 20 The electronic device 500 is a schematic diagram of a structure of an electronic device provided by an embodiment of the present invention. It should be understood that the electronic device 500 can execute each step in the above vibration abnormality discrimination threshold update method. To avoid repetition, it is not described in detail here. The electronic device 500 includes: a processing unit 501 and an acquisition unit 502.

[0388] The processing unit 501 is used to obtain the impact characteristics and the harmonic characteristics according to the vibration signal of the target device;

[0389] According to the impact feature, the harmonic feature, the first threshold and the second threshold, abnormality discrimination and alarm discrimination are performed on the samples in the vibration signal; according to the impact feature and the harmonic feature, the samples in the vibration signal are labeled and stored based on a three-segment sample labeling strategy to obtain a first labeled sample sequence; the first labeled sample sequence is labeled and verified by automatic review and manual review to obtain a second labeled sample sequence, and according to the second labeled sample sequence, the first threshold and the second threshold are updated.

[0390] Optionally, the acquisition unit 502 is configured to acquire a vibration signal of the target device.

[0391] Optionally, the impact characteristic is robust kurtosis.

[0392] Optionally, the harmonic characteristic is an envelope harmonic ratio.

[0393] Optionally, the processing unit 501 is specifically used to segment the vibration signal according to the segmented signal length to obtain multiple vibration signal sequences; obtain a first kurtosis sequence corresponding to the multiple vibration signal sequences based on the multiple vibration signal sequences; obtain the mean and variance of the first kurtosis sequence based on the first kurtosis sequence; determine whether there is a kurtosis value in the first kurtosis sequence whose difference with the mean exceeds three times the variance; if the processing unit 501 determines that there is a kurtosis value in the first kurtosis sequence whose difference with the mean exceeds three times the variance, eliminate the kurtosis value in the first kurtosis sequence whose difference with the mean exceeds three times the variance to obtain a second kurtosis sequence; obtain the mean of the second kurtosis sequence based on the second kurtosis sequence; and use the mean of the second kurtosis sequence as the impact feature.

[0394] Optionally, the processing unit 501 is specifically configured to use the mean of the first kurtosis sequence as the impact feature if the processing unit 501 determines that there is no kurtosis value in the first kurtosis sequence whose difference from the mean exceeds three times the variance.

[0395] Optionally, the processing unit 501 is specifically configured to obtain an envelope signal of the vibration signal according to a Hilbert transform and the vibration signal; and obtain the harmonic characteristic based on an autocorrelation sequence of the envelope signal.

[0396] Optionally, the processing unit 501 is specifically used to determine whether the impact characteristic is greater than the first threshold and whether the harmonic characteristic is greater than the second threshold; if the processing unit 501 determines that the impact characteristic is greater than the first threshold and the harmonic characteristic is greater than the second threshold, the abnormal judgment result is abnormal; starting from the current sample, trace back to the samples in the first value operating state to obtain multiple samples; determine whether the proportion of abnormal samples of the multiple samples is greater than or equal to the second value; if the processing unit 501 determines that the proportion of abnormal samples of the multiple samples is greater than or equal to the second value, an alarm is issued; if the processing unit 501 determines that the proportion of abnormal samples of the multiple samples is less than the second value, no alarm is issued.

[0397] Optionally, the processing unit 501 is specifically configured to determine that the abnormality determination result is normal if the processing unit 501 determines that the impact characteristic is less than or equal to the first threshold, or the harmonic characteristic is less than or equal to the second threshold.

[0398] Optionally, the processing unit 501 is specifically used to establish a characteristic coordinate system based on the impact dimension and the harmonic dimension; set a first interval division point in the harmonic dimension, and set a second interval division point in the impact dimension; according to the characteristic coordinate system, the first interval division point, the second interval division point, the impact feature, and the harmonic feature, based on the three-segment sample labeling strategy, the samples in the vibration signal are labeled and stored to obtain a first labeled sample sequence.

[0399] Optionally, the first interval division points include a first division point and a second division point, and the second interval division points include a third division point and a fourth division point;

[0400] The absolute value of the value of the second dividing point is greater than the absolute value of the value of the first dividing point;

[0401] The absolute value of the value of the fourth division point is greater than the absolute value of the value of the third division point.

[0402] Optionally, the three-segment sample labeling strategy includes:

[0403] When the impact characteristic and the harmonic characteristic meet the first condition, marking the sample as normal;

[0404] When the impact characteristic and the harmonic characteristic meet the second condition, marking the sample as abnormal;

[0405] When the impact characteristic and the harmonic characteristic meet the third condition, the sample is marked as pending.

[0406] Optionally, the first condition includes: the impact characteristic is less than or equal to the value of the fourth division point, or the harmonic characteristic is less than or equal to the value of the second division point.

[0407] Optionally, the second condition includes: the impact characteristic is greater than or equal to the value of the fourth division point, and the harmonic characteristic is greater than or equal to the value of the second division point.

[0408] Optionally, the third condition is different from the first condition and the second condition.

[0409] Optionally, the processing unit 501 is specifically used to perform impact feature enhancement on the first labeled sample sequence through a minimum entropy deconvolution method to obtain an enhanced signal sequence; obtain characteristic indicators of the enhanced signal sequence according to the enhanced signal sequence; determine whether the characteristic indicators of the enhanced signal sequence belong to a normal area; if the processing unit 501 determines that the characteristic indicators belong to the normal area, correct the labels of the labeled samples marked as pending in the first labeled sample sequence to normal, and obtain the second labeled sample sequence.

[0410] Optionally, the characteristic indicators of the enhanced signal sequence include the robustness kurtosis and envelope harmonic rate of the enhanced signal sequence.

[0411] Optionally, the normal area includes: the robustness kurtosis of the enhanced signal is greater than or equal to the value of the fourth division point, and the envelope harmonic rate of the enhanced signal is greater than or equal to the value of the second division point.

[0412] Optionally, the processing unit 501 is specifically used to obtain a first envelope spectrum according to the enhanced signal sequence if the processing unit 501 determines that the characteristic indicator does not belong to the normal area; obtain a second envelope spectrum according to the first labeled sample sequence; determine whether the double frequency characteristics of the first envelope spectrum and the second envelope spectrum in the low-frequency domain are consistent; if the processing unit 501 determines that the double frequency characteristics of the first envelope spectrum and the second envelope spectrum in the low-frequency domain are consistent, in response to the user's operation, correct the annotation of the labeled sample marked as pending in the first labeled sample sequence to abnormal, and obtain the second labeled sample sequence; if the processing unit 501 determines that the double frequency characteristics of the first envelope spectrum and the second envelope spectrum in the low-frequency domain are inconsistent, in response to the user's operation, correct the annotation of the labeled sample marked as pending in the second labeled sample sequence to normal, and obtain the second labeled sample sequence.

[0413] Optionally, the processing unit 501 is specifically used to obtain multiple labeled samples and their characteristic indicators within a preset interval from the second labeled sample sequence; obtain a fourth condition and a fifth condition based on the characteristic indicators of the multiple labeled samples and the preset interval; obtain multiple first normal labeled samples from the multiple labeled samples based on the fourth condition; obtain a harmonic rate sequence by extracting the harmonic rate values ​​of the multiple first normal labeled samples; update the second threshold according to the harmonic rate sequence; obtain multiple second normal labeled samples from the multiple labeled samples based on the fifth condition; obtain a third kurtosis sequence by extracting the robust kurtosis values ​​of the multiple second normal labeled samples; and update the first threshold according to the third kurtosis sequence.

[0414] Optionally, the interval division point of the preset interval in the harmonic dimension is a first interval division point.

[0415] Optionally, the interval division point of the preset interval in the impact dimension is a second interval division point.

[0416] Optionally, the target device comprises a rotating mechanical device.

[0417] Optionally, the target device includes a reducer, a motor, a gearbox, a fan, a pump or a turbine rotor.

[0418] Optionally, the acquiring unit 502 is specifically configured to acquire a vibration signal of the target device through a vibration sensor.

[0419] Optionally, the installation position of the vibration sensor includes a driving end of the device, a non-driving end of the device, a top of the device, a bottom of the device, or other parts connected to the device body.

[0420] Optionally, the vibration sensor includes a wired sensor or a wireless sensor.

[0421] Optionally, the sampling frequency of the vibration sensor is greater than or equal to 2000 Hz.

[0422] Optionally, a sampling duration of the vibration sensor is greater than or equal to 0.5 s.

[0423] Optionally, a sampling interval of the vibration sensor is less than or equal to 5 hours.

[0424] It should be understood that the electronic device 500 here is embodied in the form of a functional unit. The term "unit" here can be implemented in the form of software and / or hardware, and is not specifically limited to this. For example, a "unit" can be a software program, a hardware circuit, or a combination of the two that implements the above functions. The hardware circuit may include an application-specific integrated circuit (ASIC), an electronic circuit, a processor (such as a shared processor, a proprietary processor, or a group processor, etc.) and a memory for executing one or more software or firmware programs, a merged logic circuit, and / or other suitable components that support the described functions.

[0425] Therefore, the units of each example described in the embodiments of the present invention can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0426] The embodiment of the present application provides an electronic device, which may be a terminal device or a circuit device built into the terminal device. The electronic device may be used to execute the functions / steps in the above method embodiment.

[0427] An embodiment of the present application provides a computer-readable storage medium, in which instructions are stored. When the instructions are executed on a terminal device, the terminal device executes the functions / steps in the above method embodiment.

[0428] The embodiment of the present application also provides a computer program product comprising instructions, and when the computer program product is run on a computer or any at least one processor, the computer executes the functions / steps in the above method embodiment.

[0429] In the embodiments of the present application, "at least one" refers to one or more, and "plurality" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent the existence of A alone, the existence of A and B at the same time, and the existence of B alone. Among them, A and B can be singular or plural. The character " / " generally indicates that the previous and subsequent associated objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0430] Those of ordinary skill in the art will appreciate that the various units and algorithm steps described in the embodiments disclosed herein can be implemented in a combination of electronic hardware, computer software, and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0431] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0432] In several embodiments provided in the present application, any function can be stored in a computer-readable storage medium if it is implemented in the form of a software functional unit and sold or used as an independent product. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for enabling an electronic device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0433] The above is only a specific implementation of the present application. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. The protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A vibration abnormality discrimination threshold updating method, characterized in that: The method comprises: According to the vibration signal of the target equipment, the impact characteristics and harmonic characteristics are obtained; According to the impact characteristic, the harmonic characteristic, the first threshold and the second threshold, performing abnormality discrimination and alarm discrimination on the samples in the vibration signal; According to the impact characteristics and the harmonic characteristics, samples in the vibration signal are labeled and stored based on a three-segment sample labeling strategy to obtain a first labeled sample sequence; The first labeled sample sequence is labeled and verified by automatic review and manual review to obtain a second labeled sample sequence, and the first threshold and the second threshold are updated according to the second labeled sample sequence.

2. The method according to claim 1, characterized in that Before obtaining the impact characteristic and the harmonic characteristic according to the vibration signal of the target device, the method further includes: A vibration signal of the target device is obtained.

3. The method according to claim 1 or 2, characterized in that: The impact characteristic is robust kurtosis.

4. The method according to any one of claims 1 to 3, characterized in that The harmonic characteristic is the envelope harmonic ratio.

5. The method according to any one of claims 1 to 4, characterized in that According to the vibration signal of the target device, the impact characteristics are obtained, including: Segmenting the vibration signal according to the segment signal length to obtain multiple segment vibration signal sequences; According to the plurality of vibration signal sequences, obtaining a first kurtosis sequence corresponding to the plurality of vibration signal sequences; According to the first kurtosis sequence, obtaining a mean and a variance of the first kurtosis sequence; Determine whether there is a kurtosis value in the first kurtosis sequence whose difference from the mean is more than three times the variance; If it is determined that there is a kurtosis value of the variance whose difference from the mean is more than three times in the first kurtosis sequence, the kurtosis value of the variance whose difference from the mean is more than three times in the first kurtosis sequence is removed to obtain a second kurtosis sequence; According to the second kurtosis sequence, obtaining a mean value of the second kurtosis sequence; The mean value of the second kurtosis sequence is used as the impact feature.

6. The method according to claim 5, characterized in that After determining whether there is a kurtosis value in the first kurtosis sequence whose difference from the mean value exceeds three times of the variance, the method further includes: If it is determined that there is no kurtosis value in the first kurtosis sequence whose difference from the mean exceeds three times of the variance, the mean of the first kurtosis sequence is used as the impact feature.

7. The method according to any one of claims 1 to 6, characterized in that According to the vibration signal of the target equipment, the harmonic characteristics are obtained, including: Obtaining an envelope signal of the vibration signal according to Hilbert transform and the vibration signal; The harmonic characteristic is obtained based on the autocorrelation sequence of the envelope signal.

8. The method according to any one of claims 1 to 7, characterized in that The method of performing abnormality discrimination and alarm discrimination on samples in the vibration signal of the target device according to the impact characteristic, the harmonic characteristic, the first threshold and the second threshold comprises: Determining whether the impulse characteristic is greater than the first threshold and whether the harmonic characteristic is greater than the second threshold; If it is determined that the impact characteristic is greater than the first threshold value and the harmonic characteristic is greater than the second threshold value, the abnormality determination result is abnormal; Starting from the current sample, trace back the first numerical number of samples in the running state to obtain multiple samples; Determining whether the ratio of abnormal samples among the plurality of samples is greater than or equal to a second value; If it is determined that the proportion of abnormal samples among the plurality of samples is greater than or equal to a second value, an alarm is issued; If it is determined that the proportion of abnormal samples among the multiple samples is less than the second value, no alarm is issued.

9. The method according to claim 8, characterized in that After determining whether the impulse characteristic is greater than the first threshold and whether the harmonic characteristic is greater than the second threshold, the method further includes: If it is determined that the impact characteristic is less than or equal to the first threshold, or the harmonic characteristic is less than or equal to the second threshold, the abnormality determination result is normal.

10. The method according to any one of claims 1 to 9, characterized in that The step of labeling and storing samples in the vibration signal based on the three-segment sample labeling strategy according to the impact characteristics and the harmonic characteristics to obtain a first labeled sample sequence includes: Establish a characteristic coordinate system based on the impact dimension and the harmonic dimension; Setting a first interval division point in the harmonic dimension and setting a second interval division point in the impact dimension; According to the characteristic coordinate system, the first interval division point, the second interval division point, the impact characteristic, and the harmonic characteristic, samples in the vibration signal are labeled and stored based on the three-segment sample labeling strategy to obtain a first labeled sample sequence.

11. The method according to claim 10, characterized in that The first interval division points include a first division point and a second division point, and the second interval division points include a third division point and a fourth division point; The absolute value of the value of the second dividing point is greater than the absolute value of the value of the first dividing point; The absolute value of the value of the fourth division point is greater than the absolute value of the value of the third division point.

12. The method according to claim 11, characterized in that The three-stage sample annotation strategy includes: When the impact characteristic and the harmonic characteristic meet the first condition, marking the sample as normal; When the impact characteristic and the harmonic characteristic meet the second condition, marking the sample as abnormal; When the impact characteristic and the harmonic characteristic meet the third condition, the sample is marked as pending.

13. The method according to claim 12, characterized in that The first condition includes: the impact characteristic is less than or equal to the value of the third division point, or the harmonic characteristic is less than or equal to the value of the first division point.

14. The method according to claim 13, characterized in that The second condition includes: the impact characteristic is greater than or equal to the value of the fourth division point, and the harmonic characteristic is greater than or equal to the value of the second division point.

15. The method according to claim 14, characterized in that The third condition is different from the first condition and the second condition.

16. The method according to claim 1, characterized in that The step of performing label verification on the first label sample sequence by automatic review includes: Performing impact feature enhancement on the first labeled sample sequence by a minimum entropy deconvolution method to obtain an enhanced signal sequence; According to the enhanced signal sequence, obtaining a characteristic index of the enhanced signal sequence; Determining whether the characteristic index of the enhanced signal sequence belongs to a normal area; If it is determined that the characteristic index belongs to the normal area, the labels of the labeled samples marked as pending in the first labeled sample sequence are corrected to normal, so as to obtain the second labeled sample sequence.

17. The method according to claim 16, characterized in that The characteristic indicators of the enhanced signal sequence include the robustness kurtosis and envelope harmonic rate of the enhanced signal sequence.

18. The method according to claim 16, characterized in that The normal area includes: the robustness kurtosis of the enhanced signal is less than or equal to the value of the fourth division point, or the envelope harmonic rate of the enhanced signal is less than or equal to the value of the second division point.

19. The method according to claim 16, characterized in that The step of manually verifying the first labeled sample sequence includes: If it is determined that the characteristic indicator does not belong to a normal region, obtaining a first envelope spectrum according to the enhanced signal sequence; Obtaining a second envelope spectrum according to the first labeled sample sequence; Determining whether the frequency-doubled characteristics of the first envelope spectrum and the second envelope spectrum in the low-frequency domain are consistent; If it is determined that the first envelope spectrum and the second envelope spectrum have the same frequency doubling characteristics in the low frequency domain, in response to the user's operation, the annotation of the annotated sample marked as pending in the first annotated sample sequence is corrected to be abnormal, so as to obtain the second annotated sample sequence; If it is determined that the frequency doubling characteristics of the first envelope spectrum and the second envelope spectrum in the low frequency domain are inconsistent, in response to the user's operation, the labeling of the labeled samples marked as pending in the second labeled sample sequence is corrected to normal, so as to obtain the second labeled sample sequence.

20. The method according to claim 1, characterized in that The updating the first threshold and the second threshold according to the second labeled sample sequence includes: Acquire a plurality of annotated samples and their characteristic indicators within a preset interval from the second annotated sample sequence; According to the characteristic indicators of the plurality of labeled samples and the preset interval, a fourth condition and a fifth condition are obtained; According to the fourth condition, obtaining a plurality of first normal labeled samples from the plurality of labeled samples; Obtaining a harmonic rate sequence by extracting the harmonic rate values ​​of the plurality of first normal labeled samples; updating the second threshold value according to the harmonic rate sequence; According to the fifth condition, a plurality of second normal labeled samples are obtained from the plurality of labeled samples; A third kurtosis sequence is obtained by extracting the robust kurtosis values ​​of the plurality of second normal labeled samples; The first threshold is updated according to the third kurtosis sequence.

21. The method according to claim 20, characterized in that The interval division point of the preset interval in the harmonic dimension is the first interval division point.

22. The method according to claim 20, characterized in that The interval division point of the preset interval in the impact dimension is the second interval division point.

23. The method according to claim 1, characterized in that The target equipment includes a reducer, a motor, a gearbox, a fan, a pump or a turbine rotor.

24. The method according to claim 2, characterized in that The obtaining of the vibration signal of the target device includes: A vibration signal of the target device is obtained through a vibration sensor.

25. The method according to claim 24, characterized in that The installation position of the vibration sensor includes the device driving end, the device non-driving end, the device top, the device bottom or other parts connected to the device body.

26. An electronic device, characterized in that: It comprises a processor and a memory, wherein the memory is used to store a computer program, and the computer program comprises program instructions. When the processor runs the program instructions, the electronic device executes the steps of the method as claimed in any one of claims 1 to 25.

27. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program request is executed by a computer, the computer is caused to execute the method according to any one of claims 1 to 25.

Citation Information

Cited By

  • Rolling mill running state monitoring method, system and equipment and storage medium

    CN119832682A