Robust extraction of motor failure frequency components under fluctuating operating conditions

Abstract

A fault detection system for extracting fault signatures of a motor operating under varying speed or varying load conditions is provided. The fault detection system includes a sensor interface configured to acquire sensor signals from sensors disposed at predetermined locations on the induction machine, and a memory connected to a processor, the sensor signals indicating a rotor eccentricity level of the induction machine. The instructions include steps, the sensor interface configured to acquire operating signals of the induction machine, and the memory configured to store a computer-implemented fault detection method for extracting fault signals in the frequency domain from a frequency spectrum formed by minimum variance beamforming.

Description

Technical Field

[0001] The present invention relates to a system for extracting a fault signature of a motor operating under variable speed or variable load conditions based on motor current signature analysis.

Background Art

[0002] Motor Current Signature Analysis (MCSA) has been widely used for the past several decades to detect motor faults such as bearing faults, eccentricity, and open circuit faults. When any of these motor faults occur, the rotating magnetic flux in the air gap becomes asymmetric, and as a result, extra frequency components are induced in the stator current. The fault detection method based on MCSA aims to extract the fault signature in the frequency domain by analyzing the stator current.

[0003] In practice, the extraction of the fault frequency component is very difficult for the following reasons. First, the motor fault frequency component is generally much weaker than the operating frequency component, especially in the initial stage of the occurrence of the motor fault. Second, the weak fault signature is easily buried in background noise or interference. For example, when the motor is driven by an inverter, its fault signature may be interfered by the harmonics of the power electronic device generated by the switching operation of the power electronic device. Third, the motor generally operates under variable load and variable speed conditions. The non-periodic time-varying factor inevitably causes spectral distortion of MCSA. Therefore, in order to effectively extract the fault signature from the noisy measurement values, it is desirable to develop a robust fault signature extraction method for the motor under variable operating conditions.

[0004] Frequency spectral analysis is a typical signal processing problem and is widely used in all kinds of applications. In fault detection based on MCSA, the most common spectral analysis method, due to its ease of use, is the Fourier transform. This method works fairly well when the test motor is operating in a steady state, but it does not work very satisfactorily in the case of load fluctuations. For load fluctuations, a simple approach is to measure multiple time sequences and take their average so that the effects of noise and fluctuations can be averaged. However, this method requires longer measurement times and may not be effective in extracting small fault signatures. To achieve high-resolution spectra, researchers have introduced other advanced signal processing methods for motor fault detection, such as ESPRIT, MUSIC, and compressed sensing (CS). However, these methods are sensitive to noise or heavily dependent on the signal model. Low noise levels or small load fluctuations occurring during the measurement period can impair the accuracy of fault detection.

[0005] Therefore, a robust method is needed to extract fault signatures from motors operating under fluctuating speed or fluctuating load conditions. [Overview of the project]

[0006] Motor current spectrum analysis (MCSA) is widely used for motor fault detection, including bearing failures, eccentricity, and open circuits. When a motor failure is in its early stages, or when a failed motor is operating under fluctuating load conditions, fault signatures can be buried in background noise and interference, making fault detection extremely difficult. Several embodiments of the present invention provide a method and system for extracting fault signatures with small frequency components under fluctuating load and high background noise conditions. To this end, the inventors divide time-domain current measurements into overlapping sequences and treat these sequences as measurements from a linear sensor array. Next, the inventors employ a minimum dispersion beamforming method for array signal processing to generate a current spectrum with robust performance under fluctuating load operation. The inventors demonstrated the method of the present invention using experimental data collected on a motor operating under practical fluctuating conditions and with a minor eccentricity fault.

[0007] Some embodiments of the present invention provide a minimum-dispersion-based spectral analysis method (minimum-dispersion beamforming method) for extracting frequency components of motor failures in the initial stages of changing operating conditions. The minimum-dispersion beamforming method minimizes the noise dispersion of the spectrum at each frequency. Therefore, in the final frequency spectrum, noise-related frequency components are suppressed, while the frequency components of the signal containing the fault signal are preserved. Experimental results demonstrate the effectiveness of the present invention in extracting fault signatures from noisy measurements.

[0008] The objective of some embodiments is to provide a system and method suitable for extracting fault signatures of motors operating under fluctuating speed or fluctuating load conditions. Motivated by array signal processing methodologies, the inventors propose performing motor current spectral analysis using a minimum dispersion beamforming method. The detailed idea is as follows: First, the inventors divide the time-domain stator current under test into multiple overlapping time sequences with a constant time shift. Each sequence is treated as an independent measurement of a virtual current sensor. Since the time difference between adjacent sequences is identical, these virtual sensors form a linear virtual sensor array. Next, the spectral analysis problem is transformed into a beamforming problem of a linear array in the frequency domain. The inventors borrow the idea of ​​minimum dispersion (MV) beamforming from array signal processing techniques to improve detection performance under noisy measurement conditions. Due to variations in motor operating conditions, each sensor has a gain coefficient that depends on a different unknown frequency. The MV-based spectral analysis aims to achieve a robust spectrum by using a weighted sum of the spectra of each time sequence. The weights are frequency-dependent and optimized by minimizing the noise variance in the output spectrum.

[0009] To verify the proposed method of the present invention, fault detection based on MCSA is performed using motor eccentricity as an example. Specifically, in the experiment of the present invention, a motor with a minor eccentric fault is used as the experimental subject, and a magnetic powder brake is attached as the motor load. Then, the input current of the magnetic powder brake is changed, and the stator current of the motor is measured under fluctuating load conditions. By comparing the MCSA results of experimental data obtained using different methods, it is demonstrated that the method of the present invention can effectively extract fault frequency components under fluctuating load conditions, even under strong noise interference.

[0010] According to some embodiments of the present invention, a fault detection system is provided for extracting fault signatures of induction motors operating under fluctuating load or fluctuating speed conditions. The fault detection system may include a sensor interface configured to acquire operating signals of an induction motor, a memory configured to store a computer-implemented fault detection method, and a digital signal processor configured to perform steps of a computer-implemented fault detection method using the operating signals acquired via the sensor interface. The steps of the computer-implemented fault detection method include dividing the operating signals into N segments based on a time window, the division being performed for each of the N segments such that it includes overlapping periods between adjacent divided operating signals, converting each of the divided operating signals into N frequency domains, shifting the phase of each of the divided operating signals by a preset phase, and performing a beamforming method on the converted divided operating signals, to which an optimized weighting coefficient is applied, and extracting fault signals in the frequency domain from the frequency spectrum formed by the minimum dispersion beamforming method.

[0011] Furthermore, some embodiments of the present invention provide a computer-implemented fault detection method for extracting fault signatures of induction machines operating under fluctuating load or fluctuating speed conditions. The method includes the steps of: acquiring an operating signal of an induction machine via a sensor interface connected to a current sensor located on the induction machine; dividing the operating signal into N segments based on a time window, the division being performed for each of the N segments such that it includes overlapping periods between adjacent divided operating signals; converting each of the divided operating signals into N frequency domains; shifting the phase of each of the divided operating signals by frequency and sequence time shift-related phases; and performing a minimum dispersion beamforming method on the converted divided operating signals, to which optimized weighting coefficients are applied; and extracting fault signals in the frequency domain from the frequency spectrum formed by the minimum dispersion beamforming method.

[0012] The accompanying drawings included to provide a further understanding of the present invention illustrate embodiments of the invention and, together with the description, illustrate the principles of the invention. The drawings shown are not necessarily drawn to a constant scale. Instead, the drawings may be exaggerated to illustrate the principles of the embodiments of this disclosure. [Brief explanation of the drawing]

[0013] [Figure 1A] This is a schematic diagram showing a fault detection system for controlling and monitoring an induction motor according to one embodiment of the present invention. [Figure 1B] This is a schematic diagram showing a fault detection system operated by a user (operator) via a network for controlling and monitoring an induction motor, according to one embodiment of the present invention. [Figure 2A] This figure shows a normal motor having a uniform air gap according to one embodiment of the present invention. [Figure 2B] This figure shows an eccentric motor having a non-uniform air gap according to one embodiment of the present invention. [Figure 3]This is an exemplary plot showing the stator current in the time domain. [Figure 4] A block diagram shows a fault detection system based on MCSA for detecting rotor failures according to one embodiment of the present invention. [Figure 5] This block diagram shows details of a robust spectral analysis method according to one embodiment of the present invention. [Figure 6A] This is an exemplary plot showing the stator current spectrum under one load condition, using the Fourier transform spectrum. [Figure 6B] This is an exemplary plot showing the stator current spectrum under one load condition, using the average spectrum. [Figure 6C] This is an exemplary plot showing the stator current spectrum under one load condition, using the minimum variance (MV) spectrum. [Figure 7A] This is an exemplary plot showing the stator current spectrum under different load conditions using the Fourier transform spectrum. [Figure 7B] This is an exemplary plot showing the stator current spectrum under different load conditions, using the average spectrum. [Figure 7C] This is an exemplary plot showing the stator current spectrum under different load conditions, using the minimum variance (MV) spectrum. [Figure 8A] This is an exemplary plot showing the stator current spectrum under different load conditions using the Fourier transform spectrum. [Figure 8B] This is an exemplary plot showing the stator current spectrum under different load conditions, using the average spectrum. [Figure 8C] This is an exemplary plot showing the stator current spectrum under different load conditions, using the minimum variance (MV) spectrum. [Modes for carrying out the invention]

[0014] Hereinafter, various embodiments of the present invention will be described while referring to the drawings. Note that the drawings are not drawn to scale, and in all the drawings, elements having the same structure or function are denoted by the same reference numerals. Also, the drawings are intended to facilitate the explanation of specific embodiments of the present invention. The drawings are not intended as an exhaustive description of the present invention or as a limitation on the scope of the present invention. Furthermore, the features described in relation to specific embodiments of the present invention are not necessarily limited to those embodiments and can be implemented in any other embodiment of the present invention.

[0015] FIG. 1A is a schematic diagram showing a fault detection system 100 for controlling and monitoring an induction motor according to an embodiment of the present invention.

[0016] The induction motor (system) 2*00* includes a rotor assembly 102, a stator assembly 104, a main shaft 106, and two main bearings 108. In this example, the induction motor 200 is a squirrel-cage induction motor.

[0017] The controller 220 is powered by a power supply 230 and can be used to monitor and control the operation of the induction motor 200 in response to various inputs according to an embodiment of the present invention. For example, the controller 220 connected to the induction motor 200 can obtain data on the operating state of the induction motor 200 from the current sensor 150 and control the speed of the induction motor based on the input received from the fault detection system 100 configured to obtain stator current data for the induction motor 200. For example, the current sensor detects current data from one or more of the plurality of phases of the induction motor. More specifically, when the induction motor is a three-phase induction motor, the current sensor detects current data from the three phases of the three-phase induction motor. Specific embodiments of the present invention are described with respect to polyphase induction motors, but other embodiments of the present invention can be applied to other polyphase electrical machines.

[0018] Some embodiments of the present invention describe a system for detecting faults in an electromechanical device such as an induction motor 200. The configured detection system includes a fault detection circuit module 100 for detecting faults of a rotor 102 including eccentricity faults, open circuit faults, and eccentricity faults within an induction motor assembly. In one embodiment, the fault detection circuit module 100 is implemented as a subsystem of a controller 220. In some cases, the fault detection circuit module 100 may be referred to as a fault detection system. In an alternative embodiment, the fault detection circuit module 100 is implemented using a separate processor. The fault detection circuit module 100 may be a hardware circuit module operably connected to the controller 220. In some implementation forms, the fault detection circuit module 100 and the controller 220 can share information. For example, the fault detection circuit module 100 can reuse sensor data used by the controller to control the operation of the induction motor.

[0019] Further, the fault detection circuit module 100 includes a processor 110, a memory 120, and a fault detection program (computer-implemented fault detection method) stored in the memory 120, and the instructions of the fault detection program are executed by the processor 110. The circuit module 100 further includes a sensor interface 130 configured to obtain signals from a sensor 150. The sensor interface 130 includes A / D (analog / digital) and A / D (analog / digital) converters for data communication with the processor 110, the memory 120, the fault detection program, the user interface 140, and the current sensor 150. The processor 110 is a digital signal processor configured to obtain a digital signal of the stator current (motor current) of the induction motor from the current sensor 150 via the sensor interface 130 and to execute the steps of the fault detection program.

[0020] The processor 110 may consist of multiple processors, and the memory 120 may be a memory module containing multiple memory modules. The user interface 140 is configured to connect to a keyboard and display unit configured to display normal / fault status information of the induction motor 200 in response to the output of the fault detection module 200.

[0021] Figure 1B is a schematic diagram showing a fault detection system operated by a user (operator) via a network 250 for controlling and monitoring an induction motor, according to one embodiment of the present invention. In this case, the fault detection module 100 may be included in the operating system of the induction motor deployed at the user / operator's site. If the fault detection module 100 determines that a fault has occurred based on sensor signals from a sensor 150 obtained via the network (communication network) 250 during the operation of the induction machine (induction motor) 200, the fault detection module 100 can decelerate or stop the drive of the induction motor 200 by transmitting a control signal via the network 250 based on a pre-programmed algorithm (not shown) stored in memory 120. Optionally, the current sensor 150 may be calibrated based on a sensor calibration program (not shown) stored in the memory of the fault detection module 100. Optionally, the network 250 may be a data communication network consisting of an optical fiber network, a wireless network, an internet network, or a combination of at least two of the optical fiber network, a wireless network, and an internet network.

[0022] This configuration can be a maintenance system managed by a user / client operating the induction system 20, which is located separately from the induction motor's location. For example, this system configuration may be used for a user-operated power generation system and a train system that controls the induction motors that drive the train. If the fault detection module 100 detects that the induction motor's location and the fault detection system's location are far apart, it performs data communication between the induction motor's location and the fault detection system's location via the network. The network 250 may be a data communication network consisting of an optical fiber network, a wireless network, an internet network, or a combination of at least two of the optical fiber network, a wireless network, and an internet network.

[0023] Furthermore, the fault detection system is included as part of the user's maintenance system. If the determined induction motor eccentricity level is above the critical threshold level, the fault detection system transmits a control signal to the induction motor controller via the network using the sensor interface / control interface 130, and stops the operation of the induction motor.

[0024] In one embodiment of the present invention, current and voltage sensors detect stator current data from the stator assembly 104 of the induction motor 200, respectively. Current data obtained from the current sensor is transmitted to an inverter for control and to a fault detection module for further processing and analysis. The analysis includes detecting a fault in the induction motor 200 by performing motor current signature analysis (MCSA). In some embodiments, when a fault is detected using the fault detection module 100, the controller 220 stops the operation of the induction motor by receiving a fault detection signal via the interface 130 of the fault detection module 100 and sending an interruption signal to the controller 110 to interrupt the stator current of the induction motor 200 for further inspection or repair. Optionally, the current sensor 150 includes a controller interface (not shown) configured to receive a fault detection signal from the interface 130 and send a fault condition signal to the controller 220, thereby allowing the controller 220 to stop the operation of the induction motor 200 by interrupting the stator current of the induction motor 200. When the current sensor 150 does not include a controller interface, the interface 130 may be configured to connect to the controller 220. In this configuration, the controller can stop the operation of the induction motor 200 by interrupting the stator current of the induction motor 200 in response to a fault detection signal from the fault detection circuit module 100 via the interface 130.

[0025] The system also includes memory for storing signal measurements and various parameters and coefficients in order to perform a fault severity detection method.

[0026] Figures 2A and 2B show, according to one embodiment of the present invention, a normal motor with a uniform air gap and a faulty motor with an eccentric fault indicated by a non-uniform air gap, respectively. Based on the physical model of the induction motor and a fault detection method using different features of the present invention, the present invention aims to extract fault signatures of motors operating under fluctuating speed or fluctuating load conditions.

[0027] Motor current signature analysis (MCSA) has been widely used for motor fault detection for decades due to its effectiveness and non-invasive nature. The MCSA method aims to extract characteristic frequency components of different types of faults based on the stator current frequency spectrum.

[0028] For example, if a wire break occurs in a squirrel-cage induction motor, a new set of frequency components will appear in the stator current spectrum, in addition to the operating frequency component.

number

[0029]

number

number

number

number

number

number

number

[0030]

number

[0031] Therefore, for most motor fault detection problems, the objective of MCSA-based methods is to extract the corresponding fault signature component through effective frequency spectrum analysis. When a fault frequency component exceeding a certain threshold is detected, it is determined that a corresponding fault exists. Depending on the amplitude of the fault frequency component and other operating conditions, the severity level of the fault can be further estimated.

[0032]

number

[0033]

number

[0034]

number

[0035]

number

[0036]

number

[0037] Figure 4 is a block diagram showing a computer-implemented fault signature / signal detection method / program 121 for detecting rotor faults based on MCSA, and Figure 5 is a block diagram detailing a robust spectral analysis method according to some embodiments of the present invention. The computer-implemented fault signature / signal detection method 121 is stored in memory 120 or a storage device (not shown) and is configured to be executed by the processor 110 when the system performs fault detection of the induction machine 200.

[0038]

number

[0039]

number

[0040]

number

[0041]

number

[0042]

number

[0043]

number

number

[0044] w is frequency-dependent; that is, a different w is obtained for each frequency by solving equation (25). Once the estimated spectrum S(w) 425 is obtained, fault signatures are extracted at different frequencies in 430 according to (1) to (8). In 440, if the amplitude of the fault signature is greater than a specific value, e.g., -70dB, compared to the operating signal of an early-stage fault, the computer-implemented fault detection method / program generates a fault signature code indicating that a fault (fault signature) has been detected. In this case, the fault signature code indicates one of the fault severity levels based on the amplitude of the fault signature. For example, the severity level is determined by three levels corresponding to normal level (<-70dB), warning level (≒-60dB), and severe (critical) level (>-40dB). The corresponding control signal 450 is then transmitted to the controller 220 to control the induction motor 200. If no fault is detected, the fault detection system 100 is configured to continuously receive / acquire the operating current (stator current 410) of the induction motor 200 from the current sensor 150 via the sensor interface 130 and to continuously monitor the stator current 410. While the fault detection system (circuit module) 100 is acquiring the operating current 410 from the current sensor 150 via the sensor interface 130, it generates state data for the severity level of the induction motor and continues to transmit it to the user interface 140, causing the severity level message information (state information) to be displayed on the display unit. If necessary, a control signal is sent to the controller 220 via the network 250 using a keyboard connected to the user interface 140, and the rotation speed of the induction motor 220 is controlled in response to the displayed severity level state information of the induction motor 220.

[0045] Furthermore, if the severity level of the induction machine is above the critical threshold level, the fault detection system sends a control signal to the controller 220 of the induction machine 200 via the network 250 to reduce the rotational speed of the induction machine 200 or to stop the operation of the induction machine 200.

[0046] To validate the proposed method, eccentric failures are considered as an example of analysis based on MCSA. The experiment is conducted in the laboratory using a motor with an eccentric failure. To induce rotor eccentricity, the two bearings supporting the rotor are removed and replaced with two external bearings fixed outside the motor. This allows the air gap to be manually adjusted to a predetermined range. Four gap sensors are placed on the stator to monitor the horizontal and vertical air gaps at both ends and to ensure the accuracy of the adjustment. A magnetic powder brake, whose torque can be adjusted by changing the input current, is used as the load. The experimental setup is shown in Figure 1A. During operation, the three-phase time-domain stator current is recorded and further analyzed.

[0047]

number

[0048]

number

[0049]

number

[0050] It is evident that FM spectra cannot detect fault signatures under varying load conditions. Averaging spectra, while sometimes effective at reducing noise to some extent, lack consistency. On the other hand, the MV spectra achieved by the proposed method consistently and effectively detect fault signature components (in this case, eccentric components at frequencies of 30 Hz and 90 Hz) for different fluctuating loads. In other words, the frequency range of the MV spectrum is set to show fault signatures at the eccentric components at frequencies of 30 Hz and 90 Hz.

[0051] The embodiments of the present disclosure described above may be implemented in many ways. For example, the embodiments may be implemented in hardware, software, or a combination thereof. If implemented in software, the software code may run on any suitable processor or group of processors, regardless of whether it is located on a single computer or distributed across multiple computers. Such processors may be implemented as integrated circuits. One integrated circuit element may contain one or more processors. However, processors may be implemented in any suitable circuit.

[0052] Embodiments of this disclosure may be embodied as methods provided as examples. The operations performed as part of this method may be ordered in any suitable manner. Thus, embodiments can be constructed that may include performing operations in a different order than those performed sequentially in exemplary embodiments, or performing several operations simultaneously.

[0053] The use of sequential terms in a claim to modify claim elements, such as first, second, etc., does not imply priority, precedence, or order of one claim element relative to another, or a time order in which the actions of the method are performed, but is simply used as a label to distinguish claim elements, to differentiate one claim element having a particular name from another element having the same name (by using sequential terms).

[0054] While the present invention has been described with reference to preferred embodiments, it should be understood that various other modifications and alterations can be made within the spirit and scope of the invention.

[0055] Therefore, the attached claims encompass all variations and modifications that fall within the true spirit and scope of the invention.

Claims

1. A fault detection system for extracting fault signatures of induction motors operating under fluctuating load or fluctuating speed conditions, A sensor interface configured to acquire the operating signal of the induction motor, A memory configured to store a computer implementation fault detection method, The system includes a digital signal processor configured to perform steps of the computer-implemented fault detection method using the operating signals acquired via the sensor interface, The aforementioned step is, The step includes dividing the operating signal into N segments based on a time window, wherein the division is performed for each of the N segments such that it includes overlapping periods between adjacent divided operating signals. The steps include converting each of the divided operating signals into N frequency domains, A step of shifting the phase of each of the divided operating signals by a predetermined phase, The step of performing a beamforming method on the converted segmented operating signals, wherein an optimized weighting coefficient is applied to each of the converted segmented operating signals. A fault detection system comprising the step of extracting a fault signal in the frequency domain from a frequency spectrum formed by a minimum dispersion beamforming method.

2. The fault detection system according to claim 1, wherein the sensor interface repeatedly acquires the operation signal over a predetermined period of time.

3. The aforementioned digital signal processor has a sampling rate f ｓａ The fault detection system according to claim 1, which collects the operating signal for 60 seconds at 10 kHz or at least twice the maximum frequency of the signal.

4. The fault detection system according to claim 1, wherein the operating signal is the stator current of the induction motor.

5. The fault detection system according to claim 1, wherein the extracted frequency spectrum is a minimum variance (MV) spectrum.

6. The fault detection system according to claim 5, wherein the frequency range of the MV spectrum is set to show the fault signature in eccentric frequency components of 30 Hz and 90 Hz.

7. The location of the induction motor is far from the location of the fault detection system. Data communication between the location of the induction motor and the location of the fault detection system is performed via a network. The fault detection system according to claim 1, wherein the network is a data communication network consisting of an optical fiber network, a wireless network, an internet network, or a combination of at least two of the optical fiber network, the wireless network, and the internet network.

8. The fault detection system according to claim 7, wherein if the severity level of the induction machine is above a critical threshold level, the fault detection system reduces the rotational speed of the induction machine or stops the operation of the induction machine by transmitting a control signal to the controller of the induction machine via the network.

9. A computer-implemented fault detection method for extracting fault signatures of induction motors operating under fluctuating load or fluctuating speed conditions, The steps include: acquiring an operating signal of the induction motor via a sensor interface connected to a current sensor located on the induction motor; The step of dividing the operating signal into N segments based on a time window, wherein the division is performed for each of the N segments such that it includes overlapping periods between adjacent divided operating signals. The steps include converting each of the divided operating signals into N frequency domains, Steps include shifting the phase of each of the divided operating signals in terms of frequency and sequence time shift-related phases, The steps include performing a minimum dispersion beamforming method on the converted segmented operating signals, wherein an optimized weighting coefficient is applied to each of the converted segmented operating signals. A computer-implemented fault detection method comprising the step of extracting a fault signal in the frequency domain from a frequency spectrum formed by a minimum dispersion beamforming method.

10. The computer-implemented fault detection method according to claim 9, wherein the sensor interface repeatedly acquires the operation signal over a predetermined period of time.

11. The above method uses a digital signal processor to achieve a sampling rate f ｓａ The computer-implemented fault detection method according to claim 9, comprising collecting data of the operating signal for 60 seconds at 10 kHz or at least twice the maximum frequency of the signal.

12. The computer-implemented fault detection method according to claim 9, wherein the operating signal is the stator current of the induction motor.

13. The computer-implemented fault detection method according to claim 9, wherein the extracted frequency spectrum is a minimum dispersion (MV) beamforming spectrum.

14. The computer-implemented fault detection method according to claim 13, wherein the frequency range of the MV spectrum is set to show the fault signature in eccentric frequency components of 30 Hz and 90 Hz.

Images