Method of monitoring electric machine

Abstract

The invention relates to a method of monitoring an electric machine. The method comprises: a) acquiring a baseline of a monitored parameter of the motor (11) from a machine learning model of the motor (11) based on a set of measured parameter values from a sensor (13) arranged to measure a parameter of the motor (11) and a set of motor parameter values, a set of motor parameter values is from an electric drive (15) controlling the motor (11) and comprises electric drive configuration parameter values. B) acquiring a measurement of the monitoring parameter measured while controlling the motor (11) with the electric drive (15) using the same electric drive configuration parameter value as in step a), c) comparing the measurement of the monitoring parameter with the baseline, and d) determining the presence of a fault in the electric machine (11) if the measurement of the monitored parameter deviates from the baseline by more than a predetermined amount.

Description

Technical Field

[0001] This disclosure generally relates to electric motors, and more specifically to the monitoring of electric motor conditions. Background Technology

[0002] Electric motors, such as motors, can be condition monitored using motor models. This allows for the detection of unhealthy conditions. US 2021 / 0341901A1 discloses condition monitoring in industrial environments. An embedded analytics engine for motor drives monitors induction motor conditions for potential faults, including rotor and stator faults. The condition monitoring module is configured to acquire runtime signal data from a controller within the drive, derive runtime metrics from the runtime signal data based on induction motor fault conditions, provide the runtime metrics as input to a machine learning model, the machine learning model is constructed to identify the induction motor state based on the runtime metrics and output the state, and monitor induction motor fault conditions based on the induction motor state output by the machine learning model. Summary of the Invention

[0003] Electric motors can operate under various conditions, such as speed, load, and ambient temperature. When operating conditions change, existing solutions may fail to detect mechanical faults in certain situations. For example, a healthy bearing at a specific speed and load may produce a specific level of vibration at a root mean square (RMS) acceleration of X. If the bearing fails while the speed is reduced, an alarm may not be triggered if the new vibration level for the fault condition with a RMS acceleration of Y is lower than the healthy baseline (X>Y). Changes in load or operating temperature also affect the effectiveness of known methods.

[0004] In view of the above, the overall objective of the present invention is to provide a method for monitoring motors that solves or at least mitigates the problems of the prior art.

[0005] Therefore, according to a first aspect of this disclosure, a method for monitoring the condition of a motor is provided, the method comprising: a) obtaining a baseline of monitoring parameters of the motor from a machine learning model of the motor based on a set of motor parameter values ​​and a set of motor parameter values, the set of motor parameter values ​​being from sensors arranged to measure parameters of the motor, and the set of motor parameter values ​​being from an electric drive controlling the motor, including electric drive configuration parameter values; b) obtaining a measurement of the monitoring parameter while controlling the motor using the same electric drive configuration parameter values ​​as in step a); c) comparing the measurement of the monitored parameter with the baseline; and d) determining that a fault exists in the motor if the measured value of the monitored parameter deviates from the baseline by more than a predetermined amount.

[0006] Therefore, regardless of the motor's operating condition, it is possible to identify whether the behavior of the motor's mechanical components is abnormal (or unexpected) and whether a fault exists, as determined by the electric drive configuration parameter values. This makes motor condition monitoring more accurate.

[0007] The fault can be a mechanical fault or mechanical failure of the motor, such as bearing failure, misalignment, also known as motor deformed soft feet, that is, the motor is not evenly resting on all feet, and imbalance.

[0008] An electric motor can be an electric motor, such as an induction motor or a synchronous motor, or a generator.

[0009] The term "value" should be understood to include numerical values, Boolean values, or categorical values.

[0010] According to one embodiment, the set of electric drive configuration parameter values ​​includes at least one of the following: control type, modulation technique, switching pulse mode, switching frequency, PI gain, DC link control, and current offset.

[0011] The control type relates to how the switching mode of the electric drive is defined. It can typically be of many different types, such as scalar control or direct torque control (DTC).

[0012] According to one embodiment, the set of motor parameter values ​​includes at least one of the following: motor load and motor speed.

[0013] According to one embodiment, the machine learning model has been trained using the electric drive configuration parameter values ​​of that set of motor parameters. Therefore, the machine learning model takes into account the current values ​​of the electric drive configuration parameters when generating the baseline.

[0014] According to one embodiment, if the measurement of the monitored parameter deviates from the baseline by more than a predetermined amount, it is determined whether the set of electric drive configuration parameter values ​​is used to train the machine learning model, and if at least one value used for training differs from a value in the set of motor parameter values, a warning is issued: the determination in step d) may be potentially inaccurate. In cases where the machine learning model has not yet been trained with all possible values ​​of the electric drive configuration parameters, the method according to this example thus identifies this fact and provides a warning to the operator: the conclusion in step d) may be potentially inaccurate due to the possible influence of different electric drive configuration parameter values.

[0015] According to one embodiment, the warning includes presenting all electric drive configuration parameters that have values ​​different from the electric drive configuration parameter values ​​used for training.

[0016] According to one embodiment, the monitored parameter is one of the following: vibration, flux, internal motor temperature, and external motor temperature. The inventors have discovered that the electric drive configuration parameter values ​​directly affect torque fluctuations, and therefore influence vibration trajectories. Therefore, when vibration is the monitored parameter, considering the electric drive configuration parameter values ​​when assessing the presence of a fault in step d) is beneficial for achieving higher assessment accuracy. Furthermore, it has been found that drive configuration directly affects motor power losses, and therefore affects motor thermal performance.

[0017] According to one embodiment, the set of measured parameter values ​​includes measured parameter values ​​of motor speed and motor base temperature.

[0018] As an example, the measured parameters may include acceleration.

[0019] This set of motor parameter values ​​can include the motor's speed. The speed provided by the electric actuator in the set of motor parameter values ​​is typically much more accurate than the speed estimate / measurement from a sensor at a given speed (i.e., at low speeds). The difference in accuracy can be on the order of 100%. If the set of motor parameter values ​​includes the motor's speed, that speed value is used by a machine learning model to replace the speed provided by the sensor to generate a baseline.

[0020] One embodiment includes dynamically determining a predetermined amount based on a baseline.

[0021] According to one embodiment, determining the predetermined amount includes adding an offset to the baseline and / or subtracting the offset from the baseline.

[0022] One embodiment includes e) generating an alarm if a fault is determined to exist in step d).

[0023] According to one embodiment, the baseline reflects the health status of the motor for a set of motor parameter values.

[0024] According to a second aspect of the invention, there is provided computer code that, when executed by a processing circuit of a condition monitoring system, causes the condition monitoring system to perform the method of the first aspect.

[0025] According to a third aspect of this disclosure, a condition monitoring system for monitoring the condition of an electric motor is provided, the condition monitoring system comprising: a processing circuit device, and a storage medium including computer code according to the second aspect.

[0026] According to a fourth aspect of this disclosure, a motor assembly is provided, comprising: a motor, a plurality of sensors arranged to measure parameters of the motor, an electric actuator configured to control the motor, and a condition monitoring system according to a third aspect, the condition monitoring system being configured to acquire a set of measured parameter values ​​from the sensors and a set of motor parameter values ​​from the electric actuator.

[0027] Generally, all terms used in the claims will be interpreted according to their ordinary meaning in the technical field, unless otherwise expressly defined herein. Unless otherwise expressly stated, all references to “a / an / the element, device, component, apparatus, etc.” shall be openly interpreted as referring to at least one instance of the element, device, component, apparatus, etc. Attached Figure Description

[0028] Specific embodiments of the inventive concept will now be described by way of example with reference to the accompanying drawings, in which:

[0029] Figure 1 An example of a condition monitoring system for monitoring electric motors is shown schematically;

[0030] Figure 2 An example of a motor assembly is shown; and

[0031] Figure 3 Through Figure 1 The flowchart shows a method for a condition monitoring system to monitor the condition of an electric motor. Detailed Implementation

[0032] The concept of the invention will now be described more fully below with reference to the accompanying drawings, in which exemplary embodiments are shown. However, the concept of the invention may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided by way of example, so that this disclosure will be thorough and complete, and will fully convey the scope of the inventive concept to those skilled in the art. Throughout the specification, the same reference numerals denote the same elements.

[0033] Figure 1 A block diagram depicts an example of a condition monitoring system 1. Condition monitoring system 1 is configured to monitor the condition of a motor.

[0034] The condition monitoring system 1 includes an input unit 2, which is configured to receive a set of measured parameter values ​​from sensors arranged to measure the parameters of the motor. The input unit 2 is also configured to receive a set of motor parameter values, including electric drive configuration parameter values, from an electric drive that controls the motor.

[0035] Each value in this set of measured parameters is the value of the corresponding parameter detected or determined by the sensor.

[0036] Each value in this set of motor parameter values ​​is the current value of the corresponding parameter provided by the electric driver.

[0037] The condition monitoring system 1 can be configured to receive the set of measured parameter values ​​and the set of motor parameter values ​​from the sensor and the electric drive, respectively, via wireless, wired, or a combination of wireless and wired communication.

[0038] The condition monitoring system 1 includes a processing circuit 5, which is configured to receive the set of measured parameter values ​​and the set of motor parameter values ​​from the input unit 2.

[0039] The processing circuit 5 can be any combination of one or more suitable central processing units (CPUs), multiprocessors, microcontrollers, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), etc., which can perform any of the operations disclosed herein regarding the monitoring of the motor.

[0040] The condition monitoring system 1 may include a storage medium 7. The storage medium 7 may include a computer program that includes computer code, which, when executed by the processing circuitry 5, causes the condition monitoring system 1 to perform the methods disclosed herein.

[0041] The storage medium 7 may be implemented, for example, as a memory, such as random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or electrically erasable programmable read-only memory (EEPROM), and more specifically as a non-volatile storage medium of a device in external memory, such as USB (Universal Serial Bus) memory or flash memory, such as compact flash memory.

[0042] Condition monitoring system 1 can be, for example, a dedicated device locally located near the motor. Alternatively, condition monitoring system 1 can be part of a cloud, with sets of measured parameter values ​​and sets of motor parameter values ​​transmitted from sensors and electric actuators to the cloud for processing.

[0043] Figure 2 The motor assembly 9 is shown schematically. The motor assembly 9 includes a motor 11, such as an electric motor or a generator.

[0044] The motor assembly 9 includes multiple sensors arranged to measure parameters of the motor. The sensors may include one or more sensors 13 arranged to measure the temperature of the motor housing 11. Other sensors among the multiple sensors may include one or more accelerometers / accelerometers and / or speed sensors.

[0045] The motor assembly 9 also includes a condition monitoring system 1.

[0046] The motor assembly 9 also includes an electric drive 15 configured to control the motor 11. The electric drive 15 may be a variable speed drive.

[0047] The electric drive 15 is configured to send a set of motor parameter values ​​to the condition monitoring system 1. This set of motor parameter values ​​includes electric drive configuration parameter values. These electric drive configuration parameter values ​​reflect the current configuration of the electric drive. Electric drive configuration parameters may include, for example, control type, modulation technique, switching pulse mode, switching frequency, PI gain, DC link control, and current offset. Any remaining motor parameters that are not electric drive configuration parameters are parameters reflecting the current state of the motor, such as speed and / or load.

[0048] See Figure 3 The method for monitoring the condition of motor 11 through condition monitoring system 1 will now be described.

[0049] In step a), a baseline of the motor's monitoring parameters is obtained from the machine learning model of motor 11. The baseline is obtained from the machine learning model based on a set of measured parameter values ​​received from sensors and a set of motor parameter values, including electric drive configuration parameter values ​​received from electric drive 15. Therefore, the baseline is generated by the machine learning model based on this set of measured parameter values ​​and the set of motor parameter values, which serve as inputs to the machine learning model.

[0050] The monitored parameters can be, for example, vibration, flux, internal motor temperature, and external motor temperature. Therefore, the baseline can be an estimate of the vibration, flux in motor 11, internal motor temperature, or external motor temperature (i.e., the temperature on the outer surface of motor 11) generated by a machine learning model.

[0051] This baseline reflects the health status of motor 11 used for this set of motor parameter values.

[0052] The machine learning model can be trained by feeding a set of motor parameter values ​​from an electric drive, such as electric drive 15, preferably including electric drive configuration parameter values, and a set of measurement parameter values ​​from sensors.

[0053] In one example, the machine learning model has been trained using the same values ​​of the set of electric drive configuration parameters used to obtain the baseline in step a).

[0054] When training a machine learning model, the past behavior of motor 11 can be considered by adding filtered or moving average versions of motor parameter values ​​and measured parameter values.

[0055] In one example, RMS acceleration can be used to train a machine learning model. Other signals, such as vibration and / or velocity, can also be used.

[0056] In one example, the remaining network can be used to create the structure of a machine learning model.

[0057] In step b), the measured values ​​of the monitoring parameters are acquired. The monitoring parameters can be measured in real time from one of the sensors 13. The measurement of the monitoring parameters is performed while the electric drive 15 controls the motor 11 with the same electric drive configuration parameter values ​​as in step a). Therefore, for example, the same control type, the same PI gain, and the same switching frequency are used when the motor 11 is controlled and the set of measured parameter values ​​are measured by these sensors and provided to the machine learning model in step a) to generate a baseline, and when the measurement of the monitoring parameters is performed by one of these sensors and acquired in step b).

[0058] In step c), the measured value of the monitored parameter is compared to a baseline. Thus, for example, vibration or flux measured by one of the sensors is compared to a baseline. If the monitored parameter is vibration, the baseline obtained from the machine learning model is an estimate of the vibration; or if the monitored parameter is flux, the baseline obtained from the machine learning model is a measurement of the flux.

[0059] In step d), if the measured value of the monitored parameter deviates from the baseline by more than a predetermined amount, it is determined that there is a fault in the motor 11.

[0060] In one example, if the machine learning model has not yet been trained with all possible values ​​of the electric drive configuration parameters, and if the measured value of the monitored parameter deviates from the baseline by more than a predetermined amount, it is determined whether the set of electric drive configuration parameter values ​​used in step a) was used to train the machine learning model. If at least one value used for training differs from the electric drive configuration parameter values ​​in the set of motor parameter values ​​used in step a), the method may include producing an alert in step d) indicating that the determination may be potentially inaccurate. This is because at least one electric drive configuration parameter value differs from the value used for training. Therefore, even if the measured value of the monitored parameter deviates from the baseline by more than a predetermined amount, it is not necessarily possible to conclude that a fault exists in this case. Conversely, even if the measured value of the monitored parameter does not deviate from the baseline by more than a predetermined amount, it is not possible to conclude that a fault does not exist. In a variation of this example, the alert includes presenting all electric drive configuration parameters used in step a) that have electric drive parameter values ​​different from those used to train the machine learning model.

[0061] According to one example, a predetermined amount can be dynamically determined based on a baseline. For example, determining the predetermined amount may include adding an offset to the baseline and / or subtracting the offset from the baseline. In this way, dynamic upper and lower thresholds can be obtained, the magnitude of which varies depending on the value of the baseline. Alternatively, the permissible deviation of the measured value of the monitored parameter from the baseline can be static, and the actual deviation can be determined, for example, by subtracting the measured value of the monitored parameter from the baseline at each time step.

[0062] In one instance, step e) is performed optionally. Step e) includes generating an alarm if a fault is determined to exist in step d). The alarm can be visual, such as displayed on a monitor, or it can be an audible alarm.

[0063] In one example, step e) could include generating recommendations. For example, recommendations could be to reduce the speed or load, check the motor foundation, or check the bearings.

[0064] In one example, with permission granted, the electric drive 15 can be configured to control the motor to reduce vibration if the monitored parameter is vibration.

[0065] The concept of the invention has been described above with reference to several examples. However, it will be readily understood by those skilled in the art that other embodiments besides those disclosed above are also possible within the scope of the concept of the invention as defined by the appended claims.

Claims

1. A method for monitoring the condition of a motor (11), the method comprising: a) Based on a set of measured parameter values ​​and a set of motor parameter values, a baseline of monitored parameters of the motor (11) is obtained from a machine learning model of the motor (11), the set of measured parameter values ​​coming from sensors (13) arranged to measure the parameters of the motor (11), and the set of motor parameter values ​​including electric drive configuration parameter values ​​from an electric drive (15) controlling the motor (11). b) Obtain the measurement of the monitored parameters while using the electric drive (15) to control the motor (11) with the same electric drive configuration parameter values ​​as in step a). c) Compare the measurements of the monitored parameters with the baseline, and d) If the measurement of the monitored parameter deviates from the baseline by more than a predetermined amount, it is determined that there is a fault in the motor (11).

2. The method according to claim 1, wherein the set of electric driver configuration parameter values ​​is an electric driver configuration parameter value including at least one of the following: control type, modulation technique, switching pulse mode, switching frequency, PI gain, DC link control, and current offset.

3. The method according to any one of the preceding claims, wherein the set of motor parameter values ​​includes at least one of: the load of the motor (11) and the speed of the motor (11).

4. The method according to any one of the preceding claims, wherein the machine learning model has been trained using the set of electric drive configuration parameters.

5. The method according to any one of the preceding claims, wherein if the measurement of the monitored parameter deviates from the baseline by more than the predetermined amount, it is determined whether the set of electric drive configuration parameter values ​​is used to train the machine learning model, and if at least one value used for the training is different from the value in the set of electric drive configuration parameter values, warning: the determination in step d) may be potentially inaccurate.

6. The method of claim 5, wherein the warning comprises presenting all electric drive configuration parameters having values ​​different from the electric drive configuration parameter values ​​used for the training.

7. The method according to any one of the preceding claims, wherein the monitored parameter is one of the following: vibration, flux, internal motor temperature, and external motor temperature.

8. The method according to any one of the preceding claims, wherein the set of measurement parameter values ​​includes the speed of the motor (11) and the temperature of the motor base.

9. The method according to any one of the preceding claims, comprising dynamically determining the predetermined amount based on the baseline.

10. The method of claim 9, wherein the determination of the predetermined amount comprises adding an offset to the baseline and / or subtracting an offset from the baseline.

11. The method according to any one of the preceding claims, comprising: e) If a fault is identified in step d), generate an alarm.

12. The method according to any one of the preceding claims, wherein the baseline reflects the health status of the motor (11) for the set of motor parameters.

13. A computer code that, when executed by the processing circuitry (5) of a condition monitoring system (1), causes the condition monitoring system (1) to perform the method of any one of the preceding claims.

14. A condition monitoring system (1) for monitoring the condition of a motor (11), the condition monitoring system comprising: Processing circuit device (5), and Storage medium (7) including the computer code as described in claim 13.

15. A motor assembly (9), comprising: Motor (11) Multiple sensors (13) are arranged to measure the parameters of the motor (11). An electric drive (15) is configured to control the motor (11), and The condition monitoring system (1) according to claim 14 is configured to acquire a set of measurement parameter values ​​from the sensor (13) and a set of motor parameter values ​​from the electric driver (15).

Images