Method for detecting an electrical fault in a stator of an electrical machine, in particular in a generator of a wind turbine

Abstract

A method for detecting an electrical fault in a stator of an electrical machine, wherein the stator comprises a plurality of sets of windings, wherein each set of windings is assigned to a respective phase of the electrical machine, the method comprising the steps of determining a respective current: first, between a subset of one of the sets of windings and a different further subset of the same set of windings, and / or second, between a subset of one of the sets of windings and a neutral point, and / or third, between the neutral point and a further neutral point or at least a common neutral point connected to the neutral point and the further neutral point, evaluating a fault condition, wherein a fulfillment of the fault condition depends on the respective determined current, and outputting a fault signal to a person and / or a device when the fault condition is fulfilled.

Description

Technical Field

[0001] This invention relates to a method for detecting electrical faults in the stator of a motor, particularly in a wind turbine generator, wherein the stator comprises multiple sets of windings, each set of windings being assigned to a corresponding phase of the motor. The invention also relates to a motor assembly and a wind turbine. Background Technology

[0002] Electrical faults in the stator, especially short circuits between adjacent turns in the windings (also known as turn faults), are a common failure mode in motors. When the motor is used as a generator, such electrical faults can cause intense localized heating, thus posing a risk of damage to the generator and surrounding components. Similar problems arise when the motor is used as an electric motor. Therefore, it is highly relevant to detect such electrical faults as quickly as possible to allow for shutdown, changes to motor operation, and / or notification of maintenance personnel before further damage is likely.

[0003] Methods for detecting turn faults are known in the art. Typically, for example, an electrical imbalance in a motor due to a turn fault is detected by using a negative sequence voltage. This method is disclosed, for example, in document US 2011 / 0187304A1. Performing this measurement while the generator is driving a load can potentially generate very high currents, and therefore lead to strong localized heating when a turn fault is present. This could potentially cause a fire or damage to other components.

[0004] Turn faults can also be potentially detected by measuring the vibration of the generator caused by torque pulsations due to electrical faults. However, in large machines, the torque pulsations caused by a turn fault in a single path of multiple parallel paths are relatively small, and therefore the resulting vibrations are relatively small, making this method relatively insensitive. Even when a fault is detected, it is usually impossible to identify which of the many windings in a large machine contains the electrical fault. Summary of the Invention

[0005] Therefore, the problem to be solved is to provide an improved fault detection method that has a lower risk of damaging other components and / or increased sensitivity, which preferably allows for a certain degree of fault localization.

[0006] The problem was solved using the method initially described, which includes the following steps:

[0007] - Determine the corresponding current:

[0008] First, between a subgroup of a group of windings that includes at least one winding in a group and a different additional subgroup of the same group of windings that includes at least one other winding in the same group, and / or

[0009] Second, between a subgroup of a group comprising at least one winding in a group and a neutral point connected to at least one winding in each group, and / or

[0010] Third, between the neutral point connected to at least one winding of each group and another neutral point connected to at least one other winding of each group, or at least a common neutral point connected to the neutral point and the other neutral point,

[0011] - Evaluate the fault conditions, wherein the satisfaction of the fault conditions depends on the corresponding determined current, and

[0012] - When the aforementioned fault conditions are met, a fault signal is output to the individual and / or device.

[0013] Each group of windings is assigned to a corresponding phase of the motor. The windings of each corresponding group can be connected to the corresponding phase of the motor. The corresponding phase connection can be, for example, connected to a junction box and / or a generator circuit breaker. A subgroup can, for example, include a single winding or all windings from a given segment of the stator.

[0014] An electric motor can be a generator or an electric motor. When an electric motor is used as a generator, it can rotate and / or its speed can be increased by external torque, for example, by manual rotation or by another electric motor, while measuring one or more currents.

[0015] This method leverages the fact that an electrical fault in the stator alters the impedance of the winding containing that fault. This is particularly true for inter-turn faults that create short circuits between adjacent turns of the winding. Since individual windings are connected to the optional grounded neutral point of the multiphase system, an impedance change in one of the windings causes a potential shift on the phase-connected side of that winding's connection to the corresponding group. This results in a potential difference between points in the circuit that would otherwise be at the same potential, and thus generates a loop current within the stator winding.

[0016] This fact is utilized by measuring the loop current flowing through the windings or by measuring at least one current that is at least partially caused by such current. An advantage of this type of measurement is that it can be performed even when the motor is not supplying any current, especially when the motor is not connected to a load. For example, it is possible to perform fault detection according to the previously discussed methods during the accelerated rotation of a wind turbine before it reaches cut-in speed and is connected to an external load.

[0017] Because the measured current is typically high enough to reliably detect such faults before a significant temperature rise occurs due to loop current caused by the fault, the risks of damage to other components or even fire, as discussed earlier, can be avoided. By performing this measurement on different groups and / or subgroups within the same group, it is also possible to locate windings or subgroups of windings that contain electrical faults.

[0018] When all the stator windings are connected to a single neutral point (also called a star point), the loop current between the parallel windings, and therefore the current measured in the methods described above, can be particularly high. However, it should be understood that when the stator has several individual neutral points, there is also a sufficiently large loop current for performing the methods discussed, where each neutral point is connected to at least one winding of each phase. If an electrical fault exists in the winding of a particular phase, current transfer between the different neutral points is required to allow the loop current to flow in this situation. This current can be measured directly between the different neutral points, or the current between the respective neutral point and a common neutral point connecting, for example, the individual neutral points of different segments of the stator, can be measured to detect the electrical fault.

[0019] When different neutral points are used that are not connected via a common neutral point, these neutral points are still connected via the windings of other phases, thus still allowing loop current to flow in the phases containing electrical faults. Although this results in additional impedance to the loop current and therefore a lower loop current than when using a single neutral point, the resulting loop current is still sufficient to allow fault detection by the methods discussed earlier.

[0020] The current can be measured while the motor rotor is rotating. Preferably, the current measurement is performed when there is essentially no electrical load on the motor, for example, when the circuit breaker between the motor and the load is open and / or when a current converter including a semiconductor switch disconnects the windings from the load.

[0021] The windings of each corresponding group can be connected to the corresponding phase connection of the motor. This phase connection is typically connected to the junction box and / or generator circuit breaker. The different windings of the corresponding phase or group can be connected in essentially parallel between the phase connection and the common neutral point or the corresponding neutral point. When different neutral points are used, there may be a small potential difference between these neutral points, resulting in a slight deviation from the strict parallel connection of the windings of the corresponding group or phase.

[0022] The other subgroups of the corresponding group of windings can include all windings of that group that are not part of the first subgroup. Therefore, the measured current can be the current between the corresponding subgroup and all other windings in the same group.

[0023] The corresponding current can be determined using a current sensor connected between the sub-group of the winding and the phase connection or neutral point of that winding. Both the sub-group and another sub-group can be connected to the phase connection of the corresponding phase, and thus can be connected via a current sensor connected between the sub-group and the phase connection.

[0024] In a preferred embodiment, the current is measured while the phase connection of the motor is disconnected from the load, for example, via a circuit breaker and / or via a semiconductor switch. The semiconductor switch can be, for example, a switch used in a power converter. For example, it is possible to operate the converter using pulse width modulation. By reducing the pulse width of all switches to zero, the load can be disconnected. In this case, no current or only negligible leakage current can pass through the corresponding phase connection, so the current through the current sensor can be entirely caused by a loop current, which is very low in the absence of an electrical fault but increases sharply in the presence of one. The electrical fault can then be detected simply, for example, by comparing the measured current to a threshold. However, it may be advantageous to perform some preprocessing to, for example, combine and / or compare different measured currents.

[0025] A corresponding current can be determined for at least two groups and / or for at least two different subgroups of at least one group, wherein the corresponding subgroup includes at least one winding, and the current of the corresponding subgroup is the current between the corresponding subgroup and a corresponding additional subgroup or neutral point of the same group winding, which includes at least one additional winding of the same group, wherein the satisfaction of the fault condition depends on the determined current. The corresponding additional subgroup may in particular include all windings of the corresponding group that are not part of the corresponding subgroup. In other words, for each of a plurality of subgroups, the current from the windings of the corresponding subgroup to another winding of the same group that is not part of that subgroup can be determined.

[0026] A fault condition may, for example, include multiple sub-conditions, each depending on one of the determined currents. A fault condition is satisfied if at least one sub-condition is met. The satisfaction of a corresponding sub-condition may, for example, indicate the presence of a fault in the corresponding group and / or subgroup for which current measurements are performed.

[0027] The corresponding current can be determined when the motor is not providing power and / or when the stator phase connection is disconnected from the load and / or when the current driven through the phase connection is zero or below a threshold. For example, a very common method for accelerating the rotation of a wind turbine is to connect the motor used as a generator to the load only when a specific on-time of the wind turbine is reached, and therefore a specific on-time of the motor's rotor. Therefore, in particular, one or more current measurements and thus the detection of electrical faults can be performed during the accelerated rotation of the wind turbine before reaching the on-time and therefore before the generator is connected to the load.

[0028] The fact that the current driven by the corresponding phase connection is below a threshold can be definitively verified, for example, by measuring the corresponding current driven by the corresponding phase connection. However, it can also be assumed that the current is below this threshold when other parameters of the wind turbine clearly indicate this, for example, when the electronic switch connecting the corresponding phase connection to the load is turned off, for example by disabling the converter by setting the pulse width of the switch control signal to zero, thereby allowing only negligible leakage current to pass through the switch.

[0029] The corresponding current can be measured multiple times within a given time interval, and the satisfaction of the fault condition can depend on the multiple measurements. For example, the maximum value of the current during that time interval can be determined. Alternatively or additionally, the temporal evolution of the corresponding current, such as its correlation with the rotational speed of the motor rotor, can be considered.

[0030] The rotational speed of the motor rotor can be increased during a given time interval. As mentioned earlier, measurements can be taken, in particular, during the accelerated rotation of a wind turbine.

[0031] The motor can be a generator for a wind turbine, wherein the wind turbine is started by pitching the blades of the wind turbine to increase the speed of the motor rotor from a standstill, wherein a given time interval covers at least a portion of the time interval between the start time of the rotor being at rest and the end time of the motor being connected to the load. In other embodiments, the motor can be, for example, a large generator with multiple sets of parallel windings or even an electric motor. The increase in speed can be driven, for example, by an external electric motor that drives the motor.

[0032] The end time can specifically correspond to the time when the motor rotor speed reaches a given threshold, particularly the previously mentioned switching speed. The time interval can also end before the generator is switched on. This, for example, allows for the suppression of load connection if fault conditions are met.

[0033] As an alternative, the load can still be connected to the generator even when the fault conditions are met. For example, it may be possible to send a warning to an individual or device only when the fault conditions are met and the generator is otherwise operating normally, or to modify only certain operating parameters, such as the maximum value of the generator speed, maximum load, etc.

[0034] A corresponding maximum value of a corresponding current can be determined from multiple measurements of that current, wherein the satisfaction of a fault condition may depend on that corresponding maximum value. This can be achieved, for example, by digitally recording a series of current measurements and selecting the maximum value from these measurements, or by using a circuit that maintains, for example, the maximum value of the output of an envelope follower that follows the current.

[0035] The stator segments may each include at least one winding in each group, wherein the windings of the corresponding segments are connected to the neutral point of that segment.

[0036] - wherein, at least one segment of the segments has a winding that provides a unique electrical connection between the neutral point of that segment and the windings of the other segments.

[0037] -Or in particular, the neutral point is connected to a common neutral point via a corresponding current sensor.

[0038] The first alternative is an example of the alternative initially discussed, in which several separate neutral points are used. Alternatively, all windings could be connected to a common neutral point.

[0039] Using a separate neutral point typically simplifies stator wiring. As mentioned earlier, using a separate neutral point also simplifies monitoring different phases because it requires the loop current caused by an electrical fault in one phase to be conducted through the windings of other phases. When measuring current between subgroups of windings within the same group, the use of a single neutral point would require a current sensor for each phase of the motor if an electrical fault in each phase should be detectable. However, if an electrical fault exists in the first phase, using a single neutral point does indeed cause an increase in the measured current in the remaining phases. Thus, a substantially similar increase in the measured current in two phases already indicates an electrical fault in the third phase. Therefore, using a single neutral point allows one of these phases to remain unmonitored. However, when using a common neutral point, it is also possible to measure the current between the neutral point of a separate segment and the common neutral point, thus potentially using even fewer sensors, especially a single sensor per segment and / or neutral point.

[0040] The corresponding measure of the amplitude of motor vibration and / or torque oscillation acting on the motor rotor can be determined, wherein the satisfaction of fault conditions may depend on the corresponding measure.

[0041] For example, it is possible to record torque and / or acceleration at a given time interval, particularly the aforementioned time interval, and detect the maximum value of the oscillation amplitude of acceleration and / or torque. It is particularly possible to consider only oscillations at specific frequencies, such as oscillations at specific harmonics of the electrical frequency output by the motor. It has been recognized that electrical faults and the resulting torque oscillations will typically cause a strong increase in vibration and torque oscillations at twice the electrical frequency. This frequency band can be isolated, for example, by a bandpass filter, or the measurement data can be digitally sampled and transformed to the frequency domain, for example, by a Fourier transform, to determine the vibration amplitude within the relevant frequency range.

[0042] The present invention also relates to an electric motor device, particularly for wind turbines, comprising a motor having a stator having multiple sets of windings, wherein each set of windings is assigned to a corresponding phase of the motor, and a monitoring device, wherein the monitoring device and the motor are configured to perform the method according to the invention.

[0043] The monitoring device may include at least a current sensor for determining the corresponding current and a processing unit for assessing fault conditions and optionally for performing additional steps of the above method.

[0044] Monitoring equipment can be integrated within the motor and, for example, attached to the stator or stator housing. Alternatively, at least some components of the monitoring equipment can be spaced apart from the motor. For example, it is possible that current sensors are integrated into the motor, and processing units and / or signal transmission units for outputting fault signals to individuals and / or equipment are spaced apart from the motor. These components can be mounted, for example, to another part of the wind turbine, such as within the tower or base, or even externally to the wind turbine when the motor is part of the wind turbine. It is even possible, for example, that the processing unit is implemented as a network-based or cloud-based solution. For example, it is possible that one or more current sensors communicate with off-site units via network protocols (e.g., via TCP / IP), or more typically with local control units assigned to individual wind turbines or small clusters of wind turbines.

[0045] Furthermore, the present invention relates to a wind turbine including a motor device according to the invention. As previously stated, the method discussed, and therefore the motor device discussed, is particularly advantageous when using large motors, such as in wind turbines. Attached Figure Description

[0046] Other objects and features of the invention will become apparent from the following detailed description taken in conjunction with the accompanying drawings. However, the drawings are merely schematic diagrams designed for illustrative purposes and are not intended to limit the invention. The drawings show:

[0047] Figure 1 This is an exemplary embodiment of a wind turbine according to the invention, including an exemplary embodiment of the motor device according to the invention;

[0048] Figure 2 yes Figure 1 A schematic diagram of the motor device shown;

[0049] Figure 3-5 At different speeds of the motor Figure 2 The current measurement of the three current sensors shown;

[0050] Figure 6-8 It acts on the motor at different speeds Figure 2 Measurement of torque oscillation on the rotor of the motor device shown;

[0051] Figure 9 This is a flowchart illustrating an exemplary embodiment of a method for detecting electrical faults in the stator of an electric motor according to the present invention, and...

[0052] Figure 10 This is a schematic diagram of another exemplary embodiment of the motor device according to the present invention;

[0053] Figure 11 This is a schematic diagram of another exemplary embodiment of the motor device according to the present invention. Detailed Implementation

[0054] Figure 1 A wind turbine 1 including an electric motor assembly, which in turn includes a motor 2, in this example a generator, and a monitoring device for detecting electrical faults in the stator 4 of the motor 2, is shown. The rotor 3 of the motor 2 is connected to the hub 5 of the wind turbine, which carries the blades 6. The stator 4 is attached to a nacelle 7 carried by a tower 8.

[0055] In this example, the rotor 3 is positioned outside the stator 4 in the motor 2; however, a stator surrounding the rotor can also be used.

[0056] For the sake of simplicity and clarity, in Figure 1 Only a few components of the monitoring device are shown, namely the current sensors 9-11 and the processing unit 12. Figure 2 A more detailed schematic diagram of the monitoring device 14 and its interaction with the motor 2 (i.e., with the different windings 19-27 of the stator 4 of the motor 2) is shown below. For simplicity, Figure 2 Only the use of monitoring device 14 for monitoring a single segment 16 of stator 4 is shown. Clearly, additional current sensors can be used to monitor the current through the additional windings 22-27, and thus also detect electrical faults in the additional segments 17, 18. To be able to locate faults in all segments, n-1 sensors per phase are sufficient, where n is the number of stator segments.

[0057] like Figure 2As shown, motor 2 includes three sets of windings 19-27, 38, 51, and 52, wherein windings 19-27 of each set 38, 51, and 52 are assigned to corresponding phases 28-30 of motor 2. In this example, windings 19, 22, and 25 are part of set 38, windings 20, 23, and 26 are part of set 51, and windings 21, 24, and 27 are part of set 52. One side of windings 19-27 of each set 38, 51, and 52 is connected to the corresponding phase connection 31-33 of that set. Phase connections 31, 32, and 33 can be connected, for example, to a circuit breaker 34 of motor 2 or other electronic components of motor 2.

[0058] exist Figure 2 In the example shown, stator 4 includes several segments 16, 17, 18, each of which includes a single winding 19-27 from each of groups 38, 51, 52, wherein three windings 19-27 of the respective segments 16, 17, 18 are connected to the respective neutral points 35, 36, 37 of the three-phase system. As shown at point 84, stator 4 with more than three segments 16, 17, 18 is typically used, for example, a stator with twelve segments.

[0059] Using separate neutral points 35, 36, and 37 for each segment 16, 17, and 18 reduces the complexity of stator 4 wiring. In an alternative embodiment, a common neutral point or direct connection can also be used. Figure 2 All neutral points 35, 36, and 17. In this case, an electrical fault in windings 19-27 will result in a higher loop current, thus resulting in a higher current measured by current sensors 9, 10, and 11 than in the illustrated embodiment. However, it has been found that electrical faults can also be robustly detected when individual neutral points 35, 36, and 37 are used for different segments 16, 17, and 18.

[0060] For the following discussion, it is assumed that an electrical fault exists in winding 19, particularly an inter-turn fault. This electrical fault can be detected even when there is no load 55 driven by motor 2, for example, when circuit breaker 34 is open.

[0061] An electrical fault in winding 19 will cause winding 19 to have a different impedance than the other windings 22 and 25 of group 38. If the neutral points 35, 36, and 37 of different segments 16, 17, and 18 are directly connected, a loop current will be induced in phase 28 when rotor 3 rotates.

[0062] Even when different segments 16, 17, and 18 use separate neutral points 35, 36, and 37, a similar effect can be achieved because, in this case, current transfer between neutral points 35, 36, and 37 is possible through other phases 29 and 30. Therefore, an electrical fault in winding 19 will result in a strong increase in the current measured by current sensor 9, and a lower, typically about half the strength, increase in the current measured by current sensors 10 and 11.

[0063] Figures 3 to 5 The currents 42-44, measured by current sensors 9-11, are shown for different rotational speeds of rotor 3. The corresponding x-axis 40 corresponds to time, and the corresponding y-axis 41 responds to the measured current. Figure 3 The data shown was obtained at a very low engine speed of 1 rpm. Figure 4 The data shown was obtained at a relatively high engine speed of 4 rpm, and Figure 5 The data shown was obtained at a speed slightly below the wind turbine’s on-time, at which motor 2 will be connected to load 55, which in this example is 10.3 rpm.

[0064] At very low speeds, the effects of electrical faults in winding 19 are not clearly detectable. Increasing the speed to 4 rpm does allow for clear detection of electrical faults, as... Figure 4 As can be seen, although the slightly larger magnitude of current 42 measured by current sensor 9 compared to currents 43 and 44 measured by current sensors 10 and 11 may be difficult to detect due to noise levels, the fact that the currents in phases 29 and 30 detected by both current sensors 10 and 11 have the same phase and are 180° phase-shifted relative to the current in phase 28 detected by current sensor 9 clearly indicates that the observed currents are caused by the impedance difference between windings 19, 22, and 25 in phase 28, and that the increased currents 43 and 44 measured by current sensors 10 and 11 are due to the current loop closing the loop current of phase 28 via phases 29 and 30, since neutral points 35, 36, and 37 are not directly connected.

[0065] When Figure 5 As shown, when measuring currents 42 to 44 at even higher speeds, it is immediately noticeable that current sensor 9 measures a considerably high current 42 due to an electrical fault in winding 19, and these currents 42 are significantly higher than the currents 43 and 44 measured by current sensors 10 and 11 in other phases 29 and 30. Therefore, an electrical fault in stator 4, i.e., in winding 19, can be easily detected, for example, by comparing the maximum measured current with a threshold.

[0066] Another effect of the electrical fault in winding 19 is the fluctuation of torque acting on rotor 3. This effect... Figures 6 to 8 As shown in the diagram, the corresponding x-axis 45 represents time, and the corresponding y-axis represents the calculated torque, in this example in 10... 4 The unit is Nm. Figure 6 Measurement 49 was recorded at 1 rpm. Figure 7 The measurement 48 was recorded at 4 rpm, and Figure 8 Measurement 47 was recorded at 10.3 rpm.

[0067] It is immediately apparent from these figures that, especially at the higher speeds immediately preceding the on-state of motor 2, the torque oscillations are significantly large. These oscillations can be measured directly. However, in many cases, it is simpler to use accelerometer 13 because the oscillations in the applied torque will immediately result in vibration. It has been recognized that when an electrical fault is present in one of the windings 19-27, the vibration, and therefore the oscillations in acceleration, are caused at a frequency twice that of the supplied current.

[0068] Based on these observations, we will now refer to Figure 9 The flowchart illustrating an exemplary embodiment of this method discusses in detail the operation of the monitoring device 14 and the method for detecting electrical faults in the stator 4. In its most general sense, the method includes determining, in step S2, the currents between corresponding subgroups 64, 65, 66 of the respective groups 38, 51, 52 of windings 19-27 and different additional subgroups 39, 53, 54 of the same groups 38, 51, 52 of windings 19-27, and in step S6, evaluating fault conditions 63 dependent on these currents. If fault condition 63 is met, a signal 83 is output to the individual and / or device 15.

[0069] In the initial state of step S0, the wind turbine 1 is stopped. In step S1, the actuator 50 is used to tilt the rotor blades 6. Therefore, the hub 5 and the rotor 3 begin to rotate at an increased speed.

[0070] As the rotational speed increases slowly, in step S2, the processing unit 12 acquires corresponding measurements 57 of each current 42, 43, 44 via the respective current sensors 9, 10, 11. Simultaneously, the processing unit 12 acquires a measurement 58 of vibration 56 via accelerometer 13. In particular, it is possible to measure vibration using an accelerometer attached to a fixed component (e.g., a bearing or stator). As discussed below, step S2 will be repeated multiple times during the time interval between the rotational speed increasing from rest to a speed slightly below the connection speed at which the motor will be connected to the load 55. Therefore, a corresponding time series of measurements 57 of each of the currents 42, 43, 44 and the measurement 58 of vibration 56 are acquired.

[0071] As previously described, each of the measured currents 42, 43, and 44 is the current between subgroups 64, 65, and 66 of windings 19-27 of the corresponding groups 38, 51, and 52 of windings 19-27 and at least one additional subgroup 39, 53, or 54 of another winding 19-27 that includes the corresponding groups 38, 51, and 52 of windings 19-27. Subgroup 64 includes only a single winding 19, subgroup 65 includes only a single winding 20, and subgroup 66 includes only a single winding 21. The additional subgroup 39 includes windings 22 and 25, the additional subgroup 53 includes windings 23 and 26, and the additional subgroup 54 includes additional windings 24 and 27.

[0072] In step S3, the rotational speed 59 is determined, for example, by an additional sensor (not shown). In step S4, the rotational speed 59 is compared with a threshold 60, and if the rotational speed 59 is lower than the threshold 60, step S2 is repeated.

[0073] Once the rotational speed 59 exceeds the threshold 60, a corresponding maximum value 61 is determined for each of the currents 42, 43, and 44 based on the maximum value of measurement 57. An additional maximum value 62 is determined for vibration 56.

[0074] In step S6, fault condition 63 is evaluated. Fault condition 63 may be satisfied, for example, when at least one of the maximum values ​​61 of currents 42, 43, and 44 exceeds a given threshold. Alternatively, the fault condition may be satisfied, for example, only when the maximum value 62 of vibration 56 also exceeds the threshold, or when a combined fault value calculated based on the maximum values ​​61 and 62, for example by multiplying by the maximum values, exceeds the threshold.

[0075] If fault condition 63 is not met, then in step S8, for example by closing circuit breaker 34 once the connection speed is exceeded and thus providing power to load 55, normal operation of wind turbine 1 continues.

[0076] On the other hand, if fault condition 63 is met, processing unit 12 can output a control signal to output device 86. In its simplest case, output device 86 can be a device for outputting light or sound signals to individuals, such as signal lights, loudspeakers, etc. However, it may be advantageous to output signal 83 instead to a device 15 in the internal control unit of the wind turbine or the generator itself, such as to the central unit of the wind farm, communication equipment for service personnel, etc. Generally, it is advantageous to provide the signal to a controller that can locally and / or quickly prevent and / or stop further operation of the motor.

[0077] like Figure 2As shown in the dashed square 85, instead of current sensors 9-11 that measure the corresponding currents between the two subgroups 64-66 and 39, 53, 54 of windings 19-27, measurements of the corresponding currents between the respective subgroups 64, 65, 66 and the neutral point 35 of at least one winding 19, 20, 21 connected to each group 38, 51, 52 can be used. The measured currents are then processed as described above to assess fault condition 63.

[0078] Figure 10 It shows that it can be replaced Figure 2 Different examples of motor devices used in the motor devices shown. Figure 10 The motor device shown includes a motor 82, wherein, in Figure 10 Only the stator windings 19-27, 69-77 of the motor 82 are shown, along with the monitoring device 14 using six current sensors 9, 10, 78-81. The windings 19-27, 69-77 are divided into groups assigned to phases 28, 29, 30 and into segments 16, 17, 18, 67, 68, 87, wherein each segment 16, 17, 18, 67, 68, 87 comprises the windings 19-27, 69-77 of each group and therefore each phase 28, 29, 30.

[0079] According to Figure 2 The main difference in the embodiments is the use of a single current measurement, and therefore a single current sensor 9, 10, 78, 79, 80, 81, to measure the current between two subgroups, each subgroup comprising multiple windings 19-69, 70-77. Current sensor 9, for example, measures the current between a first subgroup comprising windings 19 and 22 and another subgroup comprising windings 25, 69, 72, and 75. On the other hand, current sensor 10 measures the current between a first subgroup comprising windings 20 and 23 and another subgroup comprising windings 26, 70, 73, and 76. A high current measured by current sensor 9 indicates an electrical fault in winding 19 or 22. Similarly, a high current measured by current sensor 10 indicates an electrical fault in winding 20 or 23.

[0080] exist Figure 10 In the example shown, there is no current sensor that directly measures any current in phase 30. Such a current sensor is not necessary because the current, for example, from windings 21 and 24 to other windings 27, 71, 74, and 77 can be directly calculated based on the current measurements of current sensors 9 and 10. Therefore, based on the current measurements of current sensors 9 and 10, electrical faults in windings 19-24 of segments 16 and 17 can be detected, and it can be determined which winding 19-24 of phases 28, 29, and 30 in segment 16 or 17 is faulty.

[0081] In the same manner, current sensors 78 and 79 can be used to determine electrical faults in segments 18 and 87, and current sensors 80 and 81 can be used to determine electrical faults in segments 67 and 68.

[0082] Figure 10 The embodiment shown allows for monitoring of a large number of windings 19-27, 69-77 for electrical faults, with very good localization of the fault within two corresponding windings. Meanwhile, the technical complexity of the monitoring device 14 is relatively low, as only six current sensors 9, 10, 78-81 are necessary for the three segments shown. The number of necessary sensors will increase significantly with the number of segments used.

[0083] Figure 11 Another embodiment of the motor device is shown. This embodiment is similar in many respects to... Figure 2 The embodiment shown will therefore only discuss the distinguishing features in detail. For simplicity, only the wiring of windings 20-27 and the location of the current sensor are shown.

[0084] This embodiment uses a common neutral point 88, where the neutral points 35-37 of different segments are connected to this common neutral point 88 via corresponding current sensors. An electrical fault in the windings 19-27 within segments 16-18 alters the potential at the neutral point 35-37 of that segment, thus causing current to flow through the corresponding current sensor 9-11. By measuring the current between the neutral point 35-37 and the common neutral point 88, a defect in any of segments 16-18 can be located using only a single current sensor 9-11 for each segment.

[0085] Although the invention has been described in detail with reference to preferred embodiments, the invention is not limited to the disclosed examples, and those skilled in the art can derive other variations from the disclosed examples without departing from the scope of the invention.

Claims

1. A method for detecting electrical faults in the stator (4) of a motor (2), wherein, The stator (4) includes multiple sets of windings (19-27, 69-77) (38, 51, 52), wherein the windings (19-27, 69-77) of each set (38, 51, 52) are assigned to the corresponding phases (28, 29, 30) of the motor (2), and the method includes the following steps: - Determine the current (42, 43, 44) between the neutral point (35, 36, 37) of at least one winding (19-27, 69-77) connected to each group (38, 51, 52) and the additional neutral point (35, 36, 37) of at least one additional winding (19-27, 69-77) connected to each group (38, 51, 52), or at least a common neutral point connected to said neutral point and said additional neutral point. - Evaluate the fault condition (63), wherein the satisfaction of said fault condition (63) depends on the determined currents (42, 43, 44), and - When the fault condition (63) is met, a fault signal (83) is output to the person and / or equipment (15).

2. The method of claim 1, wherein a subgroup (64, 65, 66) is determined from one of the groups (38, 51, 52) of windings (19-27, 69-77) comprising at least one winding of the group (38, 51, 52) and the same group (38, 51, 52) of windings (19-27, 69-77) comprising at least one additional winding of the same group (38, 51, 52). The additional current between different subgroups (39, 53, 54), and / or between a subgroup (64, 65, 66) of one of the groups (38, 51, 52) including at least one winding (19-27, 69-77) of the group (38, 51, 52) and the corresponding additional current between the subgroup (64, 65, 66) connected to the neutral point (35, 36, 37) of at least one winding (19-27, 69-77) of each group (38, 51, 52). The satisfaction of the fault conditions also depends on the corresponding additional current determined.

3. The method according to claim 2, characterized in that, For at least two groups (38, 51, 52) and / or for at least two distinct subgroups (64, 65, 66) of at least one group (38, 51, 52), determine corresponding additional currents (42, 43, 44), wherein the corresponding subgroup (64, 65, 66) comprises at least one winding (19-27, 69-77), and the current of the corresponding subgroup (64, 65, 66) is the current of the corresponding subgroup (64, 65, 66) and the package. The current (42, 43, 44) between the corresponding additional subgroups (39, 53, 54) or neutral points (35, 36, 37) of at least one other winding (19-27, 69-77) of the same group (38, 51, 52), and the corresponding neutral points (35, 36, 37), wherein the satisfaction of fault condition (63) depends on the determined current (42, 43, 44).

4. The method according to any one of the preceding claims, characterized in that, The current (42, 43, 44) is determined when there is no power supplied by the motor (2) and / or when the phase connections (31, 32, 33) of the stator (4) are disconnected from the load (55) and / or when the current driven through the phase connections (31, 32, 33) is zero or below a threshold.

5. The method according to any one of claims 1-3, characterized in that, Multiple measurements (57) of current (42, 43, 44) are obtained within a given time interval, wherein the satisfaction of fault condition (63) depends on the multiple measurements (57).

6. The method according to claim 5, characterized in that, During the given time interval, the rotational speed (59) of the rotor (3) of the motor (2) is increased.

7. The method according to claim 5, characterized in that, The motor (2) is a generator (2) of a wind turbine (1), wherein the wind turbine (1) is started by tilting the blades (6) of the wind turbine (1) to increase the speed of the rotor (3) of the motor (2) from a stationary state, wherein the given time interval covers at least a portion of the time interval between the start-up time when the rotor (3) is stationary and the end time when the motor (2) is connected to the load (55).

8. The method according to any one of claims 6 to 7, characterized in that, The corresponding maximum value (61) for the currents (42, 43, 44) is determined based on multiple measurements (57) for the currents (42, 43, 44), wherein the satisfaction of the fault condition (63) depends on the corresponding maximum value (61).

9. The method according to any one of claims 1-3, characterized in that, The stator (4) comprises multiple segments (16, 17, 18, 67, 68, 87), each including at least one winding (19-27, 69-77) of each group (38, 51, 52), wherein the winding (19-27, 69-77) of the corresponding segment (16, 17, 18, 67, 68, 87) is connected to the neutral point (35, 36, 37) of that segment (16, 17, 18, 67, 68, 87). - wherein, the winding (19-27, 69-77) of at least one of the segments (16, 17, 18, 67, 68, 87) provides a unique electrical connection between the neutral point (35, 36, 37) of that segment (16, 17, 18, 67, 68, 87) and the winding (19-27, 69-77) of the other segments (16, 17, 18, 67, 68, 87). -Or, in which the neutral points (35, 36, 37) are connected to the common neutral point (88).

10. The method according to any one of claims 1-3, characterized in that, Determine a corresponding metric (58) for the amplitude of the vibration (56) of the motor (2) and / or the torque oscillation acting on the rotor (3) of the motor (2), wherein the satisfaction of the fault condition (63) depends on the corresponding metric (58).

11. The method of claim 1, wherein the electrical fault is in the generator of the wind turbine (1).

12. The method of claim 9, wherein the neutral point (35, 36, 37) is connected to the common neutral point (88) via respective current sensors (9, 10, 11).

13. A motor device comprising a motor (2) and a monitoring device (14), said motor (2) having a stator (4) having a plurality of windings (19-27, 69-77) (38, 51, 52), wherein, The windings (19-27, 69-77) of each group (38, 51, 52) are assigned to the corresponding phases (28, 29, 30) of the motor (2), characterized in that the monitoring device (14) and the motor (2) are configured to perform the method according to any one of the preceding claims.

14. The motor device according to claim 13, wherein the motor device is used in a wind turbine (1).

15. A wind turbine, characterized in that, The wind turbine includes the motor unit as described in claim 13 or 14.

Images