System and method for improved anomaly detection for rotating machines

Abstract

This disclosure relates to systems and methods for improved anomaly detection of rotating machines. Example methods may include determining the rotational speed of the rotating machine. Example methods may also include using a frequency domain transform of a signal from the rotating machine to determine a frequency domain signal. Example methods may also include determining a first frequency band within the frequency domain signal for identifying fault frequencies of the rotating machine based on the rotational speed of the rotating machine. Example methods may also include determining the fault frequency of the rotating machine within the first frequency band. Example methods may also include determining a second frequency band within the first frequency band based on the fault frequency, wherein the second frequency band includes the fault frequency. Example methods may also include determining a first fault indicator and a baseline of the first fault indicator based on the second frequency band. Example methods may also include determining a fault condition of the rotating machine based on the deviation of the second fault indicator from the baseline. Example methods may also include providing an alarm based on the fault condition of the rotating machine.

Description

Technical Field

[0001] This disclosure relates to systems and methods for detecting anomalies in rotating machines (in some cases, the terms "machine" and "motor" may be used interchangeably herein). In some embodiments, this disclosure may more particularly relate to early fault detection in rotating machines (e.g., for line-fed and variable frequency drive (VFD) fed induction and synchronous machines). Background Technology

[0002] Electrical signature analysis (ESA) is a widely used method for detecting faults in rotating machinery using measured current and voltage signals. To ensure these fault detection algorithms work properly based on measurements available at the installation site, they are also used in a wide range of applications for which rotating machinery is deployed. Typically, standard settings are used in intelligent electronic devices (IEDs) for anomaly detection, which are not well-suited for different types of machines and applications.

[0003] Currently, many different methods exist for detecting such anomalies. A first example of a conventional method discussed in WO Patent Publication No. 2011158099, entitled "System and Method of speed detection in an AC induction machine," involves a linear speed estimation algorithm based on nameplate information. Nameplate information can refer to defining parameters for the machine, such as power factor, efficiency, torque, and current at rated voltage and frequency, among other parameters. This method can estimate the rotor speed using the frequency domain (e.g., based on slot harmonics used to estimate multiple rotor bars) and determine if the rotor speed is valid. If the rotor speed is valid, the adjusted rotor speed is determined partly based on the linear estimation algorithm and partly through frequency domain analysis. A second example of a conventional method discussed in WO Patent Publication No. 2019167086, entitled "A System for Assessment of multiple faults in induction motors," involves adjusting the input signal to remove fundamental components and estimating the slip speed and rotor speed by finding the mixed eccentricity fault components inherently present in all motors. The third example conventional method discussed in U.S. Patent No. 6,449,567, entitled "Apparatus and Method for determining shaft speed of a Motor," involves calculating the shaft speed based on the shaft frequency peak from a demodulated current signal (amplitude demodulation of the current to obtain the instantaneous current and finding candidate peaks). This third example conventional method also determines the pole-pass frequency (e.g., as the difference in angle between the instantaneous current phase and the instantaneous voltage phase). The final shaft speed can be determined from the pole-pass frequency, while the shaft speed can be determined from the shaft frequency peak. Summary of the Invention

[0004] This application provides a system comprising: at least one processor; and at least one memory storing computer-executable instructions. When executed by the at least one processor, the computer-executable instructions cause the at least one processor to: determine the rotational speed of a rotating machine; determine a frequency domain signal using a frequency domain transformation of a signal from the rotating machine; determine a first frequency band within the frequency domain signal for identifying a fault frequency of the rotating machine based on the rotational speed of the rotating machine; determine a fault frequency of the rotating machine within the first frequency band; determine a second frequency band within the first frequency band based on the fault frequency, wherein the second frequency band includes the fault frequency; determine a first fault indicator and a baseline of the first fault indicator based on the second frequency band; determine a fault condition of the rotating machine based on the deviation of the second fault indicator from the baseline; and provide an alarm based on the fault condition of the rotating machine. The computer-executable instructions further cause the at least one processor to: determine that a voltage signal for the rotating machine is unavailable; determine that the rotating machine is an inductive machine; and determine the rotational speed of the rotating machine based on at least one of: the torque component of the input current of the rotating machine; a determined peak corresponding to an inherent asymmetry in the frequency domain transformation of a fault signal of the rotating machine; and the synchronous speed of the rotating machine and the rated sliding speed of the rotating machine.

[0005] This application, in another aspect, provides a method comprising: determining the rotational speed of a rotating machine; determining a frequency domain signal using a frequency domain transformation of a signal from the rotating machine; determining a first frequency band within the frequency domain signal for identifying a fault frequency of the rotating machine based on the rotational speed of the rotating machine; determining a fault frequency of the rotating machine within the first frequency band; determining a second frequency band within the first frequency band based on the fault frequency, wherein the second frequency band includes the fault frequency; determining a first fault index and a baseline of the first fault index based on the second frequency band; determining a fault condition of the rotating machine based on a deviation of the second fault index from the baseline; and providing an alarm based on the fault condition of the rotating machine. The method further comprises: determining that a voltage signal for the rotating machine is unavailable; and determining that the rotating machine is an inductive machine, wherein determining the rotational speed of the rotating machine is based on at least one of: a torque component of an input current of the rotating machine; a determined peak corresponding to an inherent asymmetry in the frequency domain transformation of the fault signal of the rotating machine; and a synchronous speed of the rotating machine and a rated sliding speed of the rotating machine. Attached Figure Description

[0006] A detailed description is given with reference to the accompanying drawings. The same reference numerals may be used to indicate similar or identical items. Various embodiments may utilize elements and / or components other than those illustrated in the drawings, and some elements and / or components may not be present in various embodiments. Elements and / or components in the drawings are not necessarily drawn to scale. Throughout this disclosure, singular and plural terms may be used interchangeably depending on the context.

[0007] Figure 1 Example flowcharts are depicted based on one or more example embodiments of this disclosure.

[0008] Figure 2 Example flowcharts are depicted based on one or more example embodiments of this disclosure.

[0009] Figure 3 Example flowcharts are depicted based on one or more example embodiments of this disclosure.

[0010] Figures 4A-4C Example searches and sumbands are depicted according to one or more example embodiments of this disclosure.

[0011] Figure 5 Example methods are described based on one or more example embodiments of this disclosure.

[0012] Figure 6 An illustrative computing device is described according to one or more exemplary embodiments of the present disclosure.

[0013] Figure 7 An illustrative comparison of baseline and real-time data is depicted based on one or more example embodiments of this disclosure. Detailed Implementation

[0014] Overview

[0015] In some embodiments, this disclosure relates to systems and methods for improved anomaly detection of rotating machines. In some embodiments, the systems and methods described herein can improve upon conventional anomaly detection systems and methods (e.g., the three described above) in certain circumstances, and more particularly, can describe a broader approach to improved anomaly detection for different types of machine failure modes (e.g., eccentricity, broken rotor bars, or bearing failure failure modes). In some embodiments, one improvement to the systems and methods described herein may include improved motor speed estimation. This can be advantageous across various technical solutions because the estimated motor speed can then be used to determine a baseline of failure indicators for different types of failure modes, which can then be used as a baseline to identify faults in the machine during real-time operation after the baseline is established (this can be referenced below, for example, at least). Figure 3 (To describe).

[0016] In some embodiments, the improved motor speed estimation can allow speed estimation to be performed even without voltage measurements from the machine. Voltage measurements may be unavailable, for example, the machine may not include any sensors capable of determining the machine's voltage level. As a second example, voltage measurements may also be unavailable if a voltage sensor is present but is damaged and cannot collect and / or provide voltage measurement data. As a third example, voltage measurements may also be unavailable if harmonics are present in the voltage. Voltage measurements may also be unavailable in the machine for many additional reasons. Even in the absence of voltage measurements, motor speed estimation can be performed using the torque component of the current or by utilizing inherent machine asymmetries (in inherently asymmetrical machines). However, the improved motor speed estimation methods described herein can also be used when voltage measurements are available from the machine. These improved speed estimation methods can be referenced... Figures 1-3 To describe.

[0017] In some embodiments, when voltage measurements are available, the systems and methods described herein can improve motor speed estimation by normalizing the determined input power of the machine based on the machine's operating frequency. The machine's input power can be a variable commonly used in machine speed estimation; however, when operating the machine with a variable frequency drive (VFD) (a VFD can be a type of motor controller that drives an electric motor by changing the frequency and voltage supplied to it), this normalization can improve this use by adapting to frequency variations. For example, for a rated machine power of 8 kW (at 50 Hz), assuming constant V / F operation, the maximum power drawn by the machine when operating at a frequency of 30 Hz may be less than 8 kW, even though the current drawn may be the same as at full load at 50 Hz. When the machine operates at 60 Hz, the maximum power drawn for constant V / F may be greater than 8 kW. Therefore, the calculated power may need to be normalized relative to the actual operating frequency. The normalized input power Pin_norm can be determined by, for example:

[0018] Pin = Va*Ia + Vb*Ib + Vc*Ic (Equation 1)

[0019] Pin_norm=Pin*(freq_rated / fund_freq) (Equation 2)

[0020] Pin = mean(Va*Ia + Vb*Ib + Vc*Ic) (Equation 3)

[0021] Where Va, Vb, and Vc can be the complete cycle of the acquired phase voltages, Ia, Ib, and Ic can be the complete cycle of the acquired currents, freq_rated can be the machine's nameplate frequency rating, fund_freq can be the machine's operating frequency, Pin can be the machine's calculated input power, and Pin_norm can be the normalized power used for the machine. Once the normalized input power is determined, the motor speed can be determined using the normalized input power and the motor's rated slip speed. The motor's rated slip speed can be the difference between the machine's magnetic field synchronization speed (sync_speed) and the machine's actual rated speed (rated_speed) of the shaft.

[0022] rated_slip_speed=120*freq_rated / P-rated_speed (Equation 4)

[0023] sync_speed=120*fund_freq / P (Equation 5)

[0024] For example, when the fundamental frequency is less than or equal to the machine's rated frequency, the motor speed motor_speed can be determined using the following equation:

[0025] motor_speed=sync_speed-abs(Pin_norm / Pin_rated)*rated_slip_speed (Equation 6)

[0026] For example, when the fundamental frequency is greater than the rated frequency, the motor speed motor_speed can be determined using the following equation:

[0027] motor_speed=sync_speed-abs(Pin / Pin_rated)*rated_slip_speed (Equation 7)

[0028] If normalized power is to be used, using the actual input power instead of normalized power for frequencies higher than the machine's rated frequency can correct for overestimation of speed.

[0029] In some embodiments, when voltage measurements are unavailable, the systems and methods described herein can improve speed estimation by estimating the speed using only current measurements from the machine. In particular, speed estimation can be performed using the torque component of the measured current, and more particularly, in some cases, when the machine operates at a constant V / f ratio. For example, when the operating frequency is less than the machine's rated frequency, the torque component of the current can be determined using the following equation:

[0030] Imn_rated = Irated * sin(phi) (Equation 8)

[0031] To account for nonlinear V / f operation, for frequencies higher than the rated frequency, the rated magnetizing current can be normalized using the operating frequency, for example:

[0032] Imn_rated=Irated*sin(phi)*freq_rated / fund_freq (Equation 9)

[0033] Itrq_rated=sqrt(Irated^2_Imn_rated^2) (Equation 10)

[0034] Itrq=sqrt(Irms^2-Imn_rated^2) (Equation 11)

[0035] Where φ is the rated power factor angle of the machine, Imn_rated is the rated magnetizing current of the machine, Irated is the rated current of the machine, Itrq is the actual torque component of the current of the machine, Itrq_rated is the rated torque component of the current of the machine, and Irms is calculated from the fast Fourier transform (FFT) of the stator current peak at the fundamental frequency of the machine as Ipeak / sqrt(2c). The actual torque component of the current (Itrq) can be determined by estimating the motor speed using the following equation with normalized power and rated slip speed:

[0036] rated slip speed=120*freq_rated / P-rated speed (Equation 12)

[0037] sync speed=120*fund freq / P (Equation 13)

[0038] motor speed=sync speed-abs(Itrq / Itrq_rated)*rated_slip_speed; (Equation 14)

[0039] Where fund_freq can be the machine's base frequency, and P can be the machine's pole number.

[0040] In some embodiments, a third example improvement to the system and method described herein may include an improved fault detection method and fault index calculation. Conventional systems can use a frequency search band having a range of machine operating frequencies to determine the frequency range used to search for fault frequencies. The system and method may be based on estimations such as the type of machine involved and / or execution speed (e.g., as may be discussed below). Figures 1-2These search bands can be optimized by adjusting the range of the search band (e.g., the frequency range of the search band) based on factors such as (described). The search band size can also be optimized based on the confidence level associated with the estimated motor speed used for the machine. For example, a narrower search band (e.g., this could be a more accurate speed estimate) can be used when using a particular type of speed estimation, such as when using input power, the torque component of the current, or machine asymmetry, and a wider search band (e.g., this could be a less accurate speed estimation method) can be used based on speed estimation methods that may not involve these components. Further details can be provided regarding... Figures 4A-4C Examples of such optimized search and banding are depicted. These optimized search and banding can be beneficial because they can reduce the frequency range that may need to be monitored to identify fault frequencies.

[0041] In some embodiments, additional improvements to the systems and methods described herein may include pre-determining the data length, sampling frequency (to ensure that fault frequencies can be appropriately captured based on the Nyquist criterion), and frequency resolution for collecting data from a given machine according to machine nameplate ratings and the application. Additionally, one or more data quality checks may be performed on the collected data to ensure that the quality of the collected data is sufficient and that there are no transients within the recorded time length for accurate anomaly detection to be performed. Additional data post-processing may be performed to ensure an integer number of periods for FFT analysis.

[0042] Turn to the attached diagram. Figure 1A more detailed flowchart 100 is provided, outlining the operations that may be performed for speed estimation of the machine. In some embodiments, the operations described herein may be performed by a computing device (e.g., computing device 600), which may be locally located to or remote from the machine. Flowchart 100 may begin with operation 102, which may involve receiving machine nameplate information and filtered data. As mentioned above, machine nameplate information may refer to known parameters associated with the machine, such as power factor, efficiency, torque, and current at the machine's rated voltage and frequency, as well as other parameters. Filtered data may refer to a predetermined data length and frequency resolution available for data collection from the machine, and may also correspond to any number of quality checks that may be performed on any data collected from the machine. Once the initial operations of receiving machine nameplate information and filtered data are performed, subsequent operations of flowchart 100 may depend on whether voltage measurements from the machine are available (i.e., whether voltage measurements can be obtained from the machine). As an example, voltage measurements may not be available, for example, the machine may not include any sensors capable of determining the machine's voltage level. As a second example, voltage measurement may also be unavailable if a voltage sensor is present but is undergoing damage and cannot collect and / or provide voltage measurement data. As a third example, voltage measurement may also be unavailable if harmonics are present in the voltage. Voltage measurement may also be unavailable in a machine for many additional reasons. Based on this, operation 106 of flowchart 100 may include a first condition determining whether voltage measurement is available from the machine. If it is determined that voltage measurement is available (the answer to the first condition of operation 106 is yes), then flowchart 100 may proceed to operation 108. Otherwise, if it is determined that voltage measurement is unavailable, then flowchart 100 may proceed to operation 116. Voltage measurement may be unavailable, for example, if the machine does not include any sensor capable of determining the voltage level of the machine. As a second example, voltage measurement may also be unavailable if a voltage sensor is present but is undergoing damage and cannot collect and / or provide voltage measurement data. As a third example, voltage measurement may also be unavailable if harmonics are present in the voltage. Voltage measurement may also be unavailable in a machine for many additional reasons.

[0043] In some embodiments, if a voltage measurement is determined to be available from the machine by the first condition of operation 106, flowchart 100 may proceed to operation 108, which may involve determining an integer number of cycles of one or more line currents and / or phase voltages. This integer number of cycles (in the waveforms of these values) may be determined based on the zero-crossing of Ia, where Ia may refer to the phase of the current, and the zero-crossing may refer to the zero-crossing point of the waveform of Ia. Using an integer number of cycles of the acquired current or voltage can prevent spectral leakage that could interfere with the detection of a particular fault frequency and can provide a better signal-to-noise ratio when compared to a baseline. From operation 108, flowchart 100 may proceed to operation 110, which may include estimating the fundamental frequency of the machine. The fundamental frequency may be the basic operating frequency used for the machine. The fundamental frequency can be determined from the stator current of the machine using the frequency of the maximum amplitude in the stator current spectrum, or it can be determined from a current and / or voltage phase-locked loop (PLL). That is, the PLL can operate on the signal to identify the fundamental frequency and the total duration. For example, if the fundamental frequency is known to be 50 Hz, it can be determined what time is required for one cycle. Once the machine's base frequency and synchronization speed are known, flowchart 100 can proceed to operation 111, which may involve determining the machine's synchronization speed at the determined base frequency. For example, the machine's synchronization speed at the base frequency can be determined using the equations described above. That is, the synchronization speed can be a function of the machine's base frequency and pole number. Flowchart 100 can then proceed to operation 112, which may include a second condition. The second condition may involve determining whether the machine is a synchronous or inductive machine, which can be determined by the machine's nameplate. If the machine is determined to be an inductive machine, flowchart 100 can proceed to operation 114. If the machine is determined to be a synchronous machine, flowchart 100 can proceed to operation 124. At operation 114, the machine's input power (e.g., one or more of the equations presented above) can be used to estimate the machine's motor speed. At operation 124, the motor speed can be determined to be the same as the machine's synchronization speed, since the machine is determined to be a synchronous machine. After operations 114 and / or 124, flowchart 100 can proceed to... Figure 3 Operation 302 of the flowchart 300 depicted in the figure.

[0044] continue Figure 1Following flowchart 100, if it is determined at operation 106 that a voltage measurement from the machine is unavailable, flowchart 100 can proceed to operations 116-122, which may include operations similar to operations 108-112 of flowchart 100, which can be performed if a voltage measurement is determined to be available. That is, operation 116 may involve calculating an integer number of cycles of line current and / or phase voltage. Operation 118 may involve estimating the machine's fundamental frequency. Operation 120 may involve determining the machine's synchronous speed at the fundamental frequency. Additionally, operation 122 may include a third condition identical to the first condition of operation 112. That is, the third condition may involve determining whether the machine is an inductive machine or a synchronous machine. If the machine is determined to be a synchronous machine, flowchart 100 can again proceed to operation 124, where the motor speed can be determined as the machine's synchronous speed. However, if the machine is determined to be an inductive machine, flowchart 100 can proceed to... Figure 2 Operation 202 within the flowchart 200 depicted in the diagram (which may be...) Figure 1 (Continued from flowchart 100). Therefore, the determination of the estimated motor speed of the induction machine when voltage measurement is available from the induction machine can differ from the determination of the estimated motor speed of the induction machine when voltage measurement is not available from the induction machine.

[0045] Go to Figure 2Flowchart 200 may include operations for determining an estimated motor speed of the sensing machine when voltage measurements are unavailable. Flowchart 200 may begin with operation 202. Operation 202 may include determining whether alternative speed estimation is enabled in the machine. Alternative speed estimation may be established during machine commissioning and may be based on whether a particular sensor is available for data collection. If it is determined that alternative speed estimation is not enabled, flowchart 200 may proceed to operation 210. Otherwise, flowchart 200 may proceed to operation 204. Operation 204 may include determining whether the machine has inherent asymmetry. For example, a machine may have inherent asymmetry because its manufacturing tolerances may not be perfectly balanced. A machine with inherent asymmetry may have a significant peak in the stator current spectrum at or near the machine's rotational frequency. It can be determined whether the machine is asymmetrical, for example, by determining that the machine is not operating in constant V / F mode. When a machine operates in constant V / F mode, the machine always maintains a certain V / Hz ratio. This can differ from, for example, vector control mode, in which voltage and frequency are manipulated to produce an optimal V / Hz ratio for the maximum torque of the machine. Whether the machine is operating in a constant V / F mode can be known, as it can be established during commissioning or installation. If it is determined that the machine is configured to use inherent asymmetries for speed estimation, flowchart 200 can proceed to operation 212. At operation 212, a frequency domain transformation of the machine's eccentricity fault signal can be performed. Any machine asymmetry can produce a distinct peak at the eccentricity frequency, which can correspond to the machine speed in Hz in the stator current spectrum. For example, the frequency domain transformation can be based on Fast Fourier Transform (FFT), Discrete Fourier Transform (DFT), or Short-Time Fourier Transform (STFT), as well as other frequency domain transformations. Flowchart 200 can then proceed to operation 214, which can involve determining the lowest and highest rotational frequencies for the machine's operating fundamental frequency. For example, this can be performed based on the machine's fundamental frequency and the machine's nameplate sliding speed. The lowest frequency can correspond to the minimum rotational speed under maximum machine load, and the highest frequency can correspond to the machine's synchronous speed. The lowest and highest rotational frequencies can then be used to establish a range of rotational frequencies for the machine. Once the lowest and highest rotational frequencies are determined, flowchart 200 can proceed to operation 216, which may involve determining the conditions for the presence of a sharp peak in the frequency domain transformation of the eccentricity fault signal within the rotational frequency range determined in operation 214. A sharp peak can be determined based on the current amplitude at the estimated motor speed being greater than the current amplitude at other frequencies. In some cases, the current amplitude may also need to be higher than the noise floor to be considered a sharp peak. That is, a peak can be a multiple of other current amplitudes that are considered peaks (e.g., an amplitude that is 2 to 3 times the amplitude of other current amplitudes can constitute a peak).However, the amplitude of 2 to 3 times can be arbitrarily chosen, and any other multiple can be used. If a clear peak is determined to exist in operation 216, flowchart 200 proceeds to operation 218. Otherwise, flowchart 200 proceeds to operation 206. At operation 218, the machine's rotational frequency (e.g., an estimate of the motor speed) can be determined as the frequency at which the clear peak is identified.

[0046] continue Figure 2 If it is determined at operation 204 that the machine is not configured to use inherent machine asymmetry for speed estimation, flowchart 200 can proceed to operation 206. At operation 206, it can be determined whether torque current-based speed estimation is enabled. If torque current-based speed estimation is enabled, flowchart 200 can proceed to operation 208. Torque current-based speed estimation can be enabled during commissioning and can be based on the availability of certain sensors. Otherwise, flowchart 200 can proceed to operation 210. At operation 208, the motor speed can then be estimated from the torque component of the current. For example, the speed estimate can be determined using equations 13-18 presented above. At operation 210, the motor speed can be estimated based on the assumption that the machine's slip is half of the machine's rated slip. Once the machine speed has been estimated through operations 208, 210, and / or 218, flowchart 200 can proceed to flowchart 300.

[0047] Go to Figure 3 In some embodiments, once a speed estimate for the machine is determined, for example, via flowcharts 100 and 200, the operation can proceed to flowchart 300. For example, flowchart 300 can be used to establish fault indicators for one or more different types of failure modes for the machine (examples of failure modes may include eccentricity, broken rotor bars, or bearing failure failure modes). A fault indicator corresponding to a specific failure mode can be defined as the square root of the energy or sum of squares of frequency points in the frequency domain at both sides of a peak detected in the frequency domain transformation of a specific fault signal and in the band (which may be defined below). Depending on the application and failure mode, the energy may or may not be normalized. Baselines for the machine for different types of failure modes can then be established and used to determine whether an electrical fault exists in the machine (e.g., based on the determined fault indicator baseline).

[0048] continue Figure 3Flowchart 300 may begin with operation 302, which may include determining a frequency domain transform of the fault signal corresponding to an individual failure mode. That is, different types of machine failure modes may be associated with different fault frequencies. Therefore, using these systems and methods, different fault indices and different baselines can be established for individual types of failure modes. In some cases, the frequency domain transform may be based on Fast Fourier Transform (FFT), Discrete Fourier Transform (DFT), or Short-Time Fourier Transform (STFT), as well as other frequency domain transforms. At operation 304, the widths of the search band and the sum band may be adjusted based on information associated with the machine. The search band may be a frequency range that can be determined and used to search for fault frequencies. For example, fault frequencies may be identified as peak amplitudes within the search band's frequency range (e.g., as may be performed in operation 308 described below). The sum band may be a frequency range that is narrower than the search band (or, in some cases, wider). The sum band may represent a frequency range including a certain number of data points (e.g., frequency values) to the left and right of the determined fault frequency identified in the search band. If the data resolution is higher, more data points can be selected and banded, while if the data resolution is lower, fewer data points can be selected and banded.

[0049] In some embodiments, the information used to adjust the search band and / or the band width may include, for example, the machine type (e.g., synchronous or inductive machine, and other machine types) and the speed estimation method used (e.g., in...). Figure 1 and Figure 2(The different speed estimation methods described in flowcharts 100 and / or 200). As a first example, this adaptive adjustment of the search and sum band can be a function of the machine's rated slip frequency. As a second example, the adaptive adjustment can be a function of the frequency resolution of the frequency domain transform. As a third example, the adaptive adjustment can also involve learning an optimized search and sum band size for a machine design, and those same optimized search and sum band sizes can be reapplied for machines of similar designs. As a fourth example, when using a particular type of speed estimation, such as when using input power, the torque component of the current, or machine asymmetry, a narrower search band can be used (e.g., this could be a more accurate speed estimation), and a wider search band can be used based on speed estimation methods that may not involve these factors (e.g., this could be a less accurate speed estimation method). In some cases, the search and sum band can be narrowed because if the accuracy of the machine's speed estimation is more reliable, it is more likely that the fault frequency can be located near a certain speed of the machine in the spectrum. Therefore, a narrower search band can be used to identify the fault frequency. However, if a less precise method is used to determine the velocity estimate, a larger search band may be necessary because the fault frequency may potentially be further away from the estimated velocity of the machine in the spectrum. Specific examples of search bands that can be used in different instances can be found in... Figures 4A-4C Presented in the middle.

[0050] In some embodiments, once the search band and / or band are established and / or adjusted, at operation 306, the fault frequency corresponding to the failure mode of the fault signal can be determined. The fault frequencies for different types of failure modes can be derived from estimated motor speeds, nameplate details, and machine geometry (e.g., bearing dimensions used to estimate bearing failure frequencies). Example formulas for deriving fault frequencies for eccentricity, broken rotor bars, and bearing failures are given below. The fault frequency for eccentricity in the stator current FFT can be calculated as:

[0051] f ecc＝ f s + / -kf r (Equation 15)

[0052] Where f s It can be the base frequency and f r The rotational speed can be estimated in Hz, and kay represents the harmonic order. In the square of the stator current spectrum, the eccentric characteristic frequency can be f. ecc＝ kf r (Equation 16)

[0053] For different bearing defects, the failure frequency of bearing faults used in stator current FFT depends on the bearing geometry and rotational frequency (for a single point defect in the bearing) and can be given as:

[0054] f bearing＝ f s + / -kf c (Equation 17)

[0055] Where the inner circle f0, the outer circle f i and ball defect fb of f c It can be calculated as follows:

[0056] Outer roller track:

[0057] Inner raceway:

[0058] ball:

[0059] Where Nb can be the number of rolling elements, Dc can be the cage diameter, Db can be the diameter of the rolling element (ball), and f r It can be the rotational speed in Hz. Typically, the contact angle β can be assumed to be zero.

[0060] In the square of the stator current spectrum, the bearing characteristic frequency can be f. bearing＝ kf c (Equation 21)

[0061] The fault frequency of the broken rotor bar used in the stator current FFT can be calculated as follows:

[0062] f brb＝ f s + / -2ksf s (Equation 22)

[0063] Where s can be the machine's sliding speed determined from the machine's operating speed, and sf s This can be the electric slip frequency of the machine, measured in Hz. The characteristic frequency of the broken rotor bar can be determined as f in the square of the stator current spectrum. brb＝ 2ksf s (Equation 23)

[0064] continue Figure 3At operation 308, one or more peaks can be identified in the search band. In some cases, the peak can be the maximum amplitude within the defined search band. In some cases, the peak can be located around the fault frequency in the frequency domain transformation of the fault signal. The peak can be a multiple of other current amplitudes considered to be peaks (e.g., an amplitude that is 2 to 3 times the amplitude of other currents can constitute a peak. However, 2 to 3 times the amplitude can be arbitrarily chosen, and any other multiple can be used). At operation 310, one or more fault indicators can be determined for a failure mode as the energy on both sides and in the band of the determined peak of the frequency domain fault signal. At operation 312, baselines of fault indicators corresponding to certain failure modes can be established, and these baselines can be used to identify deviations above a threshold as faults. That is, during the commissioning of the machine, baselines of fault indicators corresponding to different failure modes can be established. Thresholds can be calculated for each failure mode based on statistics of data collected in the baselines and additional adjustable pre-configured conditions. Once baselineization is complete with a minimum number of points captured, the system can be configured to compare a moving average of the failure metric points to a calculated threshold, and generate a failure alarm when the failure metric exceeds a defined threshold. For example, the baseline may include one or more defined failure metrics (note that this document may refer to "failure metrics," but multiple failure metrics may be defined for a failure mode to establish a baseline), and the threshold may be based on the average, standard deviation, or any other statistical measure of the baseline failure metrics. Then, if, during the machine's real-time operation, the average of the captured data is higher than the average of the baseline failure metrics by a threshold amount for a specific type of failure mode, a failure indicating that specific type of failure mode may have occurred. For example, if a metric corresponding to the failure frequency of a bearing failure is determined to be above a baseline threshold, it can be determined that a bearing failure may have occurred. This can be done in... Figure 7 Further explanation is provided below. Figure 7 A first plotting plot 702 may be drawn, including a baseline 704, which is established by one or more failure indicators 706 of a type of failure mode determined using the systems and methods described herein. Figure 7 A second plot 710 may also be drawn to indicate the occurrence of a fault condition. The second plot 710 may show one or more data points 712 that can be captured during the real-time operation of the machine. During the time period, the average value 715 of the one or more data points 712 is depicted as being a specific amount greater than the baseline 704. Based on this, it can be determined that a fault of a specific type of failure mode has occurred.

[0065] It should be noted that, Figures 1-3The operations described and depicted in the illustrative flowcharts can be performed or carried out in any suitable order as desired in the various exemplary embodiments of this disclosure. Additionally, in some exemplary embodiments, at least a portion of the operations can be performed in parallel. Furthermore, in some exemplary embodiments, more operations can be performed than... Figure 1-3 The operations described in the text are fewer, more, or different operations.

[0066] Figure 1-3 One or more operations of the process flow can be described above as being performed by a user device, or more specifically, by one or more program modules, applications, etc., executing on the device. However, it should be understood that Figure 1-3 Any operation in the process flow can be performed, at least in part, by one or more other devices in a distributed manner, or more particularly, by one or more program modules, applications, etc., executing on such devices. Furthermore, it should be understood that processing performed in response to the execution of computer-executable instructions provided as part of an application, program module, etc., can be interchangeably described herein as being performed by the application or program module itself or by the device on which the application, program module, etc., is executed.

[0067] In some embodiments, Figures 4A-4C Example searches and bands are described according to one or more example embodiments of this disclosure. In some embodiments, Figure 4A Examples of search bands that can be depicted to determine the frequency of failures for a specific type of failure mode. Figure 4AExample plot 400 can be depicted. Plot 400 can be a plot of the frequency domain transform of a signal from the machine. That is, plot 400 can describe the IaSqFFT of the signal (which can be the square of the fast Fourier transform of the stator current) as a function of frequency in Hertz. Plot 400 can include a frequency range 401, which can be the operating range of the machine. For example, the frequency range can include a lower operating frequency 402 and an upper operating frequency 403. The lower operating frequency 402 can include the minimum rotational frequency of the machine at maximum load. The upper operating frequency 403 can be the maximum rotational frequency of the machine at no load. The upper operating frequency 403 can also be the synchronous speed of the machine. In some cases, the lower operating frequency 402 and the upper operating frequency 403 can be determined by multiplying the rated slip speed of the machine by a factor. The lower operating frequency 402 can be extended to the left by an amount corresponding to the resulting value, while the upper operating frequency 403 can be extended to the right by an amount corresponding to the resulting value. That is, the lower operating frequency 402 and the upper operating frequency 403 can be extended outwards from the estimated speed of the machine by amounts corresponding to the calculated result. The plot 400 can also depict a search band 404, which can be a frequency range within the machine's frequency range 401. The range of the search band 404 can be from the lower search band frequency 405 to the upper search band frequency 406, and can be used to identify peak amplitudes. For example, it can be based on a description of... Figure 3 The adjustments performed in flowchart 300 determine the search band, or the search band can be determined in any number of other ways.

[0068] In some embodiments, Figure 4B Example plotting 420 can be depicted. Example plotting 420 can be identical to plotting 400, to illustrate how to... Figure 4A The search band 404 established in the plotting graph 400 is defined. That is, the plotting graph 400 may include the same operating frequency range 401 and may include the same search band 404, including the same lower search band frequency 405 and upper search band frequency 406. The plotting graph 420 may also include a depiction of the search band 404 and band 408 within the machine. Band 408 may include a lower band frequency 409 and an upper band frequency 410. In some instances, the lower band frequency 409 and the upper band frequency 410 may be determined to be located at a predetermined distance from the estimated speed of the machine, which may be represented as fr. Indicators may indicate the frequency values ​​in the plotting graph.

[0069] In some embodiments, Figure 4C Example plotting 430 can be depicted. Example plotting 430 can be compared with... Figure 4A plotting 400 and Figure 4B The plot 420 is similar in depiction to the example search bands 404 and 408. However, Figure 4CThe plot 430 may differ in that the search band 404 may be depicted as wider in plot 430 than in plot 400 or plot 420. Figure 4C This could be an example of a wider search band that could be used if a less precise estimate of the machine speed is employed. For instance, when using a less precise method for estimating motor speed (e.g., when the speed estimate is not determined using input power, the torque component of the current, or machine asymmetry), a wider search band could be used. Figure 4C The wider search band 404 is depicted in the diagram. Similarly, when using these more precise velocity estimation methods, Figures 4A-4B The text describes a narrower search band, 404.

[0070] Explanatory methods

[0071] Figure 5 This is an example method 500 according to one or more example embodiments of this disclosure. Figure 5 At block 502 of method 500, the method may include determining the rotational speed of a rotating machine. Block 504 of method 500 may include determining a frequency domain signal using a frequency domain transformation of a signal from the rotating machine. Block 506 of method 500 may include determining a first frequency band within the frequency domain signal for identifying fault frequencies of the rotating machine based on the rotational speed of the rotating machine. Block 508 of method 500 may include determining a fault frequency of the rotating machine within the first frequency band. Block 510 of method 500 may include determining a second frequency band within the first frequency band based on the fault frequency, wherein the second frequency band includes the fault frequency. Block 512 of method 500 may include determining a first fault indicator and a baseline of the first fault indicator based on the second frequency band. Block 514 of method 500 may include determining a fault condition of the rotating machine based on the deviation of the second fault indicator from the baseline. Block 516 of the method may include providing an alarm based on the fault condition of the rotating machine.

[0072] exist Figure 5 The operations described and depicted in the illustrative process flow can be performed or carried out in any suitable order as desired in the various exemplary embodiments of this disclosure. Additionally, in some exemplary embodiments, at least a portion of the operations can be performed in parallel. Furthermore, in some exemplary embodiments, more operations can be performed than... Figure 5 The operations described herein are fewer, more, or different operations.

[0073] Figure 5 One or more operations of the process flow can be described above as being performed by a user device, or more specifically, by one or more program modules, applications, etc., executing on the device. However, it should be understood that Figure 5Any operation in the process flow can be performed, at least in part, by one or more other means (e.g., locally on a machine, on a remote system such as a cloud system, etc.) in a distributed manner, or more particularly, by one or more program modules, applications, etc., executing on such means. Furthermore, it should be understood that processing performed in response to the execution of computer-executable instructions provided as part of an application, program module, etc., is interchangeably described herein as being performed by the application or program module itself or by the means on which the application, program module, etc., is executed.

[0074] Explanatory System Architecture

[0075] Figure 6 The illustration shows an example computing device 600 according to one or more embodiments of the present disclosure. The computing device 600 may represent any operation that can be used to perform the operations described herein (e.g., regarding...). Figure 1-3 and Figure 5The computing device 600 may be a means for performing any operations described in the flowcharts depicted herein, or any other operations associated with anomaly detection. For example, computing device 600 may be a digital signal processing (DSP) device, a relay, or any other type of means suitable for performing the operations described herein. Computing device 600 may also be located locally on a machine or may be remote (e.g., a remote server). Computing device 600 may include at least one processor 602 that executes instructions stored in one or more memory devices (referred to as memory 604). The instructions may be, for example, instructions for implementing functionality described as being performed by one or more modules and systems disclosed above, or instructions for implementing one or more methods disclosed above. The processor(s) 602 may be embodied, for example, in a CPU, multiple CPUs, GPUs, multiple GPUs, TPUs, multiple TPUs, multi-core processors, combinations thereof, etc. In some embodiments, the processor(s) 602 may be arranged in a single processing device. In other embodiments, the processor(s) 602 may be distributed across two or more processing devices (e.g., multiple CPUs; multiple GPUs; combinations thereof; etc.). A processor may be implemented as a combination of processing circuitry systems or computing processing units (such as CPUs, GPUs, or combinations of both). Therefore, for the sake of illustration, a processor can refer to a single-core processor; a single processor with software multithreading capabilities; a multi-core processor; a multi-core processor with software multithreading capabilities; a multi-core processor employing hardware multithreading technology; a parallel processing (or computing) platform; and a parallel computing platform with distributed shared memory. Additionally, or as another example, a processor can refer to an integrated circuit (IC), an ASIC, a digital signal processor (DSP), an FPGA, a PLC, a complex programmable logic device (CPLD), discrete gate or transistor logic, discrete hardware components, or any combination thereof designed or otherwise configured (e.g., manufactured) to perform the functions described herein.

[0076] One or more processors 602 can access memory 604 through a communication architecture 606 (e.g., a system bus). The communication architecture 606 may be adapted to a specific arrangement (localized or distributed) and type of processors 602. In some embodiments, the communication architecture 606 may include one or more bus architectures, such as a memory bus or memory controller; a peripheral bus; an accelerated graphics port; a processor or local bus; combinations thereof, etc. For example, such architectures may include an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MCA) bus, an Enhanced ISA (EISA) bus, a Video Electronics Standards Association (VESA) local bus, an Accelerated Graphics Port (AGP) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express bus, a Personal Computer Memory Card International Association (PCMCIA) bus, a Universal Serial Bus (USB), etc.

[0077] The memory components or memory devices disclosed herein may be embodied in volatile or non-volatile memory, or may include both volatile and non-volatile memory. Furthermore, the memory components or memory devices may be removable or non-removable, and / or internal or external to a computing device or component. Examples of various types of non-transient storage media may include hard disk drives, zip drives, CD-ROMs, digital versatile discs (DVDs) or other optical storage devices, magnetic tape cassettes, magnetic tapes, disk storage devices or other magnetic storage devices, flash memory or other types of memory cards, cassette tapes, or any other non-transient medium suitable for retaining desired information and accessible by a computing device.

[0078] For illustration, non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) used as external cache memory. For illustration and not limitation, RAM is available in many forms, such as synchronous RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct Rambus RAM (DRRAM). The memory devices or memories disclosed in the operating or computing environments described herein are intended to include one or more of these and / or any other suitable types of memory.

[0079] Each computing device 600 may also include a mass storage device 608 accessible by one or more processors 602 via a communication architecture 606. The mass storage device 608 may include machine-accessible instructions (e.g., computer-readable instructions and / or computer-executable instructions). In some embodiments, the machine-accessible instructions may be encoded in the mass storage device 608 and may be arranged in components that can be built (e.g., linked and compiled) and retained in the mass storage device 608 in a computer-executable form, or arranged in one or more other machine-accessible non-transient storage media included in the computing device 600. Such components may embody or constitute one or more of the various modules disclosed herein. Such a module is illustrated as module 614.

[0080] The execution of module 614 by processor(s) 602, individually or in combination, can cause computing device 600 to perform any of the operations described herein (e.g., regarding...). Figure 1-3 and Figure 5 (Any operation described in the flowcharts depicted, or any other operation associated with anomaly detection).

[0081] Each computing device 600 may also include one or more input / output interface devices 610 (referred to as I / O interfaces 610) that may allow or otherwise facilitate communication between external devices and computing device 600. For example, I / O interfaces 610 may be used to receive data and / or instructions from and to external computing devices. Computing device 600 also includes one or more network interface devices 612 (referred to as network interfaces 612(one or more)) that may allow or otherwise facilitate functional coupling between computing device 600 and one or more external devices. Functionally coupling computing device 600 to external devices may include establishing a wired or wireless connection between computing device 600 and external devices. Such communication processing devices may process data according to the defined protocols of one or more radio technologies. Radio technologies may include, for example, 3G, LTE, LTE-Advanced, 5G, IEEE 802.11, IEEE 802.16, Bluetooth, ZigBee, Near Field Communication (NFC), etc.

[0082] As used herein, the terms “environment,” “system,” “unit,” “module,” “architecture,” “interface,” “component,” etc., refer to entities related to a computer or to an operating device having one or more defined functionalities. The terms “environment,” “system,” “module,” “component,” “architecture,” “interface,” and “unit” are used interchangeably and can collectively refer to functional elements. Such entities can be hardware, a combination of hardware and software, software, or software in execution. As an example, modules can be embodied in a process running on a processor, a processor, an object, an executable portion of software, an execution thread, a program, and / or a computing device. As another example, a software application executing on a computing device and the computing device can embody a module. As yet another example, one or more modules can reside within a process and / or an execution thread. Modules can reside on a single computing device or be distributed among two or more computing devices. As disclosed herein, modules can be executed from various computer-readable non-transient storage media on which various data structures are stored. Modules can communicate via local and / or remote processes based on signals (analog or digital) having one or more data packets (e.g., data from a component that interacts with a local system, another component in a distributed system, and / or interacts with other systems via signals across a network such as a wide area network).

[0083] As yet another example, a module may be embodied in or may include a device having defined functionality provided by mechanical components operated by an electrical or electronic circuitry system controlled by a software or firmware application executed by a processor. Such a processor may be internal or external to the device and may execute at least a portion of the software or firmware application. Furthermore, in another example, a module may be embodied in or may include a device providing defined functionality through electronic components without mechanical components. The electronic components may include a processor executing software or firmware that at least partially allows or otherwise facilitates the functionality of the electronic components.

[0084] In some embodiments, modules may communicate via local and / or remote processes based on signals (analog or digital) having one or more data packets (e.g., data from a component that interacts with a local system, another component in a distributed system, and / or interacts with other systems via signals across a network such as a wide area network). Additionally, or in other embodiments, modules may communicate or be otherwise coupled via thermal, mechanical, electrical, and / or electromechanical coupling mechanisms (e.g., conduits, connectors, combinations thereof). Interfaces may include input / output (I / O) components and associated processors, applications, and / or other programming components.

[0085] Furthermore, in this specification and accompanying drawings, terms such as “storage,” “storage device,” “data storage,” “data storage apparatus,” “memory,” “repository,” and virtually any other information storage component relating to the operation and functionality of the components of this disclosure refer to a memory component, an entity embodied in one or more memory devices, or a component forming a memory device. Note that the memory components or memory devices described herein embody or include non-transitory computer storage media that can be read or otherwise accessed by a computing device. Such media can be implemented in any method or technique for storing information, such as machine-accessible instructions (e.g., computer-readable instructions), information structures, program modules, or other information objects.

[0086] Unless otherwise specified or otherwise understood in the context in which they are used, conditional language such as “can,” “able,” “may,” or “possibly” is generally intended to convey, among other things, that certain implementations may include certain features, elements, and / or operations, while other implementations do not. Therefore, such conditional language is generally not intended to imply that features, elements, and / or operations are required in any way for one or more implementations, or that one or more implementations must include logic for determining whether such features, elements, and / or operations are included in or to be performed in any particular implementation, with or without user input or prompting.

[0087] The contents described in this specification and accompanying drawings include examples of systems, apparatuses, technologies, and computer program products that, individually or in combination, allow for centralized AI-based topology processes for differential protection in power substations. It is certainly impossible to describe every conceivable combination of components and / or methods in order to describe the various elements of this disclosure, but it will be appreciated that many other combinations and arrangements of the disclosed elements are possible. Therefore, it may be understood that various modifications can be made to the invention without departing from its scope. Additionally, or as an alternative, other embodiments of this disclosure may be apparent in consideration of the practice of the invention as set forth herein and in the specification and drawings. It is intended that the embodiments set forth in the specification and drawings be considered illustrative rather than restrictive in all respects. Although specific terminology is used herein, it is used only in a general and descriptive sense and not for limiting purposes.

Claims

1. A system comprising: at least one processor; and at least one memory storing computer executable instructions that, when executed by the at least one processor, cause the at least one processor to: determine a rotational speed of a rotating machine; determine a frequency domain signal using a frequency domain transform of a signal of the rotating machine; determine a first frequency band within the frequency domain signal for identifying a fault frequency of the rotating machine based on the rotational speed of the rotating machine; determine a fault frequency of the rotating machine within the first frequency band (404); determine a second frequency band (408) within the first frequency band (404) based on the fault frequency, wherein the second frequency band (408) includes the fault frequency; determine a first fault indicator and a baseline (704) of the first fault indicator based on the second frequency band (408); determine a fault condition of the rotating machine based on a deviation of a second fault indicator from the baseline (704); and provide an alert based on the fault condition of the rotating machine, wherein the computer executable instructions further cause the at least one processor to: determine that a voltage signal for the rotating machine is not available; determine that the rotating machine is an induction machine; and determine the rotational speed of the rotating machine based on at least one of: a torque component of an input current of the rotating machine; a determined peak corresponding to an inherent asymmetry in a frequency domain transform of a fault signal of the rotating machine; and a synchronous speed of the rotating machine and a rated slip speed of the rotating machine.

2. The system of claim 1, wherein, Determining the rotational speed of the rotating machine is further based on at least one of: a type of rotating machine; nameplate details of the rotating machine; or a sampled input signal from the rotating machine.

3. The system of claim 1, wherein, The computer executable instructions further cause the at least one processor to: determine that a voltage measurement signal for the rotating machine is available; and determine that the rotating machine is an induction machine, wherein determining the rotational speed of the rotating machine is based on an input power of the rotating machine.

4. The system of claim 1, wherein, Determining the first frequency band or the second frequency band is further based on at least one of: the rotational speed of the machine; a type of speed estimation used; a frequency resolution of the frequency domain transform of the fault; or nameplate information associated with the machine.

5. The system of claim 1, wherein, The fault indicator is associated with a first type of failure mode of the machine, the first type of failure mode including at least one of: eccentricity; broken rotor bar; bearing fault; or any other mechanical, thermal, or electrical fault.

6. The system of claim 1, wherein, The baseline includes one or more fault indicators and their mean and standard deviation, and wherein determining the fault condition of the rotating machine further includes determining that a mean of fault indicator data received from the machine during operation of the machine is greater than a mean of baseline fault indicator values by a threshold amount.

7. A method comprising: determining a rotational speed of a rotating machine; determining a frequency domain signal using a frequency domain transform of a signal of the rotating machine; determining a first frequency band within the frequency domain signal for identifying a fault frequency of the rotating machine based on the rotational speed of the rotating machine; determining a fault frequency of the rotating machine within the first frequency band (404); determining a second frequency band (408) within the first frequency band (404) based on the fault frequency, wherein the second frequency band (408) includes the fault frequency; determining a first fault indicator and a baseline (704) of the first fault indicator based on the second frequency band (408); determining a fault condition of the rotating machine based on a deviation of a second fault indicator from the baseline (704); and providing an alert based on the fault condition of the rotating machine, wherein the method further comprises: determining that a voltage signal for the rotating machine is not available; and determining that the rotating machine is an induction machine, wherein determining the rotational speed of the rotating machine is based on at least one of: a torque component of an input current of the rotating machine; a determined peak corresponding to an inherent asymmetry in a frequency domain transform of a fault signal of the rotating machine; and a synchronous speed of the rotating machine and a rated slip speed of the rotating machine.

8. The method of claim 7, determining the rotational speed of the rotating machine is further based on at least one of: a type of the rotating machine; a nameplate detail of the rotating machine; or a sampled input signal from the rotating machine.

9. The method of claim 7, further comprising: determining that a voltage measurement signal for the rotating machine is available; and determining that the rotating machine is an induction machine, wherein determining the rotational speed of the rotating machine is based on an input power of the rotating machine.

10. The method of claim 7, wherein, determining the first frequency band or the second frequency band is further based on at least one of: the rotational speed of the machine; a type of speed estimation used; a frequency resolution of the frequency domain transform of the fault; or nameplate information associated with the machine.

11. The method of claim 7, wherein, the fault indicator is associated with a first type of failure mode of the machine, the first type of failure mode including at least one of: eccentricity; broken rotor bar; bearing fault; or any other mechanical, thermal, or electrical fault.

12. The method of claim 7, wherein, the baseline includes one or more fault indicators and their mean and standard deviation, and wherein determining the fault condition of the rotating machine further comprises determining that an average of fault indicator data received from the machine during operation of the machine is greater than an average of baseline fault indicator values by a threshold amount.

Images