Fault detection method and device for rotary mechanical equipment

Abstract

The invention relates to the technical field of rotary mechanical vibration noise detection, in particular to a fault detection method and device for rotary mechanical equipment. The fault detection method comprises the following steps: carrying out vibration signal collection on offline to-be-detected equipment; constructing a three-dimensional modulation order spectrum of modulation order, carrier detection and energy through two times of Fourier transform; according to the three-dimensional modulation order spectrogram, constructing a two-dimensional modulation order spectrogram of aggregation energy in a modulation order and carrier order interval; and performing fault diagnosis on the to-be-tested equipment according to the two-dimensional modulation order spectrogram and the preset threshold value of each modulation order. According to the invention, a double-Fourier three-dimensional spectrum technology is provided, a'modulation order-carrier order-energy 'three-dimensional relation is constructed, the modulation order range is positioned, and the single-dimension limitation of a traditional order spectrum is broken through; and the fault recognition rate is obviously improved.

Description

Technical Field

[0001] This application relates to the field of vibration and noise detection technology for rotating machinery, and specifically to a fault detection method and apparatus for rotating machinery equipment. Background Technology

[0002] As a core component of new energy vehicles, the NVH performance of the electric drive system directly determines the user's driving experience, and modulation characteristics are one of the main reasons for NVH complaints about electric drive systems. Modulation characteristics originate from faults such as bearing failure, shaft eccentricity, and gear meshing errors, manifesting as vibration characteristics where a high-frequency carrier signal is modulated by a low-frequency fault characteristic signal. The energy distribution pattern of these characteristics is a key basis for fault diagnosis. Accurate assessment and effective control of modulation characteristics during off-line NVH testing of electric drive systems can significantly reduce the overall vehicle NVH complaint rate and improve user satisfaction.

[0003] In related technologies, vibration characteristic assessment generally employs order spectrum analysis. However, traditional order spectrum analysis only displays the energy of each order in isolation. The order spectrum is essentially a single-dimensional energy distribution map (order-energy), with "order" as the x-axis and "energy value" as the y-axis, showing the energy strength of different order components. It fails to establish a two-dimensional correlation between "low-frequency fault signals" (modulation order: reflecting the difference frequency relationship between the fault characteristic frequency and the carrier frequency) and "high-frequency structural signals" (carrier order: the ratio of the high-frequency carrier frequency to the fundamental frequency). In complex scenarios, electric drive assemblies may exhibit multi-source modulation superposition (such as the coexistence of shaft eccentricity and bearing fault modulation). The characteristic order range and energy distribution of different modulation sources differ. Because the order spectrum lacks analysis of the two-dimensional "modulation-carrier" relationship, it cannot distinguish the energy distribution patterns corresponding to different modulation sources, leading to confusion between multi-scale modulation characteristics. Therefore, how to solve the problem of insufficient modulation characteristic assessment in the NVH testing of rotating machinery equipment has become an urgent problem for practitioners. Summary of the Invention

[0004] In related technologies, the lack of three-dimensional analysis in the fault diagnosis of rotating machinery equipment leads to low accuracy of diagnostic results and a lack of clear mechanistic support, making it difficult to trace the root cause of the fault.

[0005] In a first aspect, embodiments of this application provide a fault detection method for rotating machinery, the fault detection method comprising: Vibration signals are collected from the offline test equipment; A three-dimensional modulation order spectrum of modulation order, carrier detection, and energy is constructed by two Fourier transforms. A two-dimensional modulation order spectrum is constructed based on the three-dimensional modulation order spectrum, which shows the aggregated energy within the modulation order and carrier order intervals. Fault diagnosis of the device under test is performed based on the two-dimensional modulation order spectrum and the preset thresholds for each modulation order.

[0006] In conjunction with the first aspect, in one embodiment, constructing a two-dimensional modulation order spectrum of aggregated energy within the modulation order and carrier order intervals based on a three-dimensional modulation order spectrum includes: The modulation order under the working state is determined based on the design parameters of the device under test and is taken as the working modulation order. The key carrier order interval is selected based on the main distribution area of ​​modulation features in the three-dimensional modulation order spectrum of the device under test. Calculate the aggregated energy within the key carrier order interval for each operating modulation order, and generate a two-dimensional modulation order spectrum based on the aggregated energy and the operating modulation order.

[0007] In conjunction with the first aspect, in one implementation, calculating the aggregated energy within the key carrier order interval of each operating modulation order includes: Calculate the maximum and / or sum of the energy within the key carrier order interval for each working modulation order.

[0008] In conjunction with the first aspect, in one implementation, the construction of the three-dimensional modulation order spectrum of modulation order, carrier detection, and energy through two Fourier transforms includes: The carrier order is extracted based on the vibration signal through the first Fourier transform. The modulation order is separated by a second Fourier transform of the extracted carrier order. A three-dimensional modulation order spectrum is plotted based on carrier order, modulation order, and energy.

[0009] In conjunction with the first aspect, in one embodiment, the step of extracting the carrier order based on the vibration signal through a first Fourier transform includes: Extract the carrier order according to the formula. :

[0010] in, Let B be the starting point of time, and B be the length of a single data block. Add a window function to the time domain. For carrier components, For the vibration signal at time The sampled values ​​are denoted as e, k, and j, where e is the natural constant, k is the dummy variable, and j is the imaginary unit.

[0011] In conjunction with the first aspect, in one implementation, the step of performing a second Fourier transform to separate the output modulation order using the extracted carrier order includes: Separate modulation order according to formula :

[0012] Where T is the total signal duration and a is the overlap rate. For carrier components.

[0013] In conjunction with the first aspect, in one implementation, before performing fault diagnosis on the device under test based on the two-dimensional modulation order spectrum and preset thresholds for each modulation order, the following steps are included: Set initial preset thresholds based on historical project data; It collects data from each test as new historical data to generate a database, and uses the historical data in the database to dynamically update the preset thresholds through a self-learning mechanism.

[0014] In conjunction with the first aspect, in one implementation, the step of dynamically updating the preset threshold using historical data from the database through a self-learning mechanism includes: According to the formula, the preset threshold is... Update:

[0015] Where Average is the mean of historical data. The standard deviation of historical data, This is the experience compensation value.

[0016] In conjunction with the first aspect, in one implementation, the step of dynamically updating the preset threshold using historical data from the database through a self-learning mechanism further includes: adjusting the experience compensation value based on the bench interception rate and the overall vehicle complaint rate. Make dynamic adjustments.

[0017] Secondly, embodiments of this application provide a fault detection device for rotating machinery, the fault detection device comprising: The vibration signal acquisition unit is used to collect vibration signals from the offline test equipment; The three-dimensional modulation order spectrum generation unit is used to construct a three-dimensional modulation order spectrum of modulation order, carrier detection, and energy through two Fourier transforms. A two-dimensional modulation order spectrum generation unit is used to construct a two-dimensional modulation order spectrum of aggregated energy within the modulation order and carrier order intervals based on a three-dimensional modulation order spectrum. The fault diagnosis unit is used to diagnose faults in the device under test based on the two-dimensional modulation order spectrum and the preset thresholds for each modulation order.

[0018] The beneficial effects of the technical solutions provided in this application include: This application proposes a dual Fourier three-dimensional spectral technique to construct a three-dimensional relationship of "modulation order - carrier order - energy", thereby locating the modulation order range and breaking through the single-dimensional limitation of traditional order spectra; the fault identification rate is significantly improved. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating an embodiment of the fault detection method of this application; Figure 2 This is a spectrum of operating conditions in one embodiment of this application; Figure 3 This is a three-dimensional modulation order spectrum diagram of an embodiment of this application; Figure 4 This is a two-dimensional working condition envelope order spectrum diagram in one embodiment of this application; Figure 5 This is a schematic diagram of the hardware structure of the fault detection device involved in the embodiments of this application. Detailed Implementation

[0020] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0021] In related technologies, the lack of three-dimensional analysis in the fault diagnosis of rotating machinery equipment leads to low accuracy of diagnostic results and a lack of clear mechanistic support, making it difficult to trace the root cause of the fault.

[0022] Specifically, traditional order spectral analysis has the following drawbacks: First, simply displaying the energy of each order in isolation makes it impossible to distinguish whether the energy comes from the carrier itself or from the modulation. The essence of modulation characteristics is the coupling effect between the "low-frequency fault signal (modulation order)" and the "high-frequency structural signal (carrier order)," while the order spectrum is essentially a single-dimensional energy distribution spectrum (order-energy), with "order" as the horizontal axis and "energy value" as the vertical axis, showing the energy strength of different order components, without establishing a two-dimensional relationship between the "low-frequency fault signal (modulation order)" and the "high-frequency structural signal (carrier order)."

[0023] Second, order spectra cannot pinpoint the modulation source (such as specific vibration sources like shaft or bearing faults), making it difficult to separate fault characteristics from interference signals. In complex scenarios, electric drive assemblies may exhibit multi-source modulation superposition (e.g., the coexistence of shaft eccentricity and bearing fault modulation). The characteristic order ranges and energy distributions of different modulation sources vary. Because order spectra lack the ability to analyze the two-dimensional relationship between modulation and carrier, they cannot distinguish the energy distribution patterns corresponding to different modulation sources, leading to confusion between multi-scale modulation features. For example, the high-frequency carrier of gear meshing may be modulated by the low-frequency characteristics of bearing faults, while simultaneously superimposed with modulation interference from shaft imbalance. Traditional order spectra only present a single spectral line after energy superposition, failing to separate the independent characteristics of each modulation source and affecting the accuracy of fault location.

[0024] Third, the physical meaning of modulation characteristics is directly related to the fault mechanism. Because the order spectrum does not resolve the two-dimensional relationship between "modulation-carrier", it is impossible to establish a mapping relationship between energy distribution and fault type. For example, the modulation caused by the passing frequency of bearing rollers should be concentrated in the "specific modulation order + carrier order interval", but the traditional order spectrum only shows isolated order energy and cannot correlate the two, resulting in a lack of clear mechanistic support for the diagnostic results and making it difficult to trace the root cause of the fault.

[0025] In a first aspect, embodiments of this application provide a fault detection method for rotating machinery, the fault detection method comprising: Step S1: Collect vibration signals from the offline test equipment.

[0026] Step S1 above includes: Step S1a: Based on the main driving conditions of the whole vehicle, considering the masking of wind noise and road noise at high speeds, the complaints about modulation characteristics are concentrated in the low-speed constant speed condition (with the least background noise). The offline NVH test conditions are specifically set to simulate high-frequency complaint scenarios, and representative conditions are selected in combination with the offline test cycle requirements.

[0027] The definitions of the operating conditions in some specific embodiments can be found in the table below: Table 1. Reference Operating Conditions for NVH Evaluation of Modulation-Related Issues

[0028] Step S1b: Data acquisition parameter settings.

[0029] Specifically, this includes setting vibration measurement points and setting the sampling rate.

[0030] Optionally, vibration measurement points are set: vibration sensors are placed at key locations on the electric drive assembly housing (such as suspension points, bearing housing points, etc.) to ensure coverage of major vibration sources. The sampling rate is generally set to 100 kHz to meet the Nyquist sampling requirements of high-frequency carrier components (such as the operating frequencies of motors and gears) and avoid aliasing.

[0031] Step S2: Construct a three-dimensional modulation order spectrum of modulation order, carrier detection, and energy through two Fourier transforms.

[0032] It should be noted that carrier order refers to the ratio of the high-frequency carrier frequency to the fundamental frequency. Modulation order refers to the frequency difference between the fault characteristic frequency and the carrier frequency.

[0033] Specifically, carrier order Represented as:

[0034] in, , n is the motor's base frequency (n is the motor speed, in rpm). This refers to the carrier frequency (such as the electromagnetic force frequency of an electric motor, the meshing frequency of gears, etc.).

[0035] Modulation order To represent the frequency difference relationship between the fault characteristic frequency and the carrier frequency, the relationship between the maximum evaluable modulation order and the analysis parameters is expressed as follows:

[0036] Where B is the block length (affecting frequency resolution), u is the revolutions per block increment (rev / block, used to correlate rotational speed with order mapping), and a is the overlap rate (affecting temporal resolution).

[0037] The above step S2 specifically includes: Step S2a: Preprocess the vibration signal obtained in step S1.

[0038] Specifically, step S2a includes: anti-aliasing filtering and order tracking; wherein, anti-aliasing filtering: cutoff frequency kHz, satisfying kHz sampling rate. Order tracking: Based on resampling the rotational speed pulse signal to the angle domain, the influence of rotational speed fluctuations is eliminated.

[0039] Step S2b: Extract the carrier order based on the vibration signal through the first Fourier transform.

[0040] Specifically, the carrier order is extracted according to the formula. :

[0041] in, Let B be the starting point of time, and B be the length of a single data block. Add a window function to the time domain. For carrier components, For the vibration signal at time The sampled values ​​are denoted as e, k, and j, where e is the natural constant, k is the dummy variable, and j is the imaginary unit.

[0042] Step S2c: Perform a second Fourier transform on the extracted carrier order to separate and output the modulation order.

[0043] Specifically, the modulation order is separated according to the formula. :

[0044] Where T is the total signal duration and a is the overlap rate. Let be the carrier component, and e be the natural constant.

[0045] Step S2d: Draw a three-dimensional modulation order spectrum based on the carrier order, modulation order, and energy.

[0046] Specifically, in terms of modulation order The horizontal axis represents the carrier order. The vertical axis represents energy. To differentiate the color intensity, a three-dimensional modulation order spectrum is generated, achieving a three-dimensional representation of "modulation order - carrier order - energy". The color gradient (blue → green → yellow → red) in the graph represents energy from low to high. The right-hand scale indicates the carrier order value, and the bottom scale indicates the modulation order value.

[0047] Step S3: Construct a two-dimensional modulation order spectrum of aggregated energy within the modulation order and carrier order intervals based on the three-dimensional modulation order spectrum.

[0048] Step S3 above includes: Step S3a: Select the key carrier order interval based on the working modulation order of the device under test and the main distribution area of ​​the modulation features in the three-dimensional modulation order spectrum.

[0049] It is worth noting that various interferences frequently occur in the testing environment, leading to abnormal spectrum displays and misdiagnosis of faults. Therefore, first using a three-dimensional modulation order spectrum to determine the key modulation order intervals, and then using the key carrier order intervals to generate subsequent two-dimensional modulation order spectra, can reduce interference and make the judgment results more accurate.

[0050] Specifically, step S3a includes: Step 1: Based on the design parameters of the rotating machinery, determine its typical operating modulation order. The operating modulation order includes the electromagnetic force order of the motor, the gear meshing order, and the characteristic order of the bearings and shafts.

[0051] Step 2: Based on the main distribution areas of modulation features in the three-dimensional modulation order spectrum, select the key carrier order intervals of interest [o c,low ,o c,high ].

[0052] It is understandable that high-energy regions in a three-dimensional modulation order spectrum (such as...) can be observed. Figure 3 Identify the key carrier order intervals within the red area.

[0053] Step S3b: Calculate the aggregated energy within the key carrier order interval of each working modulation order, and generate a two-dimensional modulation order spectrum based on the aggregated energy and the working modulation order.

[0054] Specifically, step S3b above includes: Step 1, Energy Aggregation Calculation: Calculate the maximum and / or sum of the energy values ​​within the key carrier order interval for each operating modulation order: Understandably, the Max mode is used to highlight the strongest fault characteristics, while the Sum mode is used to supplement distributed faults.

[0055] Max mode:

[0056] Sum mode for calculating summation:

[0057] Here, low and high correspond to the lower and upper limits of the carrier order interval, respectively.

[0058] Step 2: Generation of two-dimensional modulation order spectrum.

[0059] Specifically, in terms of modulation order The horizontal axis represents the aggregated energy value. and / or Using the vertical axis as the ordinate, a two-dimensional spectrum is plotted to achieve quantitative characterization of modulation features. Different colors in the spectrum represent different sample data.

[0060] Step S4: Perform fault diagnosis on the device under test based on the two-dimensional modulation order spectrum and the preset thresholds for each modulation order.

[0061] It should be noted that steps S1-S3 above are preliminary data processing and spectrum plotting for a single device under test (DUT). After completion, a threshold is used to judge the spectrum of the DUT. By checking whether there are parameters in the spectrum that exceed the threshold, a fault can be identified, and the fault location can be found by looking up a table. In actual work, multiple DUTs will be continuously taken offline for fault detection.

[0062] In some preferred embodiments, the selection of the preset threshold needs to be accomplished by switching between a fixed limit mode and a self-learning limit mode based on the collected sample data. Specifically, step S4 above includes: Step S4a: Set the initial preset threshold based on historical project data.

[0063] It is worth noting that this step adopts a fixed limit mode: the threshold is directly set based on historical project data (e.g., the ModS limit is 2.5 m / s under constant speed conditions). 2 ).

[0064] It's worth noting that while the fixed limit mode offers the advantage of rapid interception, it cannot account for system drift caused by factors such as equipment aging, sensor bias, and environmental noise fluctuations. Therefore, the fixed limit mode should be used when the initial sample size is small.

[0065] Step S4b: Collect the data from each test as a new historical data database, and use the historical data in the database to dynamically update the preset threshold through a self-learning mechanism.

[0066] Understandably, step S4b, using the self-learning limit mode, has the advantage of considering system drift, but the disadvantage is that it cannot be applied when there are few samples in the early stages. Therefore, it is used in the middle and later stages of the entire diagnostic process when there are more samples in the database.

[0067] The self-learning limit mode in step S4b specifically includes: According to the formula, the preset threshold is... Update:

[0068] Where Average is the mean of historical data. The standard deviation of historical data, This is the experience compensation value.

[0069] In some preferred embodiments, a closed-loop optimization mechanism is introduced for the self-learning limit mode in step S4b, that is, the empirical compensation value is adjusted by combining the bench interception rate and the vehicle complaint rate. Make dynamic adjustments.

[0070] Specifically, the threshold selection for the self-learning limit mode can be dynamically updated and optimized by referring to the table below: Table 2 Threshold Closed-Loop Optimization Mechanism

[0071] It is worth noting that traditional fault detection methods have the following drawbacks in determining thresholds: Firstly, thresholds for order spectra are often based on the energy of a single order (such as the energy upper limit of a certain carrier order), without considering the "carrier-modulation" correlation of modulation features, leading to a disconnect between the threshold and the essence of the fault (e.g., the energy difference of the same modulation feature under different carrier orders may be misjudged). Secondly, fixed thresholds cannot cope with system drift such as equipment aging, sensor deviation, and environmental noise fluctuations, while traditional self-learning thresholds rely on the accumulation of a large number of samples, and the window parameters (such as the sample size) are difficult to balance "dynamic response" and "result stability," easily leading to misjudgments (thresholds too loose) or missed detections (thresholds too strict).

[0072] Furthermore, the preferred threshold selection method described above in this application designs a closed-loop dynamic threshold mechanism, namely a dynamic threshold control strategy of dual-mode switching and closed-loop optimization, which takes into account both "rapid interception in the early stage" and "adaptive optimization in the later stage", and establishes a bench-vehicle closed-loop feedback mechanism to take into account both fault interception rate and accuracy, thereby achieving closed-loop optimization; the control efficiency is significantly optimized, and the dynamic control mechanism reduces the threshold stabilization period from the initial 6 months to 3 months.

[0073] This application provides a specific embodiment of fault detection based on an electric drive assembly, which includes the following steps: Step S1, Operating Condition Execution: Based on the actual driving and riding conditions and feedback of the electric drive assembly in the vehicle, set the NVH evaluation conditions for tuning-related issues as shown in Table 3.

[0074] Table 3. Modulation-related issues of the electric drive assembly during offline NVH testing.

[0075] Step S2: Setting the working modulation order parameters Specifically, considering that typical modulation characteristics are bearing and shaft order, the modulation order range is set to [0, 8]. Related parameter settings are as follows: block length B = 8192 (balancing resolution and computational load), rotational speed increment u = 16 (corresponding to rotational speed and order mapping), step factor a = 32 (improving time-frequency continuity).

[0076] Step S3: Select the carrier order interval.

[0077] Specifically, it covers the range of electromagnetic force order of motors, gear meshing order, bearing and shaft characteristic order, with a focus on the main distribution range of modulation order for this type of product, with a reference setting of 50-150 orders.

[0078] Step S4: Selecting the threshold.

[0079] Specifically, this includes selecting a preset threshold through dual-mode switching.

[0080] Scenario 1: When the sample size is less than 200 units, a fixed limit mode is adopted, that is, a fixed limit is set with reference to existing project data.

[0081] Table 4 Reference for Setting Fixed Limits

[0082] Case 2: When the sample size is ≥200 units, the self-learning limit mode is adopted.

[0083] The threshold of the self-learning limit for:

[0084] Furthermore, the self-learning constraint mode is optimized in a closed loop: based on the bench interception rate and the vehicle complaint rate, the corresponding thresholds are optimized.

[0085] Table 5 Threshold Closed-Loop Optimization Reference

[0086] Step S4: Construct a three-dimensional modulation order spectrum of modulation order, carrier detection, and energy through two Fourier transforms according to the aforementioned method, and further prepare a two-dimensional modulation order spectrum.

[0087] This application provides a specific test case condition for the device under test, which includes: a project vehicle complained that there was a significant “humming” noise at a constant speed of 30 kph, which was confirmed by objective data analysis to be the 4.08th order modulation characteristic of the electric drive assembly, corresponding to the 3000 rpm constant speed drive condition in the off-line working condition.

[0088] At this point, the order spectrum corresponding to the constant speed condition during the reverse inspection of the electric drive assembly's offline test shows no outliers (e.g., ...). Figure 2 (The faulty electric drive unit is blue), so it cannot be effectively intercepted.

[0089] Generate the modulation order spectrum and modulation order spectral diagram according to step S4 above. Based on the modulation order spectral diagram (e.g. Figure 3 Further analysis of the modulation characteristics revealed that the energy of modulation order 4.08 is concentrated in the carrier order range of 50-250. Considering the significant interference in the carrier order range of 150-200, the carrier order range of 50-150 was selected, and the modulation order spectrum was obtained by summing these values ​​(e.g., Figure 4 The system can clearly distinguish between faulty and normal samples. The faulty sample shows a significant peak at order 4.08, exceeding the limit of 2.5 m / s. 2 .

[0090] Step S5: Based on the typical order table of the electric drive assembly (as shown in Table 6), it is confirmed that the modulation characteristic order 4.08 is consistent with the second harmonic of the rollers of bearings A / B. Therefore, it is determined that bearings A / B may have a loss. The faulty electric drive assembly is disassembled and bearings A / B are inspected. It is found that there are obvious scratches on the rollers of bearing A. Thus, the fault detection is completed and the location of the faulty component is found.

[0091] Secondly, this application provides a fault detection device for rotating machinery, the fault detection device comprising: a vibration signal acquisition unit, a three-dimensional modulation order spectrum generation unit, a two-dimensional modulation order spectrum generation unit, and a fault diagnosis unit. The system includes a vibration signal acquisition unit for collecting vibration signals from the offline device under test; a three-dimensional modulation order spectrum generation unit for constructing a three-dimensional modulation order spectrum of modulation order, carrier detection, and energy through two Fourier transforms; a two-dimensional modulation order spectrum generation unit for constructing a two-dimensional modulation order spectrum of aggregated energy within the modulation order and carrier order intervals based on the three-dimensional modulation order spectrum; and a fault diagnosis unit for diagnosing faults in the device under test based on the two-dimensional modulation order spectrum and preset thresholds for each modulation order.

[0092] In one embodiment, reference is made to Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the fault detection method of this application. Figure 1 As shown, the fault detection methods include: Step S1: Collect vibration signals from the offline test equipment; Step S2: Construct a three-dimensional modulation order spectrum of modulation order, carrier detection, and energy through two Fourier transforms; Step S3: Construct a two-dimensional modulation order spectrum of aggregated energy within the modulation order and carrier order intervals based on the three-dimensional modulation order spectrum; Step S4: Perform fault diagnosis on the device under test based on the two-dimensional modulation order spectrum and the preset thresholds for each modulation order.

[0093] Further, in one embodiment, the step of constructing a two-dimensional modulation order spectrum of aggregated energy within the modulation order and carrier order intervals based on the three-dimensional modulation order spectrum includes: Select key carrier order intervals based on the main distribution areas of modulation features in the working modulation order and three-dimensional modulation order spectrum of the device under test; calculate the aggregated energy in the key carrier order interval of each working modulation order, and generate a two-dimensional modulation order spectrum based on the aggregated energy and the working modulation order.

[0094] Furthermore, in one embodiment, calculating the aggregated energy within the key carrier order interval of each operating modulation order includes: calculating the maximum value and / or summation value of the energy within the key carrier order interval of each operating modulation order.

[0095] Furthermore, in one embodiment, the construction of the three-dimensional modulation order spectrum of modulation order, carrier detection, and energy through two Fourier transforms includes: The carrier order is extracted based on the vibration signal through the first Fourier transform. The modulation order is separated by a second Fourier transform of the extracted carrier order. A three-dimensional modulation order spectrum is plotted based on carrier order, modulation order, and energy.

[0096] Further, in one embodiment, the extraction of the carrier order based on the vibration signal through a first Fourier transform includes: Extract the carrier order according to the formula. :

[0097] in, Let B be the starting point of time, and B be the length of a single data block. Add a window function to the time domain. For carrier components, For the vibration signal at time The sampled values ​​are denoted as e, k, and j, where e is the natural constant, k is the dummy variable, and j is the imaginary unit.

[0098] Further, in one embodiment, the step of performing a second Fourier transform on the extracted carrier order to separate the output modulation order includes: According to the formula:

[0099] Separate modulation order In the formula, T is the total signal duration, and a is the overlap rate. For carrier components.

[0100] Furthermore, in one embodiment, before performing fault diagnosis on the device under test based on the two-dimensional modulation order spectrum and the preset thresholds for each modulation order, the process includes: setting an initial preset threshold based on historical project data; collecting each test data as a new historical data database; and using the historical data in the database to dynamically update the preset thresholds through a self-learning mechanism.

[0101] Furthermore, in one embodiment, the step of dynamically updating the preset threshold using historical data in the database through a self-learning mechanism includes: According to the formula:

[0102] For preset threshold To update, in the formula, Average is the mean of the historical data. The standard deviation of historical data, This is the experience compensation value.

[0103] Furthermore, in one embodiment, the step of dynamically updating the preset threshold using historical data in the database through a self-learning mechanism further includes: combining the benchtop interception rate and the overall vehicle complaint rate to adjust the experience compensation value. Make dynamic adjustments.

[0104] The functions of each module in the above-mentioned fault detection device correspond to the steps in the above-mentioned fault detection method embodiments, and their functions and implementation processes will not be described in detail here.

[0105] Thirdly, embodiments of this application provide a fault detection device, which can be a personal computer (PC), laptop computer, server, or other device with data processing capabilities.

[0106] Reference Figure 5 , Figure 5 This is a schematic diagram of the hardware structure of the fault detection device involved in the embodiments of this application. In the embodiments of this application, the fault detection device may include a processor, a memory, a communication interface, and a communication bus.

[0107] The communication bus can be of any type and is used to interconnect the processor, memory, and communication interface.

[0108] Communication interfaces include input / output (I / O) interfaces, physical interfaces, and logical interfaces used for interconnecting internal components of the fault detection equipment, as well as interfaces used for interconnecting the fault detection equipment with other devices (such as other computing devices or user equipment). Physical interfaces can be Ethernet interfaces, fiber optic interfaces, ATM interfaces, etc.; user equipment can be displays, keyboards, etc.

[0109] Memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0110] The processor can be a general-purpose processor, which can call a fault detection program stored in memory and execute the fault detection method provided in the embodiments of this application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed when the fault detection program is called can be referred to in the various embodiments of the fault detection method of this application, and will not be repeated here.

[0111] Those skilled in the art will understand that Figure 5 The hardware structure shown does not constitute a limitation of this application and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0112] Fourthly, embodiments of this application also provide a computer-readable storage medium.

[0113] The present application has a computer-readable storage medium storing a fault detection program, wherein when the fault detection program is executed by a processor, it implements the steps of the fault detection method described above.

[0114] The method implemented when the fault detection program is executed can be referred to in various embodiments of the fault detection method of this application, and will not be repeated here.

[0115] In summary, this application proposes a dual Fourier three-dimensional spectral technique to construct a three-dimensional relationship of "modulation order - carrier order - energy," thereby locating the modulation order range and overcoming the limitations of the single dimension of traditional order spectra. The fault identification rate is significantly improved, with the false negative rate decreasing from >0.02% to <0.002% compared to traditional methods. Furthermore, this application designs a closed-loop dynamic threshold mechanism, namely a dynamic threshold control strategy involving dual-mode switching and closed-loop optimization, balancing "rapid early-stage interception" with "adaptive optimization in the later stages." A bench-vehicle closed-loop feedback mechanism is also established to balance fault interception rate and accuracy, achieving closed-loop optimization. Control efficiency is significantly improved, with the dynamic control mechanism reducing the threshold stabilization period from the initial 6 months to 3 months.

[0116] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0117] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0118] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0119] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0120] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0121] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.

[0122] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A fault detection method for rotating machinery, characterized in that, The fault detection method includes: Vibration signals are collected from the offline test equipment; A three-dimensional modulation order spectrum of modulation order, carrier detection, and energy is constructed by two Fourier transforms. A two-dimensional modulation order spectrum is constructed based on the three-dimensional modulation order spectrum, which shows the aggregated energy within the modulation order and carrier order intervals. Fault diagnosis of the device under test is performed based on the two-dimensional modulation order spectrum and the preset thresholds for each modulation order.

2. The fault detection method as described in claim 1, characterized in that, The construction of a two-dimensional modulation order spectrum based on the three-dimensional modulation order spectrum, which aggregates energy within the modulation order and carrier order intervals, includes: The modulation order under the working state is determined based on the design parameters of the device under test and is taken as the working modulation order. Select the key carrier order intervals based on the main distribution areas of modulation features in the working modulation order and the three-dimensional modulation order spectrum. Calculate the aggregated energy within the key carrier order interval for each operating modulation order, and generate a two-dimensional modulation order spectrum based on the aggregated energy and the operating modulation order.

3. The fault detection method as described in claim 2, characterized in that, The calculation of the aggregated energy within the key carrier order interval for each working modulation order includes: Calculate the maximum and / or sum of the energy within the key carrier order interval for each working modulation order.

4. The fault detection method as described in claim 1, characterized in that, The construction of the three-dimensional modulation order spectrum, including modulation order, carrier detection, and energy, through two Fourier transforms includes: The carrier order is extracted based on the vibration signal through the first Fourier transform. The modulation order is separated by a second Fourier transform of the extracted carrier order. A three-dimensional modulation order spectrum is plotted based on carrier order, modulation order, and energy.

5. The fault detection method as described in claim 4, characterized in that, The step of extracting the carrier order based on the vibration signal through the first Fourier transform includes: According to the formula: Extracting carrier order ,in, Let B be the starting point of time, and B be the length of a single data block. Apply a windowing function to the time domain. For carrier components, For the vibration signal at time The sampled values ​​are denoted as e, k, and j, where e is the natural constant, k is the dummy variable, and j is the imaginary unit.

6. The fault detection method as described in claim 5, characterized in that, The step of performing a second Fourier transform to separate the output modulation order using the extracted carrier order includes: According to the formula: Separate modulation order Where T is the total signal duration and a is the overlap rate. For carrier components.

7. The fault detection method as described in claim 1, characterized in that, Before performing fault diagnosis on the device under test based on the two-dimensional modulation order spectrum and preset thresholds for each modulation order, the following steps are included: Set initial preset thresholds based on historical project data; It collects data from each test as new historical data to generate a database, and uses the historical data in the database to dynamically update the preset thresholds through a self-learning mechanism.

8. The fault detection method as described in claim 7, characterized in that, The method of dynamically updating the preset threshold using historical data in the database through a self-learning mechanism includes: According to the formula: For preset threshold The update is performed, where Average is the average of the historical data. The standard deviation of historical data, This is the experience compensation value.

9. The fault detection method as described in claim 8, characterized in that, The method of dynamically updating the preset threshold using historical data in the database through a self-learning mechanism also includes: combining the bench interception rate and the overall vehicle complaint rate to adjust the experience compensation value. Make dynamic adjustments.

10. A fault detection device for rotating machinery, characterized in that, The fault detection device includes: The vibration signal acquisition unit is used to collect vibration signals from the offline test equipment; The three-dimensional modulation order spectrum generation unit is used to construct a three-dimensional modulation order spectrum of modulation order, carrier detection, and energy through two Fourier transforms. A two-dimensional modulation order spectrum generation unit is used to construct a two-dimensional modulation order spectrum of aggregated energy within the modulation order and carrier order intervals based on a three-dimensional modulation order spectrum. The fault diagnosis unit is used to diagnose faults in the device under test based on the two-dimensional modulation order spectrum and the preset thresholds for each modulation order.

Images