Automatic detection of bearing defects by scanning pattern and post-scan logic filter

Abstract

A bearing defect automatic detection system includes a processor to receive condition monitoring data including vibration harmonics corresponding to a bearing coupled to a rotatable shaft. The processor performs a pattern scan process that scans a pattern through both a speed range and a bearing class defect frequency range. In response to a test pattern having a test pattern sideband, the processor scans the test pattern sideband through a sideband range for the condition monitoring data to determine a baseband frequency and a sideband frequency of the pattern. The processor determines a most likely bearing defect type associated with the bearing based on a best match value in the results associated with the test pattern. The processor also performs a post-scan logic process that compares a number (N) of most recent results from the pattern scan process to at least one conditional test to confirm that the most likely bearing defect type is present.

Description

Technical Field

[0001] This invention relates to an automatic bearing defect detection method using a swept pattern and a post-sweep logic filter. Background Technology

[0002] A train can have many railcars, each with multiple axles and corresponding axle boxes. Each axle box can house bearings from the same or different manufacturers. Over time, bearings develop defects due to various reasons (e.g., contamination, surface defects, lubrication problems, etc.), which can be detected in the bearing's vibration harmonics. The area of ​​collecting and monitoring these vibration harmonics, and addressing the defects detected in them, is called condition monitoring.

[0003] Furthermore, regardless of whether traditional condition monitoring applications are online or offline, installing, utilizing, and maintaining shaft speed sensors that support the collection and monitoring of bearing vibration harmonics can be problematic and / or expensive. For example, traditional condition monitoring applications require knowledge of the shaft speed within a few percent of the tolerance and the exact bearing defect (from the bearing's brand and designation) in order to identify the vibration frequency components / symptoms associated with the bearing defect within the bearing vibration harmonics.

[0004] Furthermore, condition monitoring applications rely on known or pre-modeled bearing defect frequencies corresponding to specific bearing brands / models. These defect frequencies can be predetermined by the bearing manufacturer and cataloged for public use. Therefore, the identified vibration spectrum frequencies can be compared with the predetermined bearing defect frequencies (corresponding to the brand / model of the bearing mounted on the axle housing) to ultimately identify one or more bearing defects. However, even knowing the approximate shaft speed, the actual brand / model of the bearing mounted on a given axle housing may differ from the expected brand / model of the bearing specified for mounting on the axle housing, potentially leading to misdiagnosis or inaccurate identification of bearing defects.

[0005] Many traditional condition monitoring applications also require managing specific parameters that can affect shaft speed calculations (such as the wheel diameter and bearing designation for each axle box). The database organizing these parameters must be constantly and promptly updated to ensure proper management of these specific parameters. However, the database update process can be costly (or time-consuming) relative to man-hours and is also prone to errors, such as incorrect data indicating the brand / model of the bearing installed on a given axle box. Summary of the Invention

[0006] According to one or more non-limiting embodiments, a method for performing automatic bearing defect detection is provided. The method includes receiving condition monitoring data from one or more sensors by a processor, the condition monitoring data including vibration harmonics of at least one bearing coupled to a rotatable shaft. The method further includes performing a pattern scanning process by the processor, the pattern scanning process scanning at least one test pattern through both a speed range and a bearing category defect frequency range. In response to the test pattern having at least one test pattern sideband, the at least one test pattern sideband is also scanned through the sideband range against the condition monitoring data to determine the baseband frequency (fundamental frequency) and sideband frequency of the pattern from best-match values. The method further includes determining, by the processor, the most probable bearing defect type associated with the at least one bearing based on the best-match value among two or more results associated with the at least one test pattern. The method further includes performing a post-sweep logic process by the processor, the post-sweep logic process comparing a number (N) of the latest results from the pattern scanning process with at least one condition test to confirm the presence of the most probable bearing defect type.

[0007] According to one or more non-limiting embodiments, an automatic bearing defect detection system is provided herein. The automatic bearing defect detection system includes a processor that communicates with one or more sensors to receive condition monitoring data, the condition monitoring data including vibration harmonics corresponding to at least one bearing coupled to a rotatable shaft. The processor is configured to perform a pattern scanning process that scans at least one test pattern through both a speed range and a bearing category defect frequency range. In response to the test pattern having at least one test pattern sideband, the processor also scans the at least one test pattern sideband through the sideband range for the condition monitoring data to determine the baseband frequency and sideband frequency of the pattern from best-match values. The processor is further configured to determine the most probable bearing defect type associated with the at least one bearing based on the best-match value among two or more results associated with the at least one test pattern. The processor also performs post-scan logic processing that compares a number (N) of the latest results from the pattern scanning process with at least one condition test to confirm the presence of the most probable bearing defect type.

[0008] According to one or more non-limiting embodiments, a computer program product for controlling an electronic hardware processor to perform automatic bearing defect detection includes a computer-readable storage medium having program instructions embodied therein, the program instructions being executable by the processor to perform operations including receiving condition monitoring data, the condition monitoring data including vibration harmonics of at least one bearing coupled to a rotatable shaft. The operations also include performing a pattern scanning process that scans at least one test pattern through both a speed range and a bearing category defect frequency range. In response to the test pattern having at least one test pattern sideband, the at least one test pattern sideband is also scanned through the sideband range with respect to the condition monitoring data to determine the baseband frequency and sideband frequency of the pattern from best-match values. The operation further includes determining the most probable bearing defect type associated with the at least one bearing based on the best-match value among two or more results associated with the at least one test pattern. The operation also includes performing post-scan logic processing that compares a number (N) of the latest results from the pattern scanning process with at least one conditional test to confirm the presence of the most probable bearing defect type.

[0009] Further features and advantages are achieved through the technology disclosed herein. Other embodiments and aspects of this disclosure are described in detail herein. For a better understanding of the advantages and features of this disclosure, please refer to the specification and accompanying drawings. Attached Figure Description

[0010] This subject matter is specifically pointed out and explicitly claimed in the claims. The foregoing and other features and advantages of the embodiments herein will become apparent from the following detailed description taken in conjunction with the accompanying drawings:

[0011] Figure 1 A system according to one or more embodiments is described;

[0012] Figure 2A and Figure 2B A processing flow according to one or more embodiments is described;

[0013] Figure 3 A graph depicting the scanning process associated with the baseband pattern (fundamental pattern) according to one or more embodiments is shown.

[0014] Figure 4 A graph depicting the sideband pattern scanning process according to one or more embodiments is shown; and

[0015] Figure 5 A flowchart illustrating an example algorithm according to one or more implementations is depicted. Detailed Implementation

[0016] The implementation herein relates to a swept pattern probability calculation (SPPC) for speed and defect identification within bearings. Bearing defects on bearings and related machinery can include, but are not limited to, spalls or flakes detaching from the bearing raceways (inner or outer raceways) and / or rollers and / or roller cages due to brinelling, false brinelling, corrosion, contamination, lack of lubrication, or excessive rolling pressure (e.g., due to spalling and fracture). According to one or more embodiments, the SPPC automatic detection algorithm can be implemented by one or more devices to automatically detect bearing defects on bearings and related machinery without knowing the exact shaft speed or the specific brand / model of the bearing mounted on a given bearing housing.

[0017] For example, because the bearings and related machinery of a railway axle-box can provide a vibration spectrum during use (e.g., bearing vibration harmonics), when a bearing defect develops, a defect component / symptom may appear in the vibration spectrum. The SPPC automatic detection algorithm according to various non-limiting embodiments of this teaching can identify the type of bearing defect even without knowing the axle speed and the exact brand / model of the bearing installed on a given axle-box. The identified bearing defect type can then be used to determine the precise axle speed, which can be used to perform further diagnostic operations and / or degradation analysis. Although the embodiments described herein are for railway axle-boxes, they are not limited thereto. That is, although the embodiments herein involve handling track condition monitoring errors (such as wheel diameter management errors required to convert GPS linear velocity to axle rotational speed), the embodiments herein are suitable for many condition monitoring applications across many industries (where tachometers or speed inputs are not installed or are unavailable).

[0018] Figure 1 A system 100 according to one or more embodiments is depicted. System 100 includes at least one railcar 101, and the railcar 101 includes at least one axle box 103. The axle box 103 includes one or more wheels 104 (e.g., railbogie wheels) attached thereto by fastening elements. Note that although only a single axle box is shown, most railcars have two bogies and therefore two axles with eight wheels and eight axle boxes attached thereto (e.g., via axle box bearings of the railbogie wheels). Typically, the bearing housing of the axle box 103 includes axle box bearings 105 (e.g., one or more bearings 105) of the railbogie wheels and bolted connections that rotatably connect the bearing housing to the corresponding wheel 104, and bolted connections that attach the bearing housing to the axle box 103. For example, a train typically includes two to more than seventy railcars 101, which means that there can be thousands of bearings within a system 100 that includes a fleet of trains.

[0019] Furthermore, system 100 is generally shown according to one or more embodiments. System 100 may include an electronic computer framework that includes and / or employs any number and combination of computing devices and networks utilizing various communication technologies (as described herein). System 100 can be readily scalable, extensible, and modular, with the ability to be changed for different services or to reconfigure some features independently of other features.

[0020] System 100 includes at least one sensor device 110 from a plurality of condition monitoring sensor devices. Each sensor device 110 is an electronic device that may include: a housing 111, a battery 112, at least one sensor 113 (e.g., a transducer for vibration, temperature, etc.), a data collector 115 (e.g., a processor and memory as described herein), a GPS 114, data transmission electronics 117 (e.g., a wireless modem and / or a near-field communication (NFC) transponder), and an attachment assembly 118 (e.g., one of its plurality of fixing bolts) for securing the sensor device 110 to a wheel 104. The attachment assembly 118 may be any bracket, flange, etc., for attaching the sensor device 110 to the mechanical system to be monitored.

[0021] For example, each sensor device 110 may be a compact, battery-operated device (e.g., using battery 112) that measures static and dynamic data (e.g., condition monitoring data) of the bearings of the wheel 104 to which it is attached (e.g., specifically, at least one of the fastening elements attached to the wheel 104). Each sensor device 110 may wirelessly transmit the condition monitoring data (as indicated by double arrow 119) to devices, servers, and systems such as computing device 120 via data transmission electronics 117.

[0022] According to one or more embodiments, the memory and / or data transmission electronics 117 of the data collector 115 of each sensor device 110 may store condition monitoring (results) and / or may be associated with a unique sensor identifier. For example, an NFC transponder may be pre-programmed with a unique sensor identifier associated with the wireless modulation internals of the sensor device 110, and / or with details relating to that particular sensor and its installation location (e.g., whether it is mounted on or near the axle box bearing of a track bogie wheel). Furthermore, the sensor device 110 records condition monitoring data at various predefined intervals and with speed gating (e.g., when the railcar 101 is moving and not parked in the rail yard).

[0023] The computing device 120 includes one or more central processing units (CPUs) (collectively or generally referred to as electronic hardware processor 121, or simply processor 121). Processor 121 is connected to memory 122 and various other components via a system bus. Memory 122 may include read-only memory (ROM) and random access memory (RAM). ROM is connected to the system bus and may include a basic input / output system (BIOS) that controls specific basic functions of the computing device 120. RAM is a read-write memory connected to the system bus for use by processor 121. Memory 122 stores data 124 and software 125.

[0024] Data 124 includes a set of values ​​for qualitative or quantitative variables organized in various data structures to support and be used by the operation of software 125. According to one or more embodiments, memory 122 may accumulate and / or store data 124 from sensor device 110 for use by computing device 120. In this regard, for example, data 124 may include condition monitoring data (e.g., bearing vibration and temperature, bearing vibration harmonics) and speed ranges (e.g., the range from the highest desired speed to the lowest desired speed at which the shaft of axle housing 103 can rotate / turn due to the bearing), approximate (i.e., imprecise) speed values. In one or more embodiments, for each type of defect to be detected, the speed value is an assumed speed when the machine (e.g., railcar 101) operates within a given speed range for a sustained threshold time (e.g., greater than 80% of the time), a root mean square (RSS) value, a bearing designation, a unique sensor identifier, a predefined interval for data accumulation, and a frequency range for bearing category defects. Although no specific bearing brand / model or specific bearing defect frequency is indicated, the range of bearing category defect frequencies can be predetermined (e.g., pre-calculated) to include bearing defect frequencies corresponding to the bearing brand / model category. In one or more examples, the shaft speed can be defined as revolutions per minute, as determined by GPS calculations using an approximate track wheel diameter.

[0025] Further note that each of the one or more defect patterns can be a set of frequencies over time (e.g., as the class of bearing defects arise) for a class of bearing defects, while the frequencies of a particular bearing type and / or a particular bearing defect are unknown. In this respect, the set of frequencies is associated with defect components / symptoms outside of healthy bearing operation. Defect test patterns can be weighted such that the maximum match (e.g., the maximum match between frequency and defect component / symptom) gives the highest value relative to other matches. Each defect component / symptom in the defect test pattern has a maximum value of 1, but is typically less than 1. Examples of one or more defect test patterns may include a Ball Over Frequency Outside (BPFO) pattern for detecting outer raceway defect frequencies, a Ball Over Frequency Inside (BPFI) radial and axial load pattern for detecting inner raceway defect frequencies, a Ball Spin Frequency (BSF) pattern for detecting ball bearing defect frequencies, and a Cage Baseband Frequency (FTF) pattern for detecting cage defect frequencies. The test patterns can be weighted so that the BPFO pattern has 1×BPFO for every 5 harmonics, the BPFI radial and axial load patterns have 1×BPFI for every 3 harmonics and have 1×N sidebands, the BSF pattern has 1×BSF or 2×BSF and a small number of harmonics with FTF sidebands, and the cage FTF pattern has 1×FTF and a small number of harmonics.

[0026] Software 125 is stored as instructions for execution on processor 121. That is, memory 122 is also an example of a tangible storage medium that can be read by processor 121, wherein software is stored as instructions for execution by processor 121 to make system 100 run (operate), such as those shown in Figures 2 to 3. Figure 3 The instructions described herein. Note that software 125 can reside anywhere within many types of condition monitoring systems and can provide storage operations, trend (analysis) operations, and alarm operations. When a defect is present, SPPC provides axle speed, defect type, and frequency for use in calculating the corresponding system condition indicator (CI). For example, according to one or more embodiments, as described herein, the software may include an SPPC automatic detection algorithm. Typically, the SPPC automatic detection algorithm can be implemented by computing device 120 to automatically detect bearing defects on bearings (e.g., track axle box bearings) of axle box 103 without knowing the accurate axle speed, thereby saving costs (e.g., labor time) and reducing errors in managing constantly changing wheel diameters.

[0027] When executing the SPPC automatic detection algorithm of the software, the computing device 120 scans several weighted test patterns across (i) a specified speed range and (ii) a bearing category defect frequency range, and calculates the RSS value of pattern-to-noise filtered spectrum correlations for each step and for each defect test pattern type. The most probable defect type is identified by the test pattern that provides the maximum value. In one or more non-limiting embodiments, the individual test patterns are weighted such that when more than one test pattern spans a set of spectral components, only the test pattern with the best match (probability) gives the highest value. In one or more non-limiting embodiments, the initial scan can be enhanced by zeroing out carpet noise and unidentifiable peaks higher than carpet noise using various methods.

[0028] In the following description, post-sweep logic is performed from a series of the latest (or last) or most recent number of "N" measurements or results. In one or more embodiments, a lookup table (LUT) stored in memory may include "N" rows for each pattern, where each row defines a measurement or result. In one or more non-limiting embodiments, the measurements or results include, but are not limited to, axis RPM, pattern matching values, baseband frequency, and sideband frequency. Post-sweep logic processing includes multiple analyses or conditional tests to confirm that the initially most likely detected defect type is indeed the actual defect type. The conditional test includes: (i) determining that more than N / 2 out of “N” measurements belong to the same defect type; (ii) wherein the pattern correlation value greater than N / 2 is greater than or equal to a specified threshold (e.g., typically a gE peak of 0.2 to 0.3, or a peak of 2 to 3 if in m / s^2); and (iii) wherein the error of the identified baseband and sideband frequencies greater than N / 2 correlated with the shaft RPM (as determined from GPS speed and wheel “mid-diameter”) is smaller than a specified GPS % accuracy range. In one or more non-limiting embodiments, the maximum error is due to a constant (within N measurements) offset caused by the unknown wheel diameter, while the GPS / NSS system adds a smaller random error. Therefore, taking into account GPS / GNSS errors, the ratio within the most recent N measurements can change less than the GPS / GNSS error 80% of the time. In other words, the GPS percentage (%) accuracy range can be set to 80% of measurements with an error of less than + / -3%. Therefore, the system and method described herein provide a post-scan logic processing capable of determining a first logic value (e.g., a logic "0" indicating negative or inaccurate detection, or a logic "1" indicating positive detection) for each bearing defect pattern, which effectively confirms that the initially most likely detected bearing defect type identified using the scan processing is indeed the actual bearing defect type. Precise defect baseband and sideband frequencies can also be provided for diagnostic purposes.

[0029] Computing device 120 includes one or more input / output (I / O) adapters 123 coupled to a system bus. The one or more I / O adapters 123 may include a Small Computer System Interface (SCSI) adapter that communicates with system memory 122 and / or any other similar components. The one or more I / O adapters 123 may include an NFC transponder that communicates with an NFC transponder of sensor device 110. For example, the one or more I / O adapters 123 may interconnect the system bus with network 130 (which may be an external network), enabling system 100 to communicate with other such systems (i.e., server 140).

[0030] System 100 also includes network 130 and server 140. Network 130 includes a group of computers connected together and sharing resources. As described herein, network 130 can be any type of network, including local area network (LAN), wide area network (WAN), or Internet. Server 140 includes processor 142 and memory 144 (as described herein) and provides various functions to computing device 120, such as sharing and storing data 124, providing processing resources, and / or performing calculations (e.g., implementing software 125).

[0031] According to one or more embodiments, for example, server 140 may be a cloud-hosted condition monitoring system that executes software (e.g., software 125 including an SPPC automatic detection algorithm) stored in memory 144 via processor 142. Furthermore, at various predefined intervals (e.g., when the railcar 101 is parked in a rail parking lot at the end of its use), the cloud-hosted condition monitoring system of server 140 downloads and stores data (e.g., data 124 including unique sensor identifiers and / or corresponding condition monitoring data) from sensor device 110. Therefore, the software of server 140 can use the data therein to perform operations similar to those of software 125 of computing device 120.

[0032] Now go to Figure 2A and Figure 2B The processing flow 200 implemented by system 100 is described according to one or more embodiments. Processing flow 200 can be implemented by any component of system 100. Typically, regarding processing flow 200, speed and exact or specific bearing details are unknown. That is, not only is the exact shaft speed (e.g., rpm) unknown, but also the exact bearing details (such as a specific bearing brand / model and the corresponding bearing defect frequency) are unknown. Processing flow 200 can also be enhanced by various methods to “zero” spectral blanket noise and unidentifiable peaks higher than the blanket (noise).

[0033] Processing flow 200 begins at operation 202, and a computer (e.g., computing device 120 and / or server 140) receives / accumulates status monitoring data from one or more sensors (e.g., sensor device 110). According to one or more embodiments, the status monitoring data, as well as other data described herein, can be transmitted from sensor device 110 (e.g., such as...). Figure 1 (As indicated by the double arrow 119 in the diagram) to the computing device 120. More specifically, the condition monitoring data includes vibration harmonics of the bearings associated with the shaft. The computing device 120 can also forward the condition monitoring data, along with other data described herein, to the server 140 via the network 130. Thus, both the computing device 120 and the server 140 accumulate sufficient information to support the execution of the processing flow 200. The accumulation of condition monitoring data can occur at predefined intervals, and in some cases, the accumulation (operation) is performed twice a day (e.g., before the railcar 101 leaves the rail parking lot and after its return).

[0034] At operation 204, a computer (e.g., computing device 120 and / or server 140) performs scanning processing on the condition monitoring data. The scanning processing includes scanning one or more test patterns through (i) a speed range and (ii) a bearing category defect frequency range calculated for each step and for each defect pattern type to be detected. Although no specific bearing brand / model or specific bearing defect frequency is indicated, the bearing category defect frequency range can be predetermined (e.g., pre-calculated) to include bearing defect frequencies corresponding to the bearing brand / model. Therefore, the bearing category defect frequency range can be used to narrow down the possible pattern components and specific baseband and sideband frequency ranges (suitable for bearing defect diagnosis and detection). According to one or more embodiments, computing device 120 and / or server 140 can store the speed range in their respective memories 122 and 144. Although no precise speed is indicated, the speed range can be predefined from the highest desired speed to the lowest desired speed for the conditions at the time of measurement and includes multiple speed steps.

[0035] According to a non-limiting embodiment, the scanning process moves the test pattern across the spectrum in several iterations, where each iteration is referred to as a "velocity step". In one or more non-limiting embodiments, the step size is set to 1 / 2 window ( / interval) (bin). Furthermore, computing device 120 and / or server 140 may execute software (e.g., software 125) to scan / apply these test patterns at each velocity step within a velocity range, the software calculating a velocity / pattern-related RSS value (e.g., a small portion of the window at the highest frequency component) for each velocity step and each test pattern type. One or more windows correspond to the spectrum such that if there is a 1000 Hz spectrum with 800 lines, each window for each line has a value representing how much vibrational energy (e.g., a width of 1.25 Hz) is associated with the center frequency of that window.

[0036] Go to Figure 3 A graph 300 is depicted according to one or more embodiments. Graph 300 illustrates an example of pattern scanning processing of the vibration harmonics 302 of the test pattern 306 scanned by the SPPC automatic detection algorithm. The pattern scanning process moves the test pattern through the spectrum in small increments of the baseband frequency.

[0037] Vibration component frequency 308 is identified by pattern component 310. In one or more embodiments, each pattern component 310 corresponds to several components defined by the order (and the number of sidebands on either side of each order, if present). During a baseband sweep, pattern component 310 becomes coincident with vibration component 308. As pattern component 310 becomes coincident with vibration component 308, the product obtained by multiplying the RSS (root mean square) value of the weighted pattern component by the corresponding spectral window value (which they align within the sweep step) reaches the maximum value of the pattern. Therefore, among several different types of defect patterns, the defect pattern with the largest maximum value identifies the most likely type of defect. As shown, graph 300 also illustrates other examples of vibration harmonics 312 and 314 scanned by pattern components 316 and 318, respectively, using the SPPC automatic detection algorithm. It should also be noted that pattern weighting means that if more than one pattern spans a set of spectral components, only the pattern with the best match (probability) will give (or assign) the highest value.

[0038] In some instances, when it is known that the selected test pattern used in the pattern scanning process includes one or more pairs of sidebands, the pattern scanning process also includes performing sideband scanning processing. For example, Figure 4A graph 350 is depicted in relation to the sideband pattern scanning process according to one or more embodiments. The graph 350 depicts the first order 352, the second order 354, and the third order 356 of the fundamental signal component.

[0039] The first-order 352 of the fundamental signal component includes sidebands having the same positive or negative (+ / -) frequency step size as the first-order 352 of the frequency component. For simplicity, additional sideband groups are not shown. However, the first-order 352 of the fundamental signal component may include additional sidebands without departing from the scope of the invention. The first set of sidebands associated with the first-order 352 of the fundamental signal component includes a first sideband scan range start position 358 and a first sideband scan range end position 360. Therefore, the sideband pattern scanning process includes performing a first sideband start scan 362 (+ / - frequency) and a first sideband end scan 364 (+ / - frequency).

[0040] The second-order 354 of the fundamental signal component includes sidebands with the same + / - frequency step size as the first-order 352 of the frequency component. The first set of sidebands associated with the second-order 354 of the fundamental signal component includes a second sideband scan range start position 366 and a second sideband scan range end position 368. Therefore, the sideband pattern scanning process includes performing a second sideband start scan 370 (+ / - frequency) and a second sideband end scan 372 (+ / - frequency).

[0041] The third-order 356 of the fundamental signal component includes sidebands with the same + / - frequency step size as the first-order 352 of the frequency component. The first set of sidebands associated with the third-order 356 of the fundamental signal component includes a third sideband scan range start position 374 and a third sideband scan range end position 376. Therefore, the sideband pattern scanning process includes performing a third sideband start scan 378 (+ / - frequency) and a third sideband end scan 380 (+ / - frequency).

[0042] The sideband pattern scanning process includes performing a full sideband scan for each increment (step) of the baseband scan process described herein. The sideband scan involves the sidebands on either side of each of the fundamental order Δ frequencies (i.e., Δ intervals), which vary the range of sideband defects from the baseband frequency of a particular sideband defect to + / - the range of the sideband defect.

[0043] For sideband scanning processing, the step size is set such that the widest sideband component increments, i.e., stepped. In at least one non-limiting embodiment, interpolation is not used in pattern correlation. For example, when interpolation is not used in pattern correlation, the step size is set to 1 / 2 window for each sideband scan step. Therefore, by reducing the step size to below 1 / 2 window, neither frequency accuracy nor amplitude accuracy increases (e.g., for the highest order in the pattern). However, when interpolation is used in pattern correlation, the step size is set to 1 / 4 window or even smaller. Therefore, both accuracies increase as the step size decreases, but the processing time increases.

[0044] In one or more non-limiting embodiments, the sideband scanning process is performed as a loop within the baseband scanning process. For example, once a defect test pattern has been created and weighted, the number of sidebands (if any) for the defect test pattern can be determined. When the defect test pattern includes sidebands, the scan range and step size are determined, and the sideband scanning process is performed as a loop within the baseband scanning loop executed according to the baseband scanning process. In one or more non-limiting embodiments, the sideband scanning process can be omitted when the defect test pattern does not include sidebands.

[0045] Refer again Figure 2A At operation 206, the computer multiplies each pattern component (e.g., one or more defect test patterns) by a matching ambient spectral component. In some example implementations, the multiplication of each pattern component in one or more patterns by the matching ambient spectral component is performed using the interpolated matching ambient spectral component. In other example implementations, the multiplication of each pattern component in one or more test patterns by the matching ambient spectral peak value is performed using quadratic peak interpolation. At operation 208, the computer adds the pattern components together. In some example implementations, the pattern components can be added together using the root mean square (RSS). At operation 210, the addition generates one or more output values. The output values ​​include, but are not limited to, the frequency of each defined pattern. In one or more implementations, the output values ​​may be stored in a database or a first-in-first-out (FIFO) memory with a buffer size of "N". Therefore, the post-scan logic processing described herein can access a database or buffer to obtain a series of up-to-date "N" measurements, which indicate the most likely type of bearing defect identified by the storage frequency of each defined pattern.

[0046] Proceed to operation 212 (see Figure 2B The computer performs a post-sweep logic process on the N most recent measurements indicating the most likely bearing defect types. According to a non-limiting embodiment, the operations included in the post-sweep logic process are indicated by dashed box 213. At operation 214, the defect type with the highest pattern matching value for each of the N most recent measurements is determined. At operation 216, the number (referred to as "N1") of the most popular (most frequent) defects of the same type among the N most recent measurements is determined. At operation 218, it is determined that the number of defects of the same type (i.e., N1) is greater than a target value (N / 2). At operation 220, it is determined that among the defects of the same type (i.e., N1), their number (referred to as "N2") satisfies a pattern correlation threshold (TH). PC The pattern correlation value (PC) is determined at operation 222. This is done when TH is satisfied. PC Of the defects of the same type (i.e., N2), their number (here referred to as "N3") is greater than the target value (i.e., N / 2). At operation 224, TH will be satisfied. PC The number of defects of the same type (i.e., N3) is determined by having identified baseband and sideband frequencies, the identified baseband and sideband frequencies having an error related to shaft RPM less than a specified GPS percentage (%) accuracy range. In one or more non-limiting embodiments, the GPS percentage (%) accuracy range is set to 80% of the measured values ​​with an error of less than + / - 3%.

[0047] At operation 226, based on the determination that TH is satisfied... PC If the number of defects of the same type (i.e., N3) is greater than the target value (i.e., N / 2), the most likely bearing defect type is identified as the exact bearing defect type. Furthermore, at operation 226, the bearing defect frequency and accurate shaft speed, i.e., the exact or actual shaft speed, are confirmed. The exact shaft speed is confirmed by the ratio of the present defect to the known bearing defect in the vibration signal. In any of operations 218-224, if the target value (i.e., not greater than N / 2) is not met, or if the TH value is not met... PC (That is, N2 is less than TH) PCWarnings can be generated when the RPM correlation is not less than the error. For example, at dashed box 228 (e.g., optional operation), the computer outputs one or more results. In this regard, technicians can easily identify any bearing problems monitored by the computer and take remedial measures (e.g., replace or repair the bearing). Note that if no defect component is present, it is not important whether the speed is known. If there are other spectral components in the condition monitoring data (e.g., from machine dynamics / mechanics), these other spectral components can also have patterns associated with them to calculate the speed even in the absence of bearing defects.

[0048] Now go to Figure 5 The flowchart illustrates an example algorithm 500 for performing sideband scanning processing according to a non-limiting embodiment. Example algorithm 400 begins with blocks 402 and 404 receiving initial conditioning monitoring data. The initial conditioning monitoring data includes, but is not limited to, the vibration spectrum of the envelope acceleration measurement for shaft rpm (as shown in block 401), the calculated velocity range, the scan step size, and the frequency range of bearing category defects calculated for each step size (as shown in block 404). Although the specific pattern components of the bearing defect and the specific baseband and sideband frequency ranges are unknown, the category of the bearing defect frequency can be used to narrow down the possible pattern components and the specific baseband and sideband frequency ranges (suitable for bearing defect diagnosis and detection). In one or more non-limiting embodiments, when the GPS has an error range (e.g., + / - 5%) and the wheel diameter differs from the diameter used to calculate RPM (axis speed) by an error value (e.g., + / - 5%), the speed range across the test pattern includes an acceptable minimum error value based at least in part on the GPS error range and the wheel diameter error value, in this example, the acceptable minimum error value would be at least about + / - 10%.

[0049] Then, at box 410, example algorithm 400 initializes the variables. For example, the correlation value, baseband frequency, and sideband frequency are initialized to zero, respectively. At box 415, for each vibration harmonic of the bearing, a FOR loop is entered. More specifically, for the baseband range (e.g., from high to low), example algorithm 400 steps across the vibration harmonics to scan the test pattern. Although the scan direction is described as from high to low, it should be understood that the scan direction can be from low to high without departing from the scope of the invention. At decision block 425 (as indicated by the DO arrow), the FOR loop includes determining whether the number of sidebands is greater than zero. If the number of sidebands is not greater than zero, example algorithm 400 proceeds to box 430 (e.g., following the "No" arrow).

[0050] At box 430, the relevant function is invoked, and at decision box 440, it is determined whether any of the relevant values ​​is greater than the stored value. If the relevant value is greater than the stored value, the example algorithm 400 proceeds to box 445 (e.g., following the "Yes" arrow) to perform an update operation. Then, the example algorithm 400 moves to the next pattern at box 450, which returns the algorithm to box 415. Upon returning to box 415, the example algorithm 400 returns the specific defect type (e.g., for any identified relevant value and baseband frequency, as shown in box 451). At box 445, the relevant value and frequency are updated. If the sideband's relevant value is not greater than the stored value, the example algorithm 400 proceeds to box 450 (e.g., following the "No" arrow).

[0051] Returning to decision box 425, if the number of sidebands is greater than zero, example algorithm 400 proceeds to box 460 (e.g., following the "Yes" arrow). At box 460, for each vibration harmonic of the bearing, another FOR loop is entered. More specifically, for the sideband range (low to high), example algorithm 400 progressively moves across the vibration harmonic scan pattern by sideband step size. After completing the FOR loop, the method proceeds to box 463 to perform post-scan logic (PSL) processing. The PSL processing is described above, for example, in reference to... Figure 2B It has been described in detail, so for the sake of brevity it will not be repeated.

[0052] At box 465, the relevant function is called. At decision box 470, it is determined whether any of the relevant values ​​is greater than the stored value. If the relevant value is not greater than the stored value, example algorithm 400 proceeds to box 450 (e.g., following the "No" arrow). If the relevant value is greater than the stored value, example algorithm 400 proceeds to box 480 (e.g., following the "Yes" arrow). At box 480, the relevant values ​​and frequencies are updated. Then, example algorithm 400 proceeds to box 450.

[0053] Various embodiments of the invention have been described herein with reference to the accompanying drawings. Alternative embodiments of the invention are conceived without departing from its scope. Various connections and positional relationships (e.g., above, below, adjacent, etc.) between elements are illustrated in the above description and drawings. Unless otherwise stated, these connections and / or positional relationships may be direct or indirect, and the invention is not intended to be limited in this respect. Thus, the connection of entities may refer to direct or indirect connection, and the positional relationship between entities may be direct or indirect positional relationship. Furthermore, the various tasks and processing steps described herein may be incorporated (or included) into a more comprehensive procedure or process with additional steps or functions not described in detail herein.

[0054] The following limitations and abbreviations are used to interpret the claims and description. As used herein, the terms “comprising,” “including,” “comprise,” “having,” “containing,” or “comprising,” or any other variations thereof, are intended to cover non-exclusive inclusion. For example, a composition, mixture, process, method, article, or apparatus that comprises a list of elements is not necessarily limited to those elements, but may include other elements not expressly listed or inherent to such a composition, mixture, process, method, article, or apparatus.

[0055] Additionally, the term "exemplary" is used herein to mean "serving as an example, instance, or illustration." Any implementation or design described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other implementations or designs. The terms "at least one" and "one or more" can be understood to include any integer greater than or equal to one, i.e., one, two, three, four, etc. The term "multiple" can be understood to include any integer greater than or equal to two, i.e., two, three, four, five, etc. The term "connection" can include both indirect "connection" and direct "connection."

[0056] The terms “about,” “generally,” “approximately,” and variations thereof are intended to include the degree of error associated with a measurement based on a specific quantity of equipment available at the time of filing this application. For example, “about” may include a range of ±8%, ±5%, or ±2% of a given value.

[0057] For the sake of brevity, conventional techniques related to the implementation and use of aspects of the invention may or may not be described in detail herein. In particular, various aspects of the computing systems and specific computer programs used to implement the various technical features described herein are well known. Therefore, for the sake of brevity, many conventional implementation details are only briefly mentioned or omitted entirely, without providing details of well-known systems and / or processes.

[0058] This invention can be a system, method, and / or computer program product at any possible level of technical detail integration. The computer program product may include a computer-readable storage medium (or media) having computer-readable program instructions on it for causing a processor to execute aspects of the invention.

[0059] Computer-readable storage media can be tangible means capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example, but not limited to, electronic storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. A non-exhaustive list of more specific examples of computer-readable storage media includes the following: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable optical disc read-only memory (CD-ROM), digital versatile optical disc (DVD), memory sticks, floppy disks, machine encoding devices (such as punched cards or raised structures in slots on which instructions are recorded), and any suitable combination of the foregoing. As used herein, computer-readable storage media should not be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses passing through fiber optic cables), or electrical signals transmitted through wires.

[0060] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to a suitable computing / processing device, or downloaded via a network (e.g., the Internet, a local area network, a wide area network, and / or a wireless network) to an external computer or external storage device. The network may include copper transmission cables, optical fiber transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to a computer-readable storage medium within the suitable computing / processing device.

[0061] Computer-readable program instructions used to perform the operations of this invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, integrated circuit configuration data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​(such as Smalltalk, C++, etc.) and procedural programming languages ​​(such as the "C" programming language or similar programming languages). The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the latter case, the remote computer may be connected to the user's computer via any type of network (including a local area network (LAN) or a wide area network (WAN)) or may be connected to an external computer (e.g., via the Internet provided by an Internet service provider). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs) may execute computer-readable program instructions by personalizing the electronic circuitry with state information utilizing the computer-readable program instructions to perform aspects of the invention.

[0062] Various aspects of the invention will be described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block (or frame) in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0063] These computer-readable program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus for production machines, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions / actions specified in one or more boxes of the flowchart and / or block diagram. These computer-readable program instructions may also be stored in a computer-readable storage medium (which can instruct a computer, programmable data processing apparatus, and / or other means to function in a particular manner), such that the computer-readable storage medium storing the instructions includes an article of manufacture comprising instructions for implementing aspects of the functions / actions specified in one or more boxes of the flowchart and / or block diagram.

[0064] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer-implemented process, and to cause the instructions to be performed on the computer, other programmable apparatus or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0065] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of feasible implementations of systems, methods, and computer program products according to various embodiments of the invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of instructions comprising one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions indicated in the blocks may not be executed in the order shown in the drawings. For example, depending on the function involved, two consecutively shown blocks may actually be executed substantially simultaneously, or these blocks may sometimes be executed in reverse order. It will also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented by a dedicated hardware-based system that performs the specified function or action or implements a combination of dedicated hardware and computer instructions.

[0066] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting. Unless the context clearly indicates otherwise, the singular form is intended to include the plural form as used herein. It will be further understood that when the terms “comprising” and / or “including” are used in this specification, they enumerate the presence of the stated features, quantities, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, quantities, steps, operations, elements, components, and / or groups thereof.

[0067] Various embodiments have been described herein for illustrative purposes, but are not intended to be exhaustive or limited to the disclosed embodiments. Many variations and modifications will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology chosen for use herein is for the best explanation of the principles, practical applications, or technical improvements to existing (or discovered) technologies in the market, or to enable those skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method comprising: receiving, by a processor, condition monitoring data from one or more sensors, the condition monitoring data comprising vibration harmonics of at least one bearing coupled to a rotatable shaft; performing, by the processor, a pattern scan process that scans at least one test pattern through both a speed range and a bearing class defect frequency range, wherein, in response to the test pattern having at least one test pattern sideband, the at least one test pattern sideband is also scanned through a sideband range for the condition monitoring data to determine a baseband frequency and a sideband frequency of the pattern from best match values; determining, by the processor, a most likely bearing defect type associated with the at least one bearing based on the best match values in two or more results associated with the at least one test pattern; and performing, by the processor, a post-scan logic process that compares a number (N) of most recent results from the pattern scan process to at least one conditional test to confirm that the most likely bearing defect type is present. The condition monitoring data does not include one or both of an exact rotational speed of the shaft and a specific bearing defect frequency of the at least one bearing.

2. The method of claim 1, wherein, The speed range comprises a plurality of individual speed steps.

3. The method of claim 2, wherein, The pattern scan process comprises performing a plurality of baseband scans on the baseband frequency, each baseband scan performed for a respective speed step included in the speed range.

4. The method of claim 3, wherein, The pattern scan process comprises performing a plurality of sideband scans, each sideband scan performed on at least one set of sidebands and after each baseband scan performed for a given speed step.

5. The method of claim 4, wherein, The at least one conditional test comprises a plurality of conditional tests performed on a selected number of most recent results generated from the pattern scan process, and wherein, in response to all of the conditional tests generating a positive result, the actual bearing defect type is confirmed.

6. The method of claim 5, wherein, At least one of the conditional tests determines a positive correlation between both a baseband frequency of the selected number of most recent results and at least one set of sidebands of the baseband frequency and a predicted rotational speed of the shaft.

7. The method of claim 6, wherein, The positive correlation is determined in response to an error in the correlation between both the baseband frequency of the selected number of most recent results and at least one set of sidebands of the baseband frequency and the predicted rotational speed of the shaft being less than a speed threshold.

8. The method of claim 7, wherein, 9. A bearing defect automatic detection system comprising: a processor in signal communication with one or more sensors and configured to receive condition monitoring data from the one or more sensors, the condition monitoring data comprising vibration harmonics corresponding to at least one bearing coupled to a rotatable shaft, wherein the processor is configured to: perform a pattern scan process that scans at least one test pattern through both a speed range and a bearing class defect frequency range, wherein, in response to the test pattern having at least one test pattern sideband, the at least one test pattern sideband is also scanned through a sideband range for the condition monitoring data to determine a baseband frequency and a sideband frequency of the pattern from best match values; ​ determining a most likely bearing defect type associated with the at least one bearing based on the best match value from the two or more results associated with the at least one test pattern; and performing a post-scan logic process that compares a number (N) of most recent results from the pattern scan process to at least one conditional test to confirm the presence of the most likely bearing defect type.

10. The automatic bearing defect detection system of claim 9, wherein The condition monitoring data does not include one or both of an exact rotational speed of the shaft and a specific bearing defect frequency of the at least one bearing.

11. The automatic bearing defect detection system of claim 10, wherein The speed range includes a plurality of individual speed steps.

12. The automatic bearing defect detection system of claim 11, wherein The pattern scan process includes performing a plurality of baseband scans on the baseband frequencies, each baseband scan being performed for a respective speed step included in the speed range.

13. The bearing defect automatic detection system of claim 12, wherein The pattern scan process includes performing a plurality of sideband scans, each sideband scan being performed on at least one set of sidebands and after each baseband scan performed for a given speed step.

14. The automatic bearing defect detection system of claim 13, wherein The at least one conditional test includes a plurality of conditional tests performed on a selected number of most recent results generated from the pattern scan process, and wherein the actual bearing defect type is confirmed in response to all of the conditional tests generating a positive result.

15. The automatic bearing defect detection system of claim 14, wherein, At least one of the conditional tests determines a positive correlation between both a baseband frequency of the selected number of most recent results and at least one set of sidebands of the baseband frequency and a predicted rotational speed of the shaft.

16. The automatic bearing defect detection system of claim 15, wherein The positive correlation is determined in response to an error in the correlation between both the baseband frequency of the selected number of most recent results and at least one set of sidebands of the baseband frequency and the predicted rotational speed of the shaft being less than a speed threshold.

17. A computer program product for controlling an electronic hardware processor to perform automatic detection of bearing defects, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by the processor to perform operations comprising: receiving condition monitoring data from one or more sensors, the condition monitoring data including vibration harmonics of at least one bearing coupled to a rotatable shaft; performing a pattern scan process that scans at least one test pattern through both a speed range and a bearing class defect frequency range, wherein, in response to the test pattern having at least one test pattern sideband, the at least one test pattern sideband is also scanned through a sideband range for the condition monitoring data to determine a baseband frequency and a sideband frequency of the pattern from a best match value; determining a most likely bearing defect type associated with the at least one bearing based on the best match value from the two or more results associated with the at least one test pattern; and performing a post-scan logic process that compares a number (N) of most recent results from the pattern scan process to at least one conditional test to confirm the presence of the most likely bearing defect type. The condition monitoring data does not include one or both of an exact rotational speed of the shaft and a specific bearing defect frequency of the at least one bearing.

18. The computer program product of claim 17, wherein, ​ 19. The computer program product of claim 18, wherein, The speed range includes a plurality of individual speed steps.

20. The computer program product of claim 19, wherein, The pattern scan process includes performing a plurality of baseband scans on the baseband frequencies, each baseband scan being performed for a respective speed step included in the speed range.

21. The computer program product of claim 20, wherein, The pattern scan process includes performing a plurality of sideband scans, each sideband scan being performed on at least one set of sidebands and after each baseband scan performed for a given speed step.

22. The computer program product of claim 21, wherein, The at least one condition test includes a plurality of condition tests performed on a selected number of latest results generated from the pattern scan process, and wherein the actual bearing defect type is confirmed in response to all of the condition tests generating positive results.

23. The computer program product of claim 22, wherein, At least one of the condition tests determines a positive correlation between both the baseband frequencies of the selected number of latest results and at least one set of sidebands of the baseband frequencies and the predicted rotational speed of the shaft.

24. The computer program product of claim 23, wherein, The positive correlation is determined in response to an error that is less than a speed threshold value in the correlation between both the baseband frequencies of the selected number of latest results and at least one set of sidebands of the baseband frequencies and the predicted rotational speed of the shaft.

Images