METHOD AND DEVICE FOR VERIFYING THE PLASIBILITIES OF A TEMPERATURE IN A COMPONENT OF AN ELECTRIC MACHINE IN A TECHNICAL SYSTEM

Description

Technical field

[0001] The invention relates to electric machines for electric drive systems, in particular methods for verifying the plausibility of a temperature measurement of a temperature sensor in a component of the electric machine. Technical background

[0002] During the operation of electric motors, current flows in the stator and / or rotor coils generate power losses, which lead to heating of the motor's components. The amount of heat generated depends, among other things, on the motor current flowing into the motor and the power dissipated within it. Overheating of the motor's components can damage or destroy coil components and can reduce the motor's performance due to the demagnetization of hard and soft magnetic fields.

[0003] The heat generation of an electric motor is therefore monitored during operation and overheating is avoided by limiting the motor torque, i.e. by limiting the current draw.

[0004] Conventional solutions involve arranging one or more temperature sensors, usually temperature-sensitive resistors such as NTC (Negative Thermal Coefficient) sensors, on stator coils. These measure a temperature at a specific point on the stator, which, however, does not usually correspond to a maximum or representative temperature in the electric motor, since the position where the maximum temperature occurs differs from the position(s) of the temperature sensor(s).

[0005] Temperatures in moving components of the electric motor, such as the rotor, are currently determined using physical modeling, which, however, generally cannot account for all thermal sources or sinks. Furthermore, the modeled heat loss terms must be determined or calibrated using very complex simulation and / or test bench measurements.

[0006] Electric motors, as electrical machines, are currently used both as drive motors and as actuators for vehicles, work machines and as servo motors.

[0007] Since determining the temperature of electrical machine components using a temperature sensor is essential to prevent damage or destruction of the machine, monitoring the proper functioning of the temperature sensor is necessary. Simple methods for validating the temperature sensor signal are already conventionally implemented, particularly by monitoring temperature changes over a specific period or by detecting implausibly high or low temperature values ​​for anomaly detection. Other methods for validating the temperature sensor reading involve monitoring the coolant temperature.

[0008] Monitoring a temperature sensor reading using a virtual sensor, which may be implemented with a complete physical temperature model, is very complex and requires a high level of computation, especially when machine learning methods are used.

[0009] Document US 7,908,893 B2 discloses a motor system comprising a motor with an excitable winding, a motor drive connected to the winding, a temperature measuring device connected to the motor drive, and a processor programmed to determine a temperature calibration parameter based on the temperature of the motor drive and a first resistance value of the winding. The processor determines a temperature value of the motor based on the calibration parameter and a second resistance value of the winding. Disclosure of the invention

[0010] According to the invention, a method for verifying the plausibility of a temperature measurement with a temperature sensor in a component of an electric motor according to claim 1, as well as a device and a motor system according to the dependent claims, are provided.

[0011] Further details are specified in the dependent claims.

[0012] According to a first aspect, a procedure for verifying the plausibility of a temperature measurement from a temperature sensor on a component of an electrical machine is provided, comprising the following steps: Monitoring a current in the electric machine; upon detection of steady-state operation, providing a first temperature measurement of the temperature of the electric machine component, the duration of steady-state operation, and an ambient temperature; acquiring a second temperature measurement at the end of steady-state operation; determining a modeled temperature value using a data-driven temperature model based on the first temperature measurement, the duration, and the ambient temperature, wherein the temperature model is trained to map the first temperature measurement, the duration, and the ambient temperature to a modeled temperature value; validating the second temperature measurement against the modeled temperature value.

[0013] Operation with a steady load is determined when a current in or out of the electrical machine is constant for a predetermined minimum duration and for a predetermined maximum duration, or deviates from a mean current during that duration by no more than a predetermined tolerance amount.

[0014] It may be possible to verify that the duration of the steady-state operation is sufficiently long to allow for the evaluation of a significant temperature difference for the detection of signal drift. In particular, if an implausible temperature value occurs, this can be signaled accordingly, especially by a visual or audible signal, and / or the operating mode of the electric machine can be adjusted accordingly by reducing power to prevent overheating of the electric machine component.

[0015] To monitor a temperature sensor integrated into a component of an electric machine, such as a stator or rotor, the above method proposes a temperature model applicable to a static operating state of the electric machine. During normal operation, the temperature sensor is continuously monitored, thus enabling timely power reduction or derating to prevent demagnetization of the electric machine components.

[0016] However, signal drift often occurs over the operating life of a temperature sensor, causing the sensor signal to become sluggish and leading to a discrepancy between the measured temperature and the actual temperature. The opposite case, where the temperature sensor signal changes more rapidly than the actual temperature, also occasionally occurs.

[0017] To validate a temperature measurement taken with a temperature sensor, it is proposed, according to the above procedure, to provide a temperature model that models the temperature behavior of a component of an electrical machine for a static operating condition. In particular, the temperature model can describe a temperature change over a period of time in a steady state, during which the electrical machine is either de-energized or subjected to a constant load.

[0018] The temperature model can be implemented as a data-driven model, for example, in a control unit, either as software or hardware. Based on an initial measured temperature value of the current component temperature at the beginning of a steady-state condition of the electric machine, the ambient temperature, and the duration of the steady-state condition, it models a temperature value at the end of that period. The temperature model can be trained using laboratory temperature measurements taken at the end of the steady-state period. Such a temperature model avoids the problem of accurately determining the temperature using a suitable physical model based on a time integration method with dynamic and highly dynamic load phases, as the time-delayed temperature profile is very difficult to model.Thus, the temperature model above only provides for modeling a temperature change or a temperature for the end of the steady state period for steady-state operating phases, especially in the case without current.

[0019] The application of such a simplified temperature model takes advantage of the fact that conventional operating modes of electric motors, whether used as drive or actuator motors, always include periods of standstill or constant load. These operating conditions are characterized by easily predictable behavior based on the model and can be evaluated according to the data-driven temperature model. For example, in the case of a drive motor for a motor vehicle, periods when the vehicle is stationary after an operating phase, such as at a traffic light or in a traffic jam, can always be used to validate the temperature sensor. These periods occur regularly during operation, allowing for correspondingly regular validation.

[0020] Even with an electric machine used as a servo motor in a machine, such downtime phases are frequent, so that here too, a signal drift of the temperature sensor can be detected early through regular plausibility checks.

[0021] Furthermore, the data-based temperature model can be designed to assign an electric current to the modeled temperature value at steady load, whereby the modeled temperature value is determined depending on the first temperature measurement, the duration, the ambient temperature and the constant electric current.

[0022] According to one embodiment, the detection of an implausible temperature reading can be signaled by an optical or acoustic signal and / or the operating mode of the electric machine can be adjusted accordingly by reducing power in order to avoid overheating of the component of the electric machine.

[0023] Alternatively, the temperature model can be calculated in a control unit for the electric machine in a technical device.

[0024] If a plausible second temperature measurement is found, the data-based temperature model can be further trained or retrained based on the first temperature measurement, the duration, the ambient temperature and the second temperature measurement.

[0025] The data-based temperature model can be designed as a probabilistic regression model, whereby further training is only carried out if, at a data point determined by the first temperature measurement, the duration and the ambient temperature, a confidence value is determined using the data-based temperature model that is lower than a predetermined confidence threshold.

[0026] The temperature model can therefore be calculated in a control unit for the electric machine. In particular, if a probabilistic regression model is used as the temperature model, further training to improve the temperature model can be carried out when a confidence value resulting from the evaluation of the probabilistic regression model for an electric machine falls below a confidence threshold.

[0027] According to another alternative, the temperature model can be calculated in a remote central unit that is in communication with a variety of technical devices and the electrical machines.

[0028] In particular, if a plausible second temperature measurement is detected in one of the many technical devices, the data-based temperature model can be further trained or retrained based on the first temperature measurement, the duration, the ambient temperature and the second temperature measurement.

[0029] It may be provided that the data-based temperature model is designed as a probabilistic regression model, whereby further training is only carried out if, at a data point determined by the first temperature measurement, the duration and the ambient temperature, a confidence value is determined using the data-based temperature model that is lower than a predetermined confidence threshold.

[0030] The temperature model can therefore also be implemented externally, for example in a cloud. In particular, when implemented in an external central unit (cloud), the temperature model can be further trained using additional operating points from a large number of similar technical systems with identical electrical machines (design). Specifically, if a probabilistic regression model is used, further training to improve the temperature model can be performed when a confidence value derived from the evaluation of the probabilistic regression model for an electrical machine falls below a certain confidence threshold. Brief description of the drawings

[0031] The embodiments are explained in more detail below with reference to the accompanying drawings. These show: Figure 1 is a schematic representation of a technical device with an electric machine and a control unit; Figure 2 is a schematic representation of an electric machine with a temperature sensor for temperature monitoring; Figure 3 is a flowchart illustrating a method for operating the technical device; Figure 4 is a diagram illustrating the function of the temperature model; and Figure 5 is a system with a plurality of similar technical devices and a central unit for carrying out the method for verifying the plausibility of a temperature measurement by a temperature sensor in an electric machine. Description of embodiments

[0032] Figure 1Figure 1 schematically shows a motor vehicle 1 as an example of a technical device with an electric machine 2 as its drive motor, which is operated via a control unit 3. The electric machine 2 contains a temperature sensor 4 to monitor its temperature. Using a plausibility check procedure implemented in the control unit 3, the measured temperature value of a component of the electric machine is compared to a modeled temperature value, and a corresponding error is signaled in case of a deviation.

[0033] Figure 2Figure 1 schematically shows a cross-sectional view through an electric machine 2 as part of a drive system. The electric machine 2 has a stator 21 and a rotor 22 mounted on a shaft, which represent the components of the electric machine. Furthermore, the stator 21 can be provided with stator coils 211, which can be electrically controlled via phase voltages and phase currents. In alternative embodiments, rotor coils, or both stator and rotor coils, can also be provided.

[0034] The electric machine 2 is controlled by the control unit 4, which provides for the control of the electric machine 2 by applying phase voltages to the stator coil 211 according to a commutation pattern. The control unit 4 can therefore also include a power driver circuit 5 in the form of a B6 bridge circuit or the like, in a manner known per se.

[0035] Furthermore, the rotor 22 can be coupled with a position sensor 23, which can detect the rotor position with respect to the arrangement of the stator coils 211.

[0036] The control unit 4 operates the electric machine 2 in a known manner by specifying the phase voltages in order to adjust certain phase currents so that a predetermined motor torque is achieved. The application of these phase currents dissipates power in the electric machine 2, which can lead to heating of its components. This heating occurs unevenly in the components.

[0037] The stator 21 has coolant channels 24 for conveying coolant. The coolant is conveyed through the coolant channels 24 at a flow velocity known in the control unit 3.

[0038] The control unit 4 can also be configured to operate a temperature model to monitor a measured temperature value from temperature sensor 4 in real time. Monitoring the function of temperature sensor 4 is necessary because a malfunction could lead to undetected overheating of components of the electric machine 2.

[0039] In Figure 3 A flowchart is shown that illustrates in more detail the procedure for validating a temperature measurement with a temperature sensor in an electric machine. This procedure can be used, for example, in control unit 3 of the [unclear text]. Figure 1 The vehicle shown serves as an example of the technical device being implemented.

[0040] In step S1, the motor current of the electric machine is continuously measured and monitored.

[0041] In step S2, it is checked whether the electric machine 2 is de-energized or operated with a constant load, i.e., a constant motor current. If this is the case (alternative: Yes), the procedure continues with step S3; otherwise (alternative: No), it returns to step S1.

[0042] In step S3, the system waits until a steady state is reached. The waiting time can be, for example, between 0 and 2 seconds, preferably 1 second.

[0043] In step S4, a temperature measurement is taken and a resulting first temperature measurement is temporarily stored.

[0044] Furthermore, in step S5, the ambient temperature is recorded and temporarily stored.

[0045] In a subsequent step S6, it is checked whether the steady-state operation has changed due to an interim change in the current draw of the electric machine or whether a predefined maximum time period has elapsed since the first temperature measurement was taken. If this is the case (alternative: Yes), the procedure continues with step S7; otherwise (alternative: No), it returns to step S6.

[0046] In step S7, a second temperature measurement is recorded using temperature sensor 4 in a second temperature measurement, and a time period between the recordings of the temperature measurements is determined as the duration of the steady phase and temporarily stored.

[0047] In step S8, a model evaluation is now performed using the temperature model, which is trained to assign the first temperature measurement, the ambient temperature and the duration of the steady phase to a modeled temperature value.

[0048] The temperature model comprises a data-driven model and can be implemented as a probabilistic regression model, for example, in the form of a Gaussian process model. Implementing it as a probabilistic regression model has the advantage that, in addition to the model output, a confidence value can also be provided to indicate the model's reliability.

[0049] An example of the functionality of the data-based temperature model is, for example, in Figure 4 The diagram illustrates the dependence of the modeled temperature value Tmod on the duration t of the steady state and the ambient temperature Tumg, using an exemplary initial temperature measurement T0 of 60°C. In further embodiments, the temperature model can also incorporate a constant load in the form of a motor current as an additional input dimension.

[0050] The modeled temperature value T mod is compared in a subsequent step S9 with the second temperature measurement T2 taken at the end of the steady-state operating period.

[0051] If a temperature difference, preferably exceeding a predefined tolerance threshold (e.g., 1°C), is detected (alternative: Yes), a corresponding error message is signaled in step S10. This can be communicated to the vehicle driver via a visual or audible signal, or a power reduction or limitation of the electric motor can be implemented.

[0052] If no temperature difference is detected (alternative: No), the procedure continues with step S11.

[0053] In step S11, the corresponding query of the temperature model, consisting of the first temperature measurement, the duration of steady-state operation, the ambient temperature, and the second temperature measurement at the end of the period, can be saved or used as a training dataset for further training of the data-driven temperature model. Specifically, it can be stipulated that the training dataset is only used for further training of the temperature model if the temperature model, at the relevant evaluation point, exhibits a confidence value below a predetermined confidence threshold based on the first temperature measurement, the duration of steady-state operation, and the ambient temperature. In this way, the temperature model can be successively improved.

[0054] In Figure 5As a further embodiment, a system is shown in which a plurality of vehicles 1 of a vehicle fleet 13 transmit the first and second temperature measurements, the duration, and the ambient temperature to a central unit 12 in the form of operating data F as soon as these are available from a measurement. The central unit 12 has a data processing unit 121 in which the evaluation of the temperature model can be carried out, and a database 122 for storing data points, model parameters, states, and the like.

[0055] The system serves to collect fleet data in the central unit 12, to validate a second temperature measurement, and to create the data-based temperature model. For this purpose, the relevant operating parameters from each of the vehicles 1 are transmitted to the central unit 12 when a steady state is detected. By querying the temperature model based on the transmission of the first temperature value, the duration of steady state operation, and the ambient temperature to the central unit 12, the corresponding modeled temperature value can be transmitted back to the respective vehicle, allowing the plausibility check of the second temperature measurement to be performed in the vehicle as described above.

[0056] Furthermore, if, as described above, the second temperature value is also transmitted to the central unit 12 at the end of the steady-state period, a corresponding plausibility check can also be performed in the central unit 12. If it is determined there that the corresponding temperature sensor in the transmitting vehicle is not functioning correctly, this can be signaled accordingly and communicated to the vehicle. However, if it is determined that the corresponding temperature sensor is functioning correctly, the recorded data set consisting of the first temperature value, the duration of steady-state operation, and the ambient temperature can be used as a training data set for further training of the temperature model in the central unit 12. In this way, a large number of similar devices or vehicles can be used for further training of the temperature model.

Claims

1. Method for checking the plausibility of a temperature measurement of a temperature sensor (4) on a component (21, 22) of an electric machine (2) in a technical device (1), wherein the method is characterized by the following steps: - monitoring (S1) a current in the electric machine (2); - when a steady-state load operation is detected, providing (S4) a first temperature measured value of a temperature of the component (21, 22) of the electric machine (2), the duration of the steady-state load operation and an ambient temperature; - recording (S7) a second temperature measured value at the end of the steady-state load operation; - determining (S8) a modelled temperature value (Tmod) using a data-based temperature model on the basis of the first temperature measured value, the duration and the ambient temperature, wherein the temperature model is trained to assign the first temperature measured value, the duration and the ambient temperature to a modelled temperature value (Tmod); - checking the plausibility (S9) of the second temperature measured value using the modelled temperature value, wherein steady-state load operation is detected when a current in or out of the electric machine (2) is constant for a predetermined minimum duration and for a predetermined maximum duration, or deviates by no more than a predetermined tolerance amount from an average value of the current during the duration.

2. Method according to Claim 1, wherein the data-based temperature model is furthermore formed to assign an electric current at steady-state load to the modelled temperature value (Tmod), wherein the modelled temperature value (Tmod) is determined on the basis of the first temperature measured value, the duration, the ambient temperature and the constant electric current.

3. Method according to either of Claims 1 and 2, wherein the detection of an implausible temperature measured value is signalled by a visual or audible signal and / or the operating mode of the electric machine (2) is adjusted accordingly by way of power reduction in order to avoid overheating of the component (21, 22) of the electric machine (2).

4. Method according to one of Claims 1 to 3, wherein the temperature model is calculated in a control device (3) for the electric machine (2) in a technical device (1).

5. Method according to Claim 4, wherein, when a plausible second temperature measured value is detected, the data-based temperature model is further trained or retrained based on the first temperature measured value, the duration, the ambient temperature and the second temperature measured value.

6. Method according to Claim 5, wherein the data-based temperature model is in the form of a probabilistic regression model, wherein the further training is carried out only if a confidence value that is lower than a predefined confidence threshold value is ascertained using the data-based temperature model at a data point determined by the first temperature measured value, the duration, and the ambient temperature.

7. Method according to one of Claims 1 to 3, wherein the temperature model is calculated in a central unit (12) that is remote from the device and that is communicatively connected to a plurality of technical devices (1) comprising the electric machines.

8. Method according to Claim 7, wherein, when a plausible second temperature measured value is detected in one of the plurality of technical devices (1), the data-based temperature model is further trained or retrained based on the first temperature measured value, the duration, the ambient temperature and the second temperature measured value.

9. Method according to Claim 8, wherein the data-based temperature model is in the form of a probabilistic regression model, wherein the further training is carried out only if a confidence value that is lower than a predefined confidence threshold value is ascertained using the data-based temperature model at a data point determined by the first temperature measured value, the duration, and the ambient temperature.

10. Apparatus for carrying out one of the methods according to one of Claims 1 to 9.

11. Computer program product comprising instructions that, when the program is executed by a computer, cause said computer to perform the steps of the method according to one of Claims 1 to 9.

12. Machine-readable storage medium comprising instructions that, when executed by a computer, cause said computer to perform the steps of the method according to one of Claims 1 to 9.