Failure frequency matching of periodic peaks in spectral machine data

Abstract

A computer-implemented method analyzes periodic information in digital vibration data associated with a machine. The method involves generating a spectral periodic information plot (PIP) based on the digital vibration data and locating amplitude peaks in the PIP at frequencies associated with a fundamental frequency of interest. Peaks appearing at the fundamental fault frequency and related harmonic frequencies are removed from the PIP, while retaining the energy values ​​associated with the removed peaks. The remaining peaks in the PIP are classified into synchronous periodic peaks and asynchronous periodic peaks. The remaining peaks in the PIP, along with the fault frequency and related harmonic frequencies, are graphically plotted on a display device using different colors or line styles to identify different frequency groups of interest. The method implements an algorithm that locates peaks in the PIP at frequencies associated with the fundamental frequency of interest, even if the located peak frequencies do not precisely match the fundamental frequency of interest.

Description

Technical Field

[0001] This invention relates to signal analysis. More specifically, this invention relates to a system for separating periodic amplitude peaks from non-periodic amplitude peaks in machine vibration data. Background Technology

[0002] Fault frequencies are frequency values ​​associated with peak amplitudes in machine vibration spectrum data, which indicates the mechanical or electrical characteristics of the machine. Each bearing in the machine has a set of associated fault frequencies. Each set of bearing fault frequencies includes the fundamental frequency and its harmonics. When a machine vibration analyst observes the lines or other markings corresponding to a specific fault frequency superimposed on the vibration spectrum, the analyst can visually determine whether the spectral peaks in the spectrum are consistent with the pattern of that specific fault frequency. Figure 2 An example of this situation is shown, which describes a spectrum display that includes markers associated with a set of bearing outer race (BPFO) fault frequencies overlaid on the spectrum indicating BPFO faults. The markers are vertical dashed lines evenly distributed across the spectrum.

[0003] like Figure 2 As shown, the fundamental frequency fault mark (lowest frequency mark) and the fundamental frequency spectrum peak appear to match. However, as the harmonic peak frequency increases, the harmonic peaks separate from the other marks. Even if the fault frequency mark does not precisely match the actual spectrum frequency, an experienced analyst reviewing this data will know that the bearing (BPFO) is faulty.

[0004] The analysis becomes more difficult if an algorithm (rather than an experienced analyst) is determining whether a particular fault frequency marker (fundamental and harmonic) matches a specific measured spectral peak. Matching a fault frequency marker to a measured spectral peak depends on how close the marker's frequency should be to the corresponding frequency of the spectral peak. Figure 3 It shows Figure 2 A magnified view around the peak of the maximum amplitude spectrum, even though the fundamental fault frequency marker looks similar to... Figure 2 The maximum amplitude peak is consistent with that in the frequency values, which are actually as follows: Figure 3 As shown, this difference becomes even larger for higher frequency harmonic values ​​of the fundamental frequency. Therefore, for automated analysis, matching analysis is initially performed only for fundamental frequency faults.

[0005] Therefore, an improved process is needed to match the peak values ​​of the vibration spectrum associated with the harmonics of the fault frequency with the expected (labeled) frequencies of those harmonics. Summary of the Invention

[0006] The process described herein can be applied to any vibration spectrum. However, if the matching process is applied to a periodic information graph (PIP) as described in U.S. Patent 10,416,126, the search for harmonic peaks at the fault frequency is enhanced, making the matching process more efficient. In the PIP, the location of harmonic peaks is easier to identify because most noise peaks have been eliminated, thus providing fewer matching options.

[0007] In the matching process described herein, frequencies of interest (FOIs) need to be isolated from each other for analysis. As used herein, a “frequency of interest” consists of the fundamental frequency peak and harmonic frequency peaks, if applicable. Each FOI is either synchronously cyclic (i.e., operating speed, gear meshing) or asynchronously cyclic (i.e., bearing failure, belt frequency, oscillating teeth). A typical feature of PIPs is that each FOI is plotted with a different color or line style than the others, making it easier for analysts to distinguish the different FOIs displayed in the graph.

[0008] It should be understood that each FOI can be considered a form of fault frequency. It is crucial to distinguish between shaft speed frequencies and bearing fault frequencies that may be multiples of the shaft speed. Fault frequencies associated with synchronous periodic signals can be calculated based on speed detection, such as using a speed algorithm, or based on an input speed value associated with the expected shaft speed. Fault frequencies associated with asynchronous periodic signals can be calculated based on the associated speed and processed differently, as described in more detail below. Asynchronous periodic fault frequencies, such as bearing fault frequencies, can be found using the same methods used to locate shaft speeds and other synchronous periodic frequencies. In some preferred embodiments described herein, the bearing fault frequency is treated as a synchronous periodic frequency in order to find the associated bearing fault frequency peak and energy. Also in some preferred embodiments, the bearing fault frequency energy is included in the calculation of the total asynchronous periodic energy, rather than the synchronous periodic energy. This total asynchronous periodic energy is incorporated into the calculation of the severity of the mechanical fault (bearing fault) as described in patent 10,416,126.

[0009] Figure 4 A PIP (Picture Injection Processing) depicts the advantages of a non-gearbox configuration with unknown bearing characteristics. The PIP preferably shows the synchronous cycle peaks related to driving speed, plotted as dashed lines, and the asynchronous peaks, shown as solid lines. For example... Figure 5 The gearbox shown here has a synchronization cycle peak that includes all shaft speeds and associated oscillating gear frequencies (plotted as dashed lines). Figure 5 As shown, asynchronous peaks can be drawn with solid lines.

[0010] Bearing failure frequency values ​​can be applied to the asynchronous cycle peaks of the PIP to determine precise bearing failure frequencies, such as the inner ring ball pass frequency (BPFI), outer ring ball pass frequency (BPFO), ball spin frequency (BSF), and fundamental training frequency (FTF - also known as cage frequency). Figure 6 The PIP shows BPFO faults (thick dashed peaks), which are determined by applying known bearing fault frequencies. Figure 4 The results shown indicate that, similarly, by applying the bearing failure frequency... Figure 5 The gearbox analysis shown in the figure is... Figure 7 The PIP shows BPFO faults (thick solid peaks). Each of these frequency groups (if present in the spectrum) can be drawn in the PIP with a different color or line style. The process for matching fault frequencies to PIP peaks is described in more detail below.

[0011] Another problem that makes fault frequency pattern detection more difficult involves resolution and the ability to distinguish the harmonics of the operating speed from the fundamental frequency of the fault. The fundamental frequency of a fault, such as a bearing fault frequency, is often very close to the fundamental harmonic of the operating speed. This is Figure 8 The bearing analysis shown here has a BPFI frequency that is 7.054 times the travel speed, while the BPFO frequency is 4.946 times the travel speed. With insufficient resolution, it is difficult to distinguish between a travel speed 7 times the BPFI frequency and a travel speed 7.054 times the BPFI frequency.

[0012] In some preferred embodiments described herein, the algorithm for finding the FOI fault frequency first determines the speed of the associated shaft, and then finds all fundamental FOIs (in this example, bearing fault frequencies) in the PIP that match the peaks. Preferably, all harmonics of the found fundamental bearing fault frequencies are located and removed from the PIP before the residual peaks in the PIP are characterized as speed-related and / or gearbox-related. Figure 9 The text describes separation and Figure 8 The results of the relevant bearing failure frequency algorithm should be noted. Figure 8 The maximum peak value in is now characterized as follows: Figure 9 The fundamental frequency of BPFI as described in [the text] Figure 9 The bearing information in the lower left corner indicates the fundamental frequency of the fault associated with a specific fault, and adjusts the energy of the peak values ​​for each FOI group accordingly. The detailed description below provides information on how to locate fault frequencies and generate [the necessary parameters]. Figure 9 Further details of the process shown in the graph.

[0013] A preferred embodiment relates to an apparatus for acquiring and analyzing periodic information in machine vibration. The apparatus includes a vibration sensor, a data collector, and a processor. The vibration sensor is securely attached to a location within the machine, providing a reliable transmission path from a vibration source within the machine to the vibration sensor. The data collector includes an analog-to-digital converter that receives vibration signals from the sensor and converts the vibration signals into digital vibration data. The data collector also includes a memory for buffering the digital vibration data.

[0014] The processor executes operation instructions to process digital vibration data, including the following instructions:

[0015] Determine the rotational speed of the rotating parts of the machine;

[0016] Vibration waveforms are generated based on this digital vibration data;

[0017] A spectrum periodic information diagram is generated based on this digital vibration data;

[0018] Locate the amplitude peak in the periodic information graph at the frequency associated with the fundamental frequency of interest, which is associated with the fundamental frequency of the fault frequency family of the peak.

[0019] Remove the amplitude peaks at the fundamental fault frequency and the relevant harmonic frequency from the periodic information graph;

[0020] The remaining amplitude peaks in the periodic information graph are classified into synchronous periodic peaks and asynchronous periodic peaks; and

[0021] The peak values ​​of the fault frequency family and the peak values ​​of the remaining amplitude in the periodic information graph are graphically plotted in different colors or with different line styles on the display device to identify different groups of frequencies of interest.

[0022] In some embodiments, the processor retains the asynchronous energy value associated with the removed amplitude peak, which is associated with the asynchronous period peak, and displays the asynchronous period energy value as a period infographic on a display device.

[0023] In some embodiments, the family of peak fault frequencies is associated with bearing faults in the machine, and the processor calculates the severity of the bearing fault based at least in part on asynchronous cyclic energy values.

[0024] In some embodiments, the processor locates the amplitude peak in the periodicity graph at a frequency associated with the fundamental frequency of interest, even if the frequency of the amplitude peak does not precisely match the fundamental frequency of interest.

[0025] In some embodiments, the processor locates each amplitude peak in the period information graph at a frequency associated with the base frequency of interest by the following steps:

[0026] Determine the frequency difference between the frequency of the peak amplitude and the frequency associated with the corresponding fundamental frequency of interest;

[0027] If the frequency difference is less than a predetermined difference, then the peak amplitude is designated as a candidate to match the corresponding fundamental frequency of interest;

[0028] Calculate the percentage energy value of the peak amplitude;

[0029] If the percentage energy value of the amplitude peak is greater than a predetermined percentage of the total energy of the periodic infographic, then the amplitude peak is designated as a candidate to match the corresponding fundamental frequency of interest; and

[0030] The candidate amplitude peak with the frequency closest to the fundamental frequency of interest is selected as the fundamental fault frequency peak.

[0031] In some embodiments, the processor analyzes the amplitude peak in a periodicity graph associated with the baseband of interest to determine the type of machine fault associated with the amplitude peak, and displays information indicating the type of machine fault, along with the periodicity graph, on a display device.

[0032] In some embodiments, the processor displays information indicating the type of machine fault, which is selected from the group consisting of the inner ring ball pass frequency (BPFI), the outer ring ball pass frequency (BPFO), the ball rotation frequency (BSF), and the basic sequence frequency (FTF).

[0033] In some embodiments, the apparatus includes a data communication network, a processor connected to the data communication network, and a periodic information graph communicating via the data communication network. The apparatus also includes an analysis computer that receives the periodic information graph via the data communication network and displays the periodic information graph for analysis by an analyst.

[0034] In some embodiments, the data collector includes a digital data logger or a vibration data collector.

[0035] In some embodiments, the processor is a component of the data collector, while in other embodiments, the processor is a component of an analysis computer that communicates with the data collector via a communication network.

[0036] In another aspect, some embodiments of the present invention provide a computer-implemented method for acquiring and analyzing periodic information in machine vibrations. In a preferred embodiment, the method includes:

[0037] (a) Use a vibration sensor that is securely attached to the machine to generate a vibration signal;

[0038] (b) Convert the vibration signal into digital vibration data;

[0039] (c) Determine the rotational speed of the rotating parts of the machine;

[0040] (d) Generating vibration waveforms based on digital vibration data;

[0041] (e) Generate a spectral period information map based on digital vibration data;

[0042] (f) Locate the amplitude peak in the periodic information plot at the frequency associated with the fundamental frequency of interest, which is associated with the fundamental frequency of the fault frequency family of the peak.

[0043] (g) Remove the amplitude peaks at the fundamental fault frequency and the relevant harmonic frequencies from the period information graph;

[0044] (h) Divide the remaining amplitude peaks in the periodic information graph into synchronous periodic peaks and asynchronous periodic peaks; and

[0045] (i) Graphically plot the peak values ​​of the fault frequency family and the peak values ​​of the remaining amplitude in the periodic information graph in different colors or with different line styles on a display device to identify different groups of frequencies of interest.

[0046] In some embodiments, the method further includes:

[0047] (j) Retain the asynchronous energy value, which is associated with the removed peak amplitude;

[0048] (k) Calculate the asynchronous periodic energy value, which is associated with the asynchronous periodic peak value; and

[0049] (l) Display the asynchronous periodic energy value on a display device using a periodic information graph.

[0050] In some embodiments, the family of peak failure frequencies is associated with bearing failures in the machine, and the method further includes calculating bearing failure severity based at least in part on asynchronous cyclic energy values.

[0051] In some embodiments, step (f) includes locating the amplitude peak in the period information graph at a frequency associated with the fundamental frequency of interest, even if the amplitude peak in the period information graph does not precisely match the fundamental frequency of interest. In some such embodiments, for each amplitude peak, step (f) includes:

[0052] Determine the frequency difference between the frequency of the peak amplitude and the frequency associated with the corresponding fundamental frequency of interest;

[0053] If the frequency difference is less than a predetermined difference, then the peak amplitude is designated as a candidate to match the corresponding fundamental frequency of interest;

[0054] Calculate the percentage energy value of the peak amplitude;

[0055] If the percentage energy value of the amplitude peak is greater than a predetermined percentage of the total energy of the periodic infographic, then the amplitude peak is designated as a candidate to match the corresponding fundamental frequency of interest; and

[0056] Select the candidate amplitude peak value of the frequency closest to the fundamental frequency of interest as the fundamental fault frequency peak value.

[0057] In some embodiments, the method includes:

[0058] (j) Analyze the amplitude peak in the periodicity graph associated with the fundamental frequency of interest to determine the type of machine fault associated with that amplitude peak; and

[0059] (k) The periodic infographic on the display device is used to display information indicating the type of machine malfunction.

[0060] In some embodiments, step (k) includes information indicating the type of machine failure, which is one or more of the following: inner race ball pass frequency (BPFI), outer race ball pass frequency (BPFO), ball rotation frequency (BSF), or fundamental sequence frequency (FTF).

[0061] The following are definitions of various terms used in describing some embodiments of the present invention.

[0062] Fault frequency: A frequency or family of frequencies associated with one or more amplitude peaks in vibration spectrum data, wherein one or more peaks are associated with mechanically related faults in the machine;

[0063] Fault frequencies (m): A list of m fault frequencies in the vibration spectrum data;

[0064] Bearing failure frequencies (m): A list of m failure frequencies in the vibration spectrum data associated with bearing-related failures;

[0065] The vibrational energy of the bearing FF frequency (Frequency Frequency) is defined as the peak amplitude of the vibration.

[0066] Frequency of Interest (FOI): The frequency of failures associated with failures in the components of interest within a machine;

[0067] Asynchronous cyclic FOI: The frequency of concern associated with the asynchronous cyclic peak that indicates an asynchronous cyclic failure (such as a bearing or belt failure);

[0068] Asynchronous periodic energy: Vibrational energy associated with asynchronous periodic peaks in the vibration spectrum;

[0069] Synchronization Period FOI: The frequency of concern associated with synchronization period failures, such as the frequency associated with machine operating speed or gear meshing frequency;

[0070] Synchronous periodic energy: Vibrational energy associated with the synchronous periodic peak in the vibration spectrum;

[0071] Synchronization (m): A list of m fault frequencies associated with the peak value of the synchronization cycle, such as those associated with machine operating speed or gear meshing frequency;

[0072] Total synchronization: Analyzes the energy of all synchronization peaks. When calculating a single synchronization value, such as speed or a single fault frequency, total synchronization is the energy of the base value and its harmonics. For azimuth analysis, there are four fault frequency "harmonic families," and total synchronization is the total energy of all four fault frequency families.

[0073] dfMult: Optional incremental frequency multiplier;

[0074] df: The frequency difference between two adjacent discrete cells in the FFT;

[0075] Periodic Information Plot (PIP): A graph of amplitude versus frequency, showing the periodic frequencies derived from the waveform. This plot is generated by performing autocorrelation on the waveform and then performing an FFT on the resulting autocorrelated waveform, along with percentage periodic energy prediction. Only the maximum amplitude peak from the autocorrelation FFT is retained. The energy of the maximum amplitude peak comprises a percentage of the periodic energy of the total energy of the spectrum. The equivalent peaks in the original FFT match the maximum autocorrelation FFT peaks in frequency. These resulting peaks are included in the periodic information plot (as described in U.S. Patent 11,002,641).

[0076] PIP Positioning Peaks (j): A list of j peaks located in the Periodic Information Graph (PIP), where each positioning peak has a positioning frequency and a positioning amplitude;

[0077] PIP Energy: Total energy associated with the peak value in the periodicity infographic;

[0078] Asynchronous PIP: Removes PIPs that spike in the synchronization cycle. As synchronization spikes are detected, they are removed sequentially. Attached Figure Description

[0079] Other embodiments of the invention will become apparent from the detailed description taken in conjunction with the accompanying drawings, in which elements are not drawn to scale in order to show details more clearly, wherein the same reference numerals indicate the same elements in several views, and wherein:

[0080] Figure 1A and 1B A functional block diagram of a system for deriving and analyzing periodic information in a signal is described according to some preferred embodiments of the present invention.

[0081] Figure 2The spectrum display is shown, which includes markers associated with a set of bearing outer ring (BPFO) failure frequencies overlaid on the spectrum indicating BPFO failure;

[0082] Figure 3 It shows Figure 2 A magnified view of the display shows the location marker of the fundamental fault frequency next to the maximum peak in the spectrum;

[0083] Figure 4 The advantages of non-gearboxes with unknown bearing characteristics are described (from Figure 2 The Peak-Derived Periodic Information Plot (PIP) shows the synchronous periodic peaks related to running speed, drawn with dashed lines, and the asynchronous periodic peaks, drawn with solid lines.

[0084] Figure 5 The PIP of the gearbox is depicted, where the synchronous cycle peak includes all shaft speeds and associated oscillating tooth frequencies, and the asynchronous cycle peak is drawn as a solid line.

[0085] Figure 6 Depicting Figure 4 The PIP for non-transmission advantages shown in the figure indicates the outer ring defects (BPFO failure frequency) drawn with thick dashed lines;

[0086] Figure 7 It shows Figure 5 The PIP of the transmission shown represents the outer ring defect (BPFO failure frequency) drawn with a thick solid line, and BPFO failure is analyzed and confirmed in it;

[0087] Figure 8 A PIP is described, which has asynchronous periodic peaks drawn in dashed lines and numerous harmonics of rotational speed drawn in solid lines.

[0088] Figure 9 The PIP (plotted in solid lines) of a bearing with many harmonics at rotational speed is shown, and the analysis indicates inner ring bearing failure (BPFI), with the associated peaks plotted in thick dashed lines.

[0089] Figure 10 A process for determining bearing failure frequency according to a preferred embodiment is shown;

[0090] Figure 11 The process for finding the fundamental frequency of interest (FOI) according to a preferred embodiment is described; and

[0091] Figure 12 The process for separating the FOI peak from other peaks according to a preferred embodiment is described. Detailed Implementation

[0092] Figure 1A and1B An exemplary system 10 for deriving and analyzing periodic information in vibration signals is described. Figure 1A In one embodiment, a sensor 14, such as an accelerometer, is attached to the machine 12 to monitor the machine's vibrations. Although in Figure 1A An accelerometer is described in the exemplary embodiments described, but it should be understood that other types of sensors, such as velocity sensors, displacement probes, ultrasonic sensors, or pressure sensors, can be used. Sensor 14 generates a vibration signal (or other types of signals for sensors other than accelerometers) containing periodic information. For repeatability and best results, it is preferable to position each sensor 14 such that there is a physical path transitioning from the signal source (e.g., a bearing) to the sensor's mounting location. The sensor 14 should also be mounted to ensure that the signal is sensed with the least possible distortion. Some preferred embodiments include one or more tachometers 16 for measuring the rotational speed of one or more rotating parts of machine 12. The vibration and tachometer signals are provided to a data collector 16, which preferably includes an analog-to-digital converter (ADC) 18 for sampling the vibration and tachometer signals, an optional low-pass anti-aliasing filter 20 (or other combinations of low-pass and high-pass filters), and a buffer memory 22. For example, the data collector 16 may be a digital data logger, a handheld vibration data collector, or a permanently or temporarily installed monitoring device. Vibration signal data is transmitted to the periodic information processor 114, which performs the information processing tasks described herein. Figure 1A In this embodiment, the cycle information processor 24 is a component of the data collector 16. In this embodiment, the cycle information processor 24 transmits processed data via a machine data network 26, which may be a HART network. TM or WirelessHART TM A network, Ethernet, or the Internet. The analysis computer 28 receives processed data via network 26 for display on display device 30.

[0093] exist Figure 1B In the alternative embodiment shown, the periodic information processor 24 is a component of the analysis computer 28. This embodiment may be preferred when data transmission and storage are not primary concerns, so that the entire dataset can be transmitted via network 26 to the analysis computer 28 or other remote processing devices for post-processing using the same algorithms and techniques.

[0094] Regarding the placement of the sensor used for bearing and gear diagnostics, sensor 14 is typically mounted orthogonally to the shaft. Preferably, sensor 14 is mounted on a rigid and heavy metal object near the signal source (i.e., the bearing or gear). Conversely, the large amount of metal on which the sensor is mounted helps prevent resonance from entering the signal due to the machine's surface. Sensor 104 should be mounted to minimize loss of signal integrity during transmission. This requires a rigid connection—typically, sensor 104 is mounted by bolts. In some cases, such as when the machine's mounting surface is rough or covered with multiple layers of paint, the surface will need to be sanded.

[0095] Figure 10 This illustrates data based on vibration spectrum data, for example, using... Figure 1A and 1B A preferred embodiment of the process 100 for determining bearing failure frequencies using data collected in an exemplary system 10 shown herein. A speed detection algorithm is executed to determine the machine's operating speed (step 102). An exemplary speed detection algorithm is described in U.S. Patent 11,009,520, the entire contents of which are incorporated herein by reference. Alternatively, the machine's operating speed can be measured directly using a tachometer 16. Initially, the values ​​in the list of failure frequencies (1…M) are set to be equal to the values ​​of the bearing failure frequencies (1…M), where M is 4 (step 104). The values ​​of the bearing failure frequencies (1…M) are typically calculated based on bearing size and components and may be provided by the bearing manufacturer. In some preferred embodiments, these values ​​are provided in an orientation information file accessed by the analysis software.

[0096] In step 106, using Figure 11 The process described in the document finds the fundamental frequency of interest (FOI), as described in more detail below.

[0097] The initial value of synchronization (1…M) is set equal to the fault frequency value (1…M), where M is 4 (step 108), and the FF analysis is set to true (step 110). In step 112, using… Figure 12 The process described in [the document] performs an FOI analysis, as described in more detail below. The bearing FF energy value is then set to be equal to [the value of the bearing FF energy]. Figure 12 The total synchronization value is determined by the process.

[0098] If vibration data to be analyzed has been collected for the transmission, the value FF analysis is set to false (step 116), and the transmission analysis is performed using the process described in U.S. Patents 10,416,126 and 11,002,641 (which describes determining speed using gear meshing frequencies), the entire contents of which are incorporated herein by reference (step 118). For the transmission, this process involves finding all shaft speeds and associated oscillating tooth frequencies, and designating each speed and oscillating tooth frequency as FOI, as done when finding the fault frequency. The process then proceeds to step 126, where the asynchronous cyclic energy is calculated as...

[0099]

[0100] This total asynchronous cyclic energy is incorporated into the severity of computational mechanical failures (bearing failures) as described in patents 10,416,126.

[0101] If no vibration data of the gearbox was collected for analysis, the value FF for analysis was set to FALSE (step 120), the value of synchronization (1) was set to the basic operating speed (step 122), and the value of synchronization (1) was used. Figure 12 The process described herein performs FOI analysis, as detailed below (step 124). The process then proceeds to step 126, where the asynchronous periodic energy is calculated.

[0102] Figure 11 Describes the method used to find in Figure 10 The preferred embodiment of the fundamental frequency of interest (FOI) process 200 referenced in step 106, the process being... Figure 10 The fault frequency list (1…M) set in step 104 begins with the following initial values ​​(step 202):

[0103] m=1

[0104] dfMult = 2.5

[0105] df = Fmax / LOR

[0106] Diff = dfmult × df

[0107] TempFF = Fault Frequency (m)

[0108] Fault Frequency(m) = 0.0

[0109] In step 204, initial values ​​for the PIP positioning peak (j = 1…J) and PIP energy are provided, and the integer j is set to 1. The initial values ​​for the PIP positioning peak (j = 1…J) and PIP energy are determined according to the methods described in U.S. Patent Nos. 10,416,126 and 11,002,641.

[0110] If |TempFF-PIP location peak frequency (j)| is less than or equal to Diff (step 206), the process proceeds to step 208. If |TempFF-PIP location peak frequency (j)| is not less than or equal to Diff (step 206), and the integer j is not equal to J (step 210), then j is incremented by 1, and the process loops back to step 206. In step 210, if j equals J, and the integer m equals M (step 212), all fault frequencies of interest have been searched, and the process is complete. In step 212, if m is not equal to M, then m is incremented by 1, TempFF is set to equal to the fault frequency (m), Diff is set to equal to dfMult×df, the fault frequency (m) is set to equal to zero, and J is set to equal to zero (step 214). Then the integer j is incremented by 1, and the process loops back to step 206.

[0111] In step 208, if [(PIP location peak amplitude(j))] 2 / (PIP Energy) 2 If the value of [(PIP location peak amplitude(j))] is not greater than 0.05, then the process proceeds to step 210 as described above. In step 208, if [(PIP location peak amplitude(j))] 2 / (PIP Energy) 2 If the value is greater than 0.05, then the value of the fault frequency (m) is set to the value of the PIP positioning peak frequency (j), and the Diff is set to be equal to |TempFF-PIP positioning peak frequency (j)| (step 216).

[0112] Figure 12 A preferred embodiment of a process 300 for separating the Frequency of Interest (FOI) peak from other peaks is described, such as... Figure 10As shown in steps 112 and 124, the process begins by setting the initial value of total synchronization to zero, setting the integer m to 1, and setting the attention frequency to be equal to synchronization(m) (step 302). If m equals 1 (step 304 - first loop iteration), a periodic information graph (PIP) is calculated, a peak list is generated, and the total PIP energy is determined according to the process described in U.S. Patent Nos. 10,416,126 and 11,002,641, the entire contents of which are incorporated herein by reference (step 306). In step 304, if m is not equal to 1, an asynchronous PIP is generated (step 308). In step 310, the periodic peak values ​​for synchronization and asynchrony are calculated based on the attention frequency, and the synchronous periodic energy and asynchronous periodic energy are calculated.

[0113] If the FF analysis is not true (step 312), the process proceeds to step 316. If the FF analysis is true (step 312), the ratio value is calculated according to the following formula:

[0114] Ratio value = (Synchronization cycle energy) 2 / (PIP Energy) 2 ,

[0115] If the ratio value is less than 0.05, the synchronization cycle energy is set to zero (step 314), and processing continues in step 316.

[0116] In step 316, if m is not equal to M, then m is incremented by 1 (step 318) and the synchronization energy (m-1) is set to equal the synchronization cycle energy (step 320). In step 322, the total synchronization is calculated according to the following formula:

[0117]

[0118] In step 324, if the process is not searching for a fault frequency (if not an FF analysis) or the synchronization energy (m-1) is greater than zero, an asynchronous PIP is generated by removing the synchronization peak associated with synchronization (m-1) and the frequency of interest is set to synchronization (m). The process then loops back to step 304.

[0119] In step 316, if m equals M, then the synchronization energy (M) is set to equal the synchronization period energy, and the total synchronization is calculated according to the following formula:

[0120]

[0121] If the process does not find the fault frequency (if not FF analysis) or the synchronization energy (m) is greater than zero, an asynchronous PIP is generated by removing the synchronization peaks associated with the synchronization (m) (step 326). The PIP is then displayed (e.g., on display device 30), where each synchronization (j) and associated harmonics, as well as the periodic peaks of the asynchronous, are drawn in different colors or line styles according to the harmonic group (step 328).

[0122] For purposes of illustration and description, the above description of some preferred embodiments of the invention has been given. These are not exhaustive, nor are they intended to limit the invention to the precise forms disclosed. Obvious modifications or variations are possible in accordance with the above teachings. Some embodiments were chosen and described to provide the best illustration of the principles of the invention and its practical application, thereby enabling those skilled in the art to utilize the invention in various embodiments and make various modifications suitable for the particular purpose contemplated. All such modifications and variations are within the scope of the invention as defined by the appended claims when interpreted according to the scope of a fair, lawful, and just authorization.

Claims

1. An apparatus for acquiring and analyzing periodic information in vibrations associated with a machine, the apparatus comprising: A vibration sensor is securely attached to the machine at a location that provides a reliable transmission path from a vibration source within the machine to the vibration sensor, the vibration sensor being used to generate vibration signals; A data collector, which communicates with the vibration sensor and is configured to receive and modulate the vibration signal, includes: Analog-to-digital converter, the analog-to-digital converter being used to convert vibration signals into digital vibration data; and A memory, the memory being used to buffer the digital vibration data; and A processor operable to receive the digital vibration data, the processor being configured to execute operational instructions for processing the digital vibration data, the operational instructions including instructions to perform the following operations when executed: Determine the rotational speed of the rotating parts of the machine; A vibration waveform is generated based on the digital vibration data; A spectral period information diagram is generated based on the digital vibration data; Locate the amplitude peak in the periodicity graph at a frequency associated with the fundamental frequency of interest, which is associated with the fundamental frequency of the fault frequency family of the peak. Remove the amplitude peaks at the fundamental fault frequency and associated harmonic frequency from the periodic information graph; The asynchronous energy value is retained, and the asynchronous energy value is associated with the peak amplitude of the removed value; The remaining amplitude peaks in the periodic information graph are classified into synchronous periodic peaks and asynchronous periodic peaks. Calculate the asynchronous periodic energy value, which is associated with the asynchronous periodic peak value; Fault frequency families and residual amplitude peaks in the periodic information graph are graphically plotted on a display device using different colors or line styles to identify different groups of frequencies of interest, and the asynchronous periodic energy value is displayed on the display device using the periodic information graph.

2. The apparatus of claim 1, wherein, The family of fault frequencies of the peak value is associated with bearing faults in the machine, and the processor calculates the severity of the bearing faults at least in part based on the asynchronous periodic energy value.

3. The apparatus of claim 1, wherein, The processor locates the amplitude peak in the periodic information graph at a frequency associated with the fundamental frequency of interest, even if the frequency of the amplitude peak does not precisely match the fundamental frequency of interest.

4. The apparatus of claim 3, wherein, The amplitude peaks in the period information plot located at the frequencies associated with the fundamental frequency of interest include, for each amplitude peak: Determine the frequency difference between the frequency of the amplitude peak and the frequency associated with the corresponding fundamental frequency of interest; If the frequency difference is less than a predetermined difference, the amplitude peak value is designated as a candidate to match the corresponding fundamental frequency of interest; Calculate the percentage energy value of the peak amplitude; If the percentage energy value of the amplitude peak is greater than a predetermined percentage of the total energy of the periodic information graph, then the amplitude peak is designated as a candidate to match the corresponding fundamental frequency of interest. as well as The candidate amplitude peak value of the frequency closest to the fundamental frequency of interest is selected as the fundamental fault frequency peak value.

5. The apparatus of claim 1, wherein, The processor: Analyze and compare the amplitude peaks in the periodic information graph associated with the fundamental frequency of interest to determine the type of machine fault associated with the amplitude peaks; and The periodic information graph is used on the display device to display information indicating the type of machine malfunction.

6. The apparatus of claim 5, wherein, The processor displays information indicating the type of machine malfunction, which is selected from a group consisting of the inner ring ball pass frequency (BPFI), the outer ring ball pass frequency (BPFO), the ball spin frequency (BSF), and the basic training frequency (FTF).

7. The apparatus according to claim 1, further comprising: A data communication network, wherein the processor is connected to the data communication network and transmits the periodic information graph through the data communication network; as well as An analysis computer, connected to the data communication network, is used to receive and display the periodic information graph for analysis.

8. The apparatus of claim 1, wherein, The data collector includes a digital data logger or a vibration data collector.

9. The apparatus of claim 1, wherein, The processor is a component of the data collector.

10. The apparatus of claim 1, wherein, The processor is a component of an analysis computer that communicates with the data collector via a communication network.

11. A computer-implemented method for acquiring and analyzing periodic information in vibrations associated with a machine, the method comprising: (a) A vibration signal is generated using a vibration sensor securely attached to the machine; (b) Convert the vibration signal into digital vibration data; (c) Determine the rotational speed of the rotating parts of the machine; (d) Generate a vibration waveform based on the digital vibration data; (e) Generate a spectral period information diagram based on the digital vibration data; (f) Locate the amplitude peak in the periodicity graph at a frequency associated with the fundamental frequency of interest, which is associated with the fundamental frequency of the fault frequency family of the peak. (g) Remove the amplitude peaks at the fundamental fault frequency and the associated harmonic frequencies from the periodic information graph; (h) Divide the remaining amplitude peaks in the periodic information graph into synchronous periodic peaks and asynchronous periodic peaks; as well as (i) Graphically plot the peak fault frequency peaks and residual amplitude peaks in the periodic information graph in different colors or line styles on a display device to identify different groups of frequencies of interest; (j) Retain the asynchronous energy value, which is associated with the removed peak amplitude; (k) Calculate the asynchronous periodic energy value, which is associated with the asynchronous periodic peak value; as well as (l) The asynchronous periodic energy value is displayed on the display device using the periodic information graph.

12. The method of claim 11, wherein, The family of failure frequencies of the peak value is associated with bearing failures in the machine, and the method further includes calculating the severity of bearing failures based at least in part on the asynchronous periodic energy value.

13. The method of claim 11, wherein, Step (f) further includes positioning the amplitude peak in the period information graph at a frequency associated with the fundamental frequency of interest, even if the amplitude peak in the period information graph does not precisely match the fundamental frequency of interest.

14. The method of claim 13, wherein, Step (f) includes, for each amplitude peak: Determine the frequency difference between the frequency of the amplitude peak and the frequency associated with the corresponding fundamental frequency of interest; If the frequency difference is less than a predetermined difference, the amplitude peak value is designated as a candidate to match the corresponding fundamental frequency of interest; Calculate the percentage energy value of the peak amplitude; If the percentage energy value of the amplitude peak is greater than a predetermined percentage of the total energy of the periodic information graph, then the amplitude peak is designated as a candidate to match the corresponding fundamental frequency of interest. as well as The candidate amplitude peak value of the frequency closest to the fundamental frequency of interest is selected as the fundamental fault frequency peak value.

15. The method of claim 11, further comprising: (j) Analyze the amplitude peaks in the periodic information graph associated with the fundamental frequency of interest to determine the type of machine fault associated with the amplitude peaks; as well as (k) Display information indicating the type of machine malfunction on the display device using the periodic information graph.

16. The method of claim 15, wherein, Step (k) also includes displaying information indicating the type of machine failure, which is selected from a group consisting of the inner ring ball pass frequency (BPFI), the outer ring ball pass frequency (BPFO), the ball rotation frequency (BSF), and the basic training frequency (FTF).

Images