System and method for monitoring a drive unit bearing

Abstract

A vehicle and a system and method for health monitoring of a drive unit bearing in the vehicle are provided. A system for health monitoring of a drive unit bearing in a vehicle includes a sensor mounted to a drive unit of the vehicle and configured to provide a vibration signal; and a processor configured to: pre-process the vibration signal from the sensor to enhance a fault feature and form a frequency domain signal; and processing the frequency domain signal to estimate a health phase of the drive unit bearing.

Description

Technical Field

[0001] The present disclosure relates to monitoring drive unit bearings. More particularly, the present disclosure relates to health monitoring of drive unit bearings using vibration sensors. Background Art

[0002] Bearings in drive units, such as those used in motor vehicles, can fail during use. For example, drive unit bearings can fail due to high mileage fatigue (spalling) or electrical damage (wrinkling) in electric vehicles.

[0003] Therefore, it is desirable to provide methods and systems for determining drive unit bearing health. Furthermore, other desirable features and characteristics of the present disclosure will become apparent from the subsequent detailed description and the appended claims, taken in conjunction with the accompanying drawings and the foregoing background. Summary of the invention

[0004] A system for health monitoring of a drive unit bearing in a vehicle is provided, the system comprising: a sensor mounted to a drive unit of the vehicle and configured to provide a vibration signal; and a processor configured to: pre-process the vibration signal from the sensor to enhance fault characteristics and form a frequency domain signal; and process the frequency domain signal to estimate a health stage of the drive unit bearing.

[0005] The processor is configured to process the frequency domain signals to extract and fuse the health indicators, and estimate the fault level and the fault location according to the health indicators.

[0006] In some embodiments of the system, the processor is configured to process the frequency domain signal to detect whether the bearing is healthy or faulty.

[0007] In some embodiments of the system, the processor is configured to process the frequency domain signal to isolate a fault in a specific bearing and determine a location of the fault on the specific bearing.

[0008] In some embodiments of the system, the processor is configured to pre-process the vibration signal from the sensor by performing phase domain transform, discrete frequency removal, impulse enhancement, envelope filtering, and short time Fourier transform (STFT).

[0009] In some embodiments of the system, the processor is configured to process the frequency domain signal by the steps of: calculating a health indicator; applying a regression model to the health indicator to find a health stage; and comparing the health stage to a predefined threshold to detect a health state of the drive unit bearing.

[0010] In some embodiments of the system, the processor is configured to perform a calibration process comprising: fitting a regression model to a health indicator to find a health stage; calculating a modeling error; sorting the health stages; selecting a specific health stage for fusion; and determining predefined weights for the health stages, and wherein output from the calibration process is stored in the system.

[0011] In some embodiments of the system, the processor is configured to: calculate a health indicator; apply a predefined model to estimate a fault level (health stage) based on the health indicator, apply predefined weights to generate a weighted health stage; sum the weighted health stages to provide a fused health stage; and compare the fused health stage to a predefined threshold to detect the health status of the drive unit bearing.

[0012] In some embodiments of the system, the processor is further configured to transmit an output indicative of a health stage of a drive unit bearing.

[0013] In some embodiments, the system further comprises a motor speed sensor for sensing motor speed and / or a torque sensor for sensing torque, and the processor is configured to determine whether the motor speed is within a speed range and / or determine whether the torque is within a torque range suitable for monitoring the drive unit bearings.

[0014] A method for monitoring the health of a drive unit bearing in a vehicle is provided, the method comprising: sensing vibration with a sensor at a known position relative to the drive unit and providing sensor data; performing a short-time Fourier transform (STFT) to form a spectrum signal over a plurality of segments; calculating a health indicator of the drive unit bearing based on the spectrum signal and a critical frequency of the drive unit bearing; and determining a health stage of the drive unit bearing based on each health indicator.

[0015] In some embodiments of the method, determining the health stage of the drive unit bearing based on each health indicator includes applying a regression model to each health indicator.

[0016] In some embodiments, the method further comprises detecting a health status of a drive unit bearing based on the health stage.

[0017] In some embodiments, the method also includes performing a calibration process, including: fitting a regression model to the health indicator to find a health stage; calculating a modeling error; sorting the health stages; selecting a specific health stage for fusion; determining a predefined weight for the health stage; and storing the output of the calibration process.

[0018] In some embodiments, the method further comprises applying predefined weights to the calculated health indicators to generate a weighted health stage.

[0019] In some embodiments, the method further comprises calculating a fused health stage based on the weighted health stages.

[0020] In some embodiments, the method further comprises comparing the fused health stage to a predefined threshold to detect a health status of the drive unit bearing.

[0021] In some embodiments, the method further comprises transmitting an output indicative of a health stage of a drive unit bearing.

[0022] In some embodiments of the method, the drive unit bearing is a first drive unit bearing, and the method further comprises monitoring the health of a second drive unit bearing.

[0023] A vehicle is provided, comprising a drive unit bearing; a vibration sensor configured to provide sensor data; and one or more processors configured individually or collectively to: pre-process the sensor data to generate a frequency spectrum; and process the frequency spectrum to estimate a health stage of the drive unit bearing. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The present disclosure will be described below in conjunction with the following drawings, wherein like reference numerals represent like elements, and wherein:

[0025] Figure 1 is a schematic diagram of a vehicle according to an exemplary embodiment.

[0026] Figure 2 According to an exemplary embodiment Figure 1 A perspective view of a portion of a drive unit shown in , illustrating the location of some of the drive unit bearings.

[0027] Figure 3 is a method for packaging according to an exemplary embodiment Figure 1 and Figure 2 A perspective view of the drive unit housing of a drive unit and showing the locations of some vibration sensors.

[0028] Figure 4 is a flow chart of a method for monitoring a drive unit bearing in a motor vehicle according to an exemplary embodiment. DETAILED DESCRIPTION

[0029] The following detailed description is merely exemplary in nature and is not intended to limit application and use. In addition, it is not intended to be bound by any express or implied theory presented in the previous background technology, the invention content, or the detailed description below. As used herein, the term module refers to any hardware, software, firmware, electronic control unit or component, processing logic and / or processor device alone or in any combination, including but not limited to: application-specific integrated circuits (ASICs), electronic circuits, processors (shared, dedicated or groups) and memories that execute one or more software or firmware programs, combinational logic circuits, and / or other suitable components that provide the described functions.

[0030] Embodiments of the present disclosure may be described herein in terms of functional and / or logical block components and various processing steps. It should be understood that these block components may be implemented by any number of hardware, software, and / or firmware components configured to perform specified functions. For example, embodiments of the present disclosure may employ various integrated circuit components, such as memory elements, digital signal processing elements, logic elements, lookup tables, etc., which may perform various functions under the control of one or more microprocessors or other control devices. In addition, those skilled in the art will appreciate that embodiments of the present disclosure may be practiced in conjunction with any number of autonomous driving systems (including cruise control systems, automatic driver assistance systems, and autonomous driving systems), and the vehicle system described herein is merely an example embodiment of the present disclosure.

[0031] For the sake of brevity, conventional techniques related to other functional aspects of signal processing, data transmission, signaling, control, and systems (and the various operating components of the systems) may not be described in detail herein. In addition, the connecting lines shown in the various figures included herein are intended to represent example functional relationships and / or physical couplings between various elements. It should be noted that many alternative or additional functional relationships or physical connections may exist in the embodiments of the present disclosure.

[0032] In embodiments herein, methods and systems capable of early detection and isolation of degraded drive unit bearings use health indicators extracted from critical bearing frequencies in a vibration spectrum. Thus, propulsion losses and potential safety issues associated with drive unit bearing failures can be avoided. Embodiments herein are provided for electric vehicles (EVs), autonomous vehicles (AVs), internal combustion engine vehicles, hybrid vehicles, or other suitable vehicles.

[0033] Referring to the drawings, wherein like reference numerals refer to like parts, Figure 1A vehicle 10 is shown. The vehicle 10 includes a powertrain 12 configured to launch and propel the vehicle, i.e., operate the vehicle in all speed ranges between low road speeds and high road speeds via drive wheels 14. An exemplary powertrain 12 includes an engine or motor 18. For example, the vehicle may be an electric vehicle (EV), an internal combustion engine (ICE) vehicle, or a hybrid vehicle. The powertrain 12 is operatively connected to an electromechanical drive unit 22.

[0034] The drive unit 22 also includes a drive unit housing 30 that serves as a stationary member to which the electric machine 18 and certain other torque transmitting devices (not shown) may be secured.

[0035] exist Figure 1 In the embodiment of the present invention, the vehicle 10 is provided with a motor speed sensor 13, a torque sensor 15, and an accelerometer / vibration sensor 60. In addition, the vehicle 10 includes a system 55 for monitoring the drive unit bearings. In some embodiments, the system 55 is coupled to the drive unit 22 as described below. The system 55 may include or communicate with the motor speed sensor 13, the torque sensor 15, and the acceleration sensor 60. In addition, the system 55 includes a controller 50, which may communicate with the motor speed sensor 13, the torque sensor 15, and the acceleration sensor 60. In some embodiments, the controller 50 may be part of a larger system processor.

[0036] It should be noted that the system 55 may be provided for health monitoring of mechanical components of a drive unit of an EV or a transmission of an ICE vehicle or components of a hybrid vehicle. Figure 2 An exemplary drive unit for an EV is shown. Figure 2 As shown in , the drive unit 22 includes a plurality of bearings 40. For example, corresponding bearings 40 may be located on the input shaft 19, the transmission shaft 17 and / or the output shaft 11. The bearings 40 may include ball bearings, cylindrical roller bearings, tapered ball bearings or other suitable bearings.

[0037] Figure 3 shows the package used Figure 2 The housing 30 of the drive unit 22 is shown in FIG. Figure 3 As shown in , the vibration sensors 60 are mounted on the outer surface 71 of the drive unit housing 30. Each vibration sensor 60 can be dedicated to the analysis of a specific bearing 40 inside the drive unit housing 30. In other embodiments, more than one vibration sensor 60 can be used to analyze one bearing 40. In other embodiments, one vibration sensor 60 can be used to analyze more than one bearing 40. Although Figure 3The vibration sensors 60 are shown mounted on an outer surface 71 of the drive unit housing 30, but in other embodiments, some or all of the vibration sensors 60 may be located within the drive unit housing 30. In either case, each vibration sensor 60 is located at a known location, a known distance from one or more bearings 40, and has a known orientation relative to one or more bearings 40.

[0038] exist Figure 1 In the embodiment of the present invention, the system 55 may include a controller 50, or receive instructions from a controller 50. The term "controller" and related terms such as electronic control unit or engine control unit (ECU) may include or refer to one or more combinations of application specific integrated circuits (ASICs), electronic circuits, central processing units (e.g., microprocessors), and associated non-transitory memory components in the form of memory and storage devices (read-only, programmable read-only, random access, hard drives, etc.).

[0039] The non-transitory memory component can store machine-readable instructions in the form of one or more software or firmware programs or routines, combinational logic circuits, input / output circuits and devices, signal conditioning and buffering circuits, and other components that can be accessed by one or more processors to provide the described functions. The input / output circuits and devices include analog / digital converters and related devices that monitor inputs from sensors, monitoring these inputs at a preset sampling frequency or in response to a triggering event. Software, firmware, programs, instructions, control routines, codes, algorithms and similar terms refer to sets of instructions executable by controllers, including calibration and lookup tables. Each controller executes a control routine to provide the desired function. The routine can be executed periodically. Alternatively, the routine can be executed in response to the occurrence of a triggering event. Communication between the controller 50, the sensor 13, 15, 60 and the system 55 is completed in various arrangements using a direct wired point-to-point link, a networked communication bus link, a wireless link or another suitable communication link. Communication includes exchanging data signals in an appropriate form, including, for example, exchanging electrical signals via a conductive medium, exchanging electromagnetic signals via air, exchanging optical signals via optical waveguides, etc. Data signals may include discrete, analog or digitized analog signals representing inputs from sensors, actuator commands, and communications between controllers. The term "signal" refers to a physically discernible indicator of information conveyed and may be a suitable waveform (e.g., electrical, optical, magnetic, mechanical, or electromagnetic) capable of propagating through a medium, such as DC, AC, sine, triangle, square, vibration, etc.

[0040] It should be noted that the data from the vibration sensor can be processed on-board (on the vehicle) or off-board. For the on-board case, the ECU can be used as the controller 50. For the off-board case, the data can be sent to the back office controller 50 and the processing can be done offline. In addition, the method for monitoring the bearing 40 can be real-time health monitoring or offline health monitoring.

[0041] like Figure 1 As shown in , the controller 50 can accept information from various sources, process the information, and provide control commands based on the information to achieve results such as the operation of the vehicle 10 and its systems (including system 55). In the illustrated embodiment, the controller 50 includes a processor 57 and a memory device 58, and is coupled to a storage device 59. The processor 57 performs the computational and control functions of the controller 50 and can be any type of processor or multiple processors, a single integrated circuit such as a microprocessor, or any suitable number of integrated circuit devices and / or circuit boards that work together to accomplish the functions of the processing unit. During operation, the processor 57 can execute one or more programs and can use data, each of which can be contained in the storage device 59, and as such, the processor 57 controls the general operation of the controller 50 in performing the processes described herein, such as the processes and methods described in more detail below.

[0042] The memory device 58 may be any type of suitable memory. For example, the memory device 58 may include volatile and non-volatile storage such as read-only memory (ROM), random access memory (RAM), and keep-alive memory (KAM). The KAM is a permanent or non-volatile memory that can be used to store various operating variables when the processor 57 is powered off. The memory device 58 may be implemented using any of a variety of known memory devices, such as a PROM (programmable read-only memory), an EPROM (erasable PROM), an EEPROM (electrically erasable PROM), flash memory, or any other electrical, magnetic, optical, or combination memory device capable of storing data, some of which represent executable instructions used by the controller 50. In the illustrated embodiment, the memory device 58 may store the above-mentioned programs as well as one or more stored data values, such as for short-term data access.

[0043] The storage device 59 stores data, such as long-term data access for automatically controlling the vehicle 10 and its systems. The storage device 59 can be any suitable type of storage device, including direct access storage devices, such as hard disk drives, flash memory systems, floppy disk drives, and optical disk drives. The storage device 59 is a non-transitory computer-readable medium that is configured to store programs and data, such as parameters related to the vehicle 10. In an exemplary embodiment, the storage device 59 is the source from which the memory device 58 receives a program that performs one or more embodiments of one or more processes of the present disclosure. In another exemplary embodiment, the program can be directly stored in the memory device 58 and / or otherwise accessed by the memory device 58. The program represents executable instructions used by the controller 50 to process information and control the vehicle 10 and its systems (including the system 55).

[0044] It will be understood that while the exemplary embodiment is described in the context of a fully functional computer system, those skilled in the art will recognize that the mechanisms of the present disclosure are capable of being distributed as a program product in which one or more types of non-transitory computer-readable signal-bearing media are used to store the program and its instructions and to perform its distribution, such as a non-transitory computer-readable medium that carries the program and contains computer instructions stored therein for causing a computer processor (such as processor 57) to execute the program. Such a program product may take a variety of forms, and the present disclosure applies equally regardless of the specific type of computer-readable signal-bearing medium used to perform the distribution. Examples of signal-bearing media include recordable media, such as floppy disks, hard disks, memory cards, and optical disks, as well as transmission media, such as digital and analog communication links. It will be understood that cloud-based storage and / or other technologies may also be used in certain embodiments. Similarly, it will be understood that the computer system of controller 50 may also differ in other ways. Figure 1 The embodiment shown in .

[0045] In some embodiments, the controller 50 operates the system 55 to perform a method for monitoring the bearing 40 , and in particular for detecting bearing degradation, for isolating a degraded bearing, and / or for locating a bearing fault in a drive unit of a vehicle.

[0046] refer to Figure 4 , a flow chart of a method 400 for monitoring bearings 40 in a drive unit 22 of a motor vehicle 10 is shown. The method 400 may include isolating a fault, ie, isolating a defective bearing among a plurality of drive unit bearings, and identifying the location of the fault within the faulty bearing.

[0047] In various implementations, method 400 includes, at act 402, receiving a vehicle signal. The vehicle signal may include vibration, motor torque and speed signals, such as signals from sensors 60, 15 and 13, and / or other signals. For example, system 55 or controller 50 may receive the vehicle signal.

[0048] At act 404, method 400 includes triggering an algorithm for processing the vibration signal to monitor drive unit bearing health. In one non-limiting example, the algorithm may be triggered when the motor speed is within a selected range and / or when the torque speed is within a selected range.

[0049] After the algorithm is triggered at act 404 , the method 400 continues to pre-process the vibration signal at pre-processing block 410 .

[0050] For example, at action block 411, method 400 transforms the vibration signal from a time-based signal to a phase-based signal. More specifically, method 400 transforms the vibration signal captured from vibration sensor 60 to the phase domain using the motor speed signal. For example, method 400 may use the pulse count component of the motor speed signal from an encoder in motor speed sensor 13, or may integrate the angular velocity value. Transforming the vibration signal from a time-based signal to a phase-based signal eliminates velocity variations in the signal.

[0051] At action block 412, method 400 filters other vibrations from the signal. For example, the discrete frequency removal process removes vibrations generated from other components, such as other vehicle components, other drive unit components including gears, and / or may remove road noise. The discrete frequency removal process includes any suitable technique for removing unwanted vibrations or interference by identifying and filtering out specific frequencies. For example, the discrete frequency removal process may use autoregressive linear prediction, adaptive noise cancellation, random / discrete separation, and / or time synchronous averaging (TSA) techniques.

[0052] Autoregressive Linear Prediction (ARLP) uses a mathematical model to predict the future values ​​of a signal based on a linear combination of past values. ARLP can be used to estimate the coefficients of a filter, which can be used to remove specific frequencies from a signal. Specifically, the filter is designed based on past values ​​of the signal, which are used to estimate the coefficients of the filter. The signal is then filtered to remove unwanted frequency components.

[0053] Adaptive Noise Cancellation (SAN) is based on the concept of adaptive filtering, where filters are designed to remove noise from a signal based on the characteristics of the noise itself. In SAN, the filter is adjusted in real time based on the input signal, thereby adapting to changing noise conditions.

[0054] Random / discrete separation techniques involve separating a signal into discrete components and then selectively removing components that contain unwanted noise. This can be achieved through techniques such as Fourier analysis, which decomposes the signal into individual frequency components. Once the individual components are identified, the unwanted frequencies can be removed, leaving behind a cleaner signal, i.e. one with a high signal-to-noise ratio.

[0055] like Figure 4 As shown in , method 400 also includes, at action block 413, enhancing the fault signature, i.e., enhancing the signal impulse. Specifically, method 400 can enhance the pulse events or transients in the signal. The vibration from the fault is pulsed. Therefore, enhancing the signal impulse will enhance the fault signature. In addition, action block 413 can remove road noise from the signal.

[0056] Therefore, to further enhance the impulsivity in the signal, an impulsivity enhancement process, such as bandpass filtering, is performed to selectively enhance the impulsive events in the signal, such as based on a desired frequency range. Bandpass filtering is a technique for filtering out unwanted frequency components from a signal while retaining a desired frequency range of impulsive events in the signal. The desired frequency range, i.e., the selection criterion, is the frequency range where the impulsivity is greatest using spectral kurtosis.

[0057] Additionally or alternatively, the impulsivity in the signal can be enhanced by using techniques such as minimum entropy deconvolution (MED). Minimum entropy deconvolution is a technique for recovering impulsive events in a signal. The technique works by minimizing the entropy of the recovered impulse response, thereby producing a deconvolved signal with enhanced impulsive content.

[0058] exist Figure 4 , method 400 may continue to apply the envelope at action block 414. For example, method 400 may use an envelope filtering process that isolates the envelope or amplitude variations of a signal while attenuating its high frequency content. In some embodiments, the envelope filtering process isolates transient components of a signal, i.e., sudden and brief changes in amplitude, such as spikes or pulses. The envelope filter may act as a low pass filter to pass only the low frequency components of the signal corresponding to the envelope. The result is a signal that contains only the low frequency components of the input, which represent its envelope.

[0059] The method 400 also includes, at action block 415, performing a short-time Fourier transform (STFT). For example, the method 400 can analyze the signal by dividing the signal into small overlapping segments, applying a window (such as Hanning or Hamming) on ​​the small segments to reduce frequency domain artifacts, and performing a Fourier transform or a fast Fourier transform (FFT) on each windowed segment to obtain its frequency content. Thus, the STFT process transforms the signal from the phase domain to the frequency domain to form a spectral signal for each short segment.

[0060] In summary, method 400 uses predefined startup criteria and vehicle signals (including but not limited to motor torque and speed signals) to determine whether to trigger the algorithm. In addition, the motor speed signal is used to transform the vibration signal captured from the vibration sensor from the time domain to the phase domain. Thereafter, a discrete frequency removal method (e.g., autoregressive linear prediction or adaptive noise elimination or random / discrete separation) is applied to the vibration signal in the phase domain to filter out vibrations of other drive units and vehicle components (including but not limited to shafts and gears). Then, the signal impulsivity is enhanced by removing noise and amplifying any potential fault features using minimum entropy deconvolution or bandpass filtering based on spectral kurtosis. Next, an envelope filter is applied to the signal with enhanced impulsivity to enhance the fault features. After enhancing the fault features, the signal is divided into small windows or buffers, a window function is applied to each window, and a short-time Fourier transform is performed on each window to generate a vibration spectrum.

[0061] As a result of action blocks 411 to 415 in pre-processing block 410 , at action block 415 , a vibration spectrum is generated from the vibration signal received from the vibration sensor at action block 402 .

[0062] The method 400 may include, at action block 420, storing critical frequencies of the bearings. Specifically, for each bearing and its position, critical frequencies (including inner race ball passing frequency (BPFI), outer race ball passing frequency (BPFO), and ball spinning frequency (BSF)) are calculated based on the known geometry of the bearing. This one-time calculation of critical frequencies is then stored in a memory in the system 55 or in a memory accessible by the system 55, such as in a memory in the controller 50 or accessible by the controller 50.

[0063] At action block 450 , method 400 selects a bearing for analysis.

[0064] At action block 451 , method 400 selects a location on a selected bearing (eg, inner race, outer race, etc.) for analysis.

[0065] At action box 452, method 400 calculates a health indicator for a selected bearing position based on the spectrum and the stored critical frequency. In some embodiments, the health indicator includes the amplitude of a fast Fourier transform (FFT) at the critical frequency, the relative value of the FFT at the critical frequency, the peak height in a small window centered on the critical frequency, the normalized peak height in a small window centered on the critical frequency, the absolute value of the energy in a small window centered on the critical frequency, the relative value of the energy in a small window centered on the critical frequency, the sideband of the critical frequency, the relative sideband of the critical frequency, or various combinations thereof. The health indicator can be calculated by one or more techniques. In some embodiments, the health indicator can be calculated based on the spectrum of each window based on the critical frequency. The health indicator can be determined for each segment (i.e., window or buffer).

[0066] In some embodiments, the health indicator module includes normalization. For example, although larger vibrations (due to larger torque or higher speed) will produce larger FFT amplitudes, larger FFT amplitudes do not indicate a higher fault level. The health indicators (absolute FFT, peak, sidebands) can be divided by the root mean square (RMS) of each segment in the phase domain. Optionally, the health indicator is divided by the median of the FFT amplitude of a window (in the frequency domain) centered on the critical frequency.

[0067] Additionally or alternatively, health indicator module 87 may perform relative energy normalization, which involves dividing the amplitude value in each frequency band by the total energy in that frequency band. This has the effect of removing amplitude variations due to variations in operating conditions or measurement equipment.

[0068] like Figure 4 As shown in , method 400 may include a calibration process 430 to be performed once. At action box 431, the calibration process 430 includes fitting a regression model to the health indicator (HI) to the actual degradation level (the ground truth of the degradation level). In an exemplary embodiment, this is performed using all windows. This calibration step determines the model for each health indicator (HI) in this manner. Specifically, action box 431 will be repeated for all health indicators (HI) because each health indicator requires a model. In other words, passing the HI (all windows) to the model is repeated for all HIs (i.e., relative peak HI, energy HI, etc.). For each health indicator, the value of the health indicator (all windows) is input and the estimated degradation level is output. The estimated degradation level is called a health stage (HS). The health stage can be an estimate of the bearing ground truth of the health condition.

[0069] Furthermore, at action block 432 , the difference between the healthy stage and the actual fault level is calculated as a modeling error.

[0070] Action blocks 431 and 432 may be repeated for all health indicators to find corresponding health stages and errors.

[0071] The calibration process 430 also includes, at action block 433, ranking the health stages (corresponding health indicators) based on criteria such as modeling error, monotonicity, linearity, receiver operating characteristic (ROC), robustness, etc.

[0072] The calibration process 430 continues at action block 434 by selecting which health stages (corresponding to which health indicators) to fuse. For example, the method 400 may use a forward selection method to determine which health stages to fuse.

[0073] In addition, the calibration process 430 includes, at action block 435, determining a predetermined weight for each selected health stage for fusion. For example, the inverse of the modeling error of each health stage, i.e., (1 / modeling error), can be used as the predetermined weight. Thus, a health stage with a smaller modeling error has a larger weight, while a health stage with a larger modeling error has a smaller weight.

[0074] The output of the calibration process 430 includes the HI-HS model, the list of HSs to be fused, and the predefined weight of each HS, i.e., 1 / modeling error. As shown, at action block 440 of method 400, these outputs of the calibration process 430 are stored and can be used as predefined parameters of the algorithm of method 400.

[0075] Thus, after calculating the health index at action block 452, method 400 may continue at action block 456 by applying the predefined regression model stored as a calibration output to the health index (HI) to find the health stage (HS).

[0076] At action block 457 , method 400 may continue by applying the predetermined weights stored as an output of the calibration process to each selected health stage for fusion.

[0077] The method 400 continues at action block 458 by summing the weighted health stages to provide a fused health stage (ie, estimated fault level) for each segment of the selected bearing.

[0078] Additionally, method 400 may include, at action block 459, averaging the health stages across segments to produce a mature health stage. The mature size (i.e., the number of buffers / segments used for averaging) may be determined based on need. The larger the number of buffers, the more mature the decision; however, the longer the detection time.

[0079] At action block 461, method 400 includes applying a predefined threshold value at the fused health stage to detect the health status, i.e., healthy or faulty, of the selected bearing at the selected location. For example, when the fused health stage is greater than the threshold value, the selected bearing at the selected location is classified as faulty, and when the fused health stage is less than the threshold value, the selected bearing at the selected location is classified as healthy.

[0080] The threshold can be determined through empirical testing and validation. Setting the threshold too high will result in false negatives, while setting the threshold too low will result in false positives.

[0081] Method 400 may then repeat from action block 451, selecting another location on the selected bearing for analysis, to applying a predefined threshold in the fused health stage at action block 461 to detect the health of the next location of the selected bearing, ie, healthy or faulty.

[0082] After all locations of the selected bearing have been analyzed, method 400 may repeat, beginning at action block 450, selecting a next bearing for analysis at each location on the next bearing.

[0083] Thus, in summary, action blocks 450 to 461 may be repeated to determine which bearing(s) are faulty and to determine the location of the fault in the faulty bearing (inner race, outer race, etc.). The critical frequencies for different locations (inner race (BPFI), outer race, etc.) are different. The critical frequencies for different bearings are also different (depending on the bearing geometry). Thus, for each bearing, action blocks 451 to 461 are repeated for its critical frequencies (various locations), and thus, method 400 determines whether the bearing is healthy or faulty in the state of health (SOH) determination, and for each location of each bearing, estimates the fault level in the health stage (HS) determination. The process is then repeated on all bearings within the drive unit to isolate the fault (i.e., which bearing), and identify the location (i.e., where in the bearing). In the event of a fault detection, an output or warning message regarding the fault detection and isolation is transmitted.

[0084] Furthermore, method 400 may be considered to detect the presence of a fault (ie, health state), estimate the level of the fault (ie, health stage), isolate the detected fault (such as a specific bearing), and locate the identified fault (such as a specific portion of a bearing).

[0085] The method 400 may also include, at action block 470, transmitting an output to a user. The output may indicate the health status and health stage (faulty or healthy) of each location in each bearing. Alternatively, the output may only report the location of the faulty bearing and the fault on the faulty bearing, as well as the fault level (health stage).

[0086] Although at least one exemplary embodiment has been proposed in the foregoing detailed description, it should be understood that there are a large number of variations. It should also be understood that one or more exemplary embodiments are merely examples and are not intended to limit the scope, applicability or configuration of the present disclosure in any way. On the contrary, the foregoing detailed description will provide a convenient roadmap for implementing one or more exemplary embodiments for those skilled in the art. It should be understood that various changes may be made to the functions and arrangements of the elements without departing from the scope of the present disclosure as set forth in the attached claims and their legal equivalents.

Claims

1. A system for health monitoring of a drive unit bearing in a vehicle, the system comprising: a sensor mounted to a drive unit of the vehicle and configured to provide a vibration signal; as well as A processor, the processor being configured to: preprocessing the vibration signal from the sensor to enhance fault characteristics and form a frequency domain signal; and The frequency domain signal is processed to estimate a health stage of the drive unit bearing.

2. The system according to claim 1, wherein: The processor is configured to process the frequency domain signals to extract and fuse health indicators, and estimate a fault level and a fault location based on the health indicators.

3. The system according to claim 1, wherein: The processor is configured to process the frequency domain signal to detect whether the bearing is healthy or faulty.

4. The system according to claim 1, wherein: The processor is configured to process the frequency domain signal to isolate a fault in a particular bearing and determine a location of the fault on the particular bearing.

5. The system according to claim 1, wherein: The processor is configured to pre-process the vibration signal from the sensor by performing phase domain transform, discrete frequency removal, impulse enhancement, envelope filtering, and short time Fourier transform (STFT).

6. The system according to claim 1, wherein: The processor is configured to process the frequency domain signal by the following steps: Calculate health indicators; applying a regression model to the health indicator to find the health stage; and The health stage is compared to a predefined threshold to detect a health status of the drive unit bearing.

7. The system according to claim 1, wherein: The processor is configured to perform a calibration process, comprising: Regression models were fitted to the health indicators to find the health stages; Calculate modeling errors; ranking said health stages; Selecting specific health stages for fusion; and Predefined weights for the health stages are determined, and wherein output from the calibration process is stored in the system.

8. The system according to claim 7, wherein: The processor is configured to: Calculate health indicators; applying the predefined weights to generate a weighted health stage; summing the weighted health stages to provide a fused health stage; and The fused health stage is compared with a predefined threshold to detect a health status of the drive unit bearing.

9. The system according to claim 1, wherein: The processor is further configured to transmit an output indicative of the stage of health of the drive unit bearing.

10. A vehicle comprising: Drive unit bearings; a vibration sensor configured to provide sensor data; as well as One or more processors that are individually or collectively configured to: preprocessing the sensor data to generate a frequency spectrum; and The frequency spectrum is processed to estimate the stage of health of the drive unit bearing.

Images