Determining electrical angular position of rotor of wind turbine generator

Abstract

A method of determining a rotor electrical angular position (17) of a generator (1a) comprising a stator (2) and a rotor (3) having a plurality of mounted permanent magnets (4), in particular of a wind turbine, is described, the method comprises: allowing reception and / or reception of at least time-dependent measurement signals (9a, 9b) from a first analog Hall sensor (10a) and a second analog Hall sensor (10b) and optionally a third analog Hall sensor (10c), a plurality of Hall sensors mounted at different circumferential positions at the stator (2) to detect a magnetic flux resulting at least in part from one or more of the permanent magnets; evaluating the received measurement signals (9a, 9b, 9c); selecting at least one measurement signal based on the evaluation in order to exclude, in particular, the measurement signals of any faulty sensor; an electrical angular position (17a, 17) is determined on the basis of the at least one selected measurement signal.

Description

Technical Field

[0001] The present invention relates to a method and a corresponding device for determining the electrical angular position of a rotor of a generator comprising a stator and a rotor with a plurality of mounted permanent magnets, in particular a generator of a wind turbine. The present invention further relates to a method for controlling a permanent magnet generator, wherein the control is based on the determined electrical angular position of the rotor. Furthermore, the present invention relates to a generator system comprising a device for determining the electrical angular position of the rotor. Background Art

[0002] For conventional permanent magnet synchronous motors, rotor position feedback (generator electrical angle) may be required to control the generator. Conventionally, various position feedback techniques can be employed. For example, during normal generator power generation, when the generator rotates above a certain speed, an EMF (electromotive force) observer can be used to estimate the rotor position. When starting the motor without assistance from the aerodynamic torque of the blades, a high-frequency injection (HFI) observer can be used to estimate the rotor position, which is then used to generate generator torque. In some specific applications that require high dynamics in terms of position feedback, encoders can also be utilized.

[0003] Each of the above-mentioned techniques and methods has its own advantages and disadvantages, and conventionally a single technique may not meet all control requirements in all situations. Often, multiple techniques of the above-mentioned techniques can be conventionally provided, and thus the system may become very complex in terms of required components and configuration.

[0004] Conventional turbine applications may include utilizing the rotor position for generating electrical energy in a normal power generation mode, for motor drive, for rotor positioning for speed control and / or torque generation, and the like.

[0005] Conventional EMF observers rely on the calculation of the motor's back EMF (electromotive force) and are typically applied when the rotor speed exceeds a certain speed level. Conventionally, EMF observers can be used during normal power generation mode. However, at zero or low speeds, this approach may not be suitable.

[0006] The electrical position of the motor can also be estimated by a so-called HFI observer, by applying a voltage or injecting a high-frequency (typically a few hundred hertz) current and measuring the response in terms of current and / or voltage. An HFI (high-frequency injection) observer can be used at zero or low speed, for example for motor drive and rotor positioning purposes. Since this method is potentially intrusive to the drive system (due to the need to inject high-frequency current and / or voltage), it cannot be applied continuously and may interfere with certain operations. Furthermore, this method may be difficult for motors with no or low saliency.

[0007] Conventional encoders can also be mounted at the center of rotation of the generator or on the outside (e.g., on a brake disc). It has been observed that encoders provide unacceptable electrical angle errors, especially when the generator diameter is relatively large and the number of poles is also large. Furthermore, encoders mounted on the outside are only suitable for low-speed operation.

[0008] Therefore, there may be a need for a method and a corresponding device for determining the electrical angular position of the rotor of a generator, which method and device provide more reliable results and are capable of providing accurate and reliable rotor positions in different operating modes, in particular at different rotational speeds of the rotor, in particular at low speed, at high speed and at nominal speed. Summary of the Invention

[0009] This need is met by the subject matter according to the independent claims.The dependent claims describe advantageous embodiments of the invention.

[0010] According to an embodiment of the present invention, a method for determining the electrical angular position of a rotor of a generator comprising a stator and a rotor with a plurality of mounted permanent magnets is provided, the generator being in particular a generator of a wind turbine, the method comprising: allowing and / or receiving at least time-correlated measurement signals from a first analog Hall sensor and time-correlated measurement signals from a second analog Hall sensor, which Hall sensors are mounted at different circumferential positions on the stator to detect a magnetic flux at least partially due to one or more of the permanent magnets; evaluating the received measurement signals; selecting at least one measurement signal based on the evaluation so as to in particular exclude the measurement signals of any faulty sensor; determining the electrical angular position based on at least one selected measurement signal (including a calculated combination of at least one selected measurement signal).

[0011] The method can be implemented by software and / or hardware.The method can be executed by the apparatus for determining the electrical angular position of the rotor according to an embodiment of the present invention.

[0012] The rotor electrical angle position θ_e may be different from the mechanical angle position θ_m, but the two can be calculated by the mathematical formula of the mechanical angle position and the number of motor pole pairs p p and possible angular offsets And for the relevant ones, that is The mechanical angular position of the rotor may correspond to or be equal to the azimuthal angular position of the rotor.

[0013] The rotor electrical angular position can be used to vector control the generator, which may involve transforming into and out of a dq coordinate system that rotates synchronously with the rotor. In the dq coordinate system, electrical quantities will be essentially DC quantities.

[0014] The stator may comprise one or more stator segments forming an entire circumference. Each of the one or more stator segments may comprise a multi-phase winding set wound around or arranged within a tooth of the corresponding stator segment.

[0015] The generator may be a synchronous permanent magnet generator.When the generator is a generator of a wind turbine, the main shaft (at which the plurality of rotor blades are mounted) may be mechanically connected to the generator rotor, optionally via a gearbox.

[0016] Each of the analog Hall sensors may be capable of detecting a magnetic field or flux generated or created at least in part by one of the permanent magnets mounted on the rotor. The Hall sensor may employ the so-called Hall effect to detect the magnetic field or flux. Each of the Hall sensors may continuously provide a corresponding measurement signal in a non-fault state. Thus, each of the measurement signals from the different one or more Hall sensors may be an analog signal and may include a time course of values ​​related to the magnetic flux or magnetic field detected by the corresponding sensor. The measurement signal may, for example, comprise an electrical signal. The Hall sensor may, for example, be mounted at or within one or more air ducts provided in the stator.

[0017] In the case of a faulty sensor, the measurement signal may not be received or may be incorrect. If this is the case, the corresponding Hall sensor can be identified or detected as a faulty sensor. Embodiments of the present invention can detect or identify when one or more of the installed Hall sensors becomes faulty or is faulty. Thus, the method can continuously monitor whether one or more of the Hall sensors is faulty. Based on dynamic monitoring, the selection of measurement signals for determining the rotor angular position can be dynamically updated.

[0018] Evaluating the received measurement signals (under normal conditions with properly functioning sensors, measurement signals from all installed sensors may be received) may involve digitizing and / or analyzing the corresponding time course defined by the measurement signals. The measurement signals may, for example, be evaluated to detect any unexpected behavior or irregularities. Thus, the measurement signals may be analyzed or evaluated individually or in combination, such as by evaluating a binary sum signal, where each bit is associated with a sensor and the bit is set by the individual sensor states. Furthermore, the measurement signals may be evaluated by forming one or more groups comprising measurement signals from different sensor combinations. The evaluation may be based on digitized measurement signals and / or analog measurement signals.

[0019] The selection of the at least one measurement signal can, for example, identify one or more non-faulty (e.g., healthy) sensors whose measurement signals do not exhibit any irregularities or any unexpected behavior (e.g., with respect to the considered time course). At least one measurement signal can be selected, but preferably at least two measurement signals can be selected, or in other embodiments, even three measurement signals of three non-faulty (i.e., properly functioning) sensors can be selected. The higher the number of measurement signals selected, the greater the fault tolerance of the rotor position determination.

[0020] Determining the electrical angular position based on the at least one selected measurement signal may involve performing calculations according to one or more mathematical equations. The equations may also be implemented in the method, or in a circuit or electrical evaluation module. Thus, analog and / or digital implementations of the determination step and also of one or more of the other steps of the method may be provided.

[0021] The evaluation of the measurement signal may, for example, include fault detection and selection based on the evaluation of the measurement signal. The determination of the electrical angle position may be implemented in a closed-loop or open-loop variant. The open-loop approach can be simple but may suffer from noise in the input signal. The closed-loop approach may use a phase-locked loop (DLL) or a synchronous frequency extractor (SFE). A low-pass filter may be integrated to minimize the effects of noise.

[0022] It should be noted that the at least two Hall sensors, and in particular the three Hall sensors, do not need to be installed at specific relative circumferential positions in the stator. These relative positions can in particular vary from a 120-degree offset or a 90-degree offset. This allows for flexible sensor installation, simplifying the method and installation effort.

[0023] Thus, embodiments of the present invention can provide a simple solution for sensor installation and signal processing for angle calculation, while also allowing for fault detection and, therefore, fault-tolerant operation. Unlike approaches such as EMF observers and / or HFI observers, embodiments of the present invention can support a wide range of desired applications and operating conditions and can provide improved accuracy and reliability across the entire speed range. Consequently, there may be no need to provide or install different types of sensors or employ one or more EMF, HFI observer techniques.

[0024] According to an embodiment of the present invention, evaluating the received measurement signals comprises: for each sensor, digitizing the received measurement signals in order to derive a time-correlated individual state pattern, in particular a binary state pattern; evaluating the digitized measurement signals, in particular in a combined manner, in order to select at least one non-faulty sensor and / or deselect at least one faulty sensor.

[0025] A time-dependent individual state pattern derived from a digitized, received measurement signal can, for example, define two or more different states depending on, for example, the negative or positive sign of the analog measurement signal within the considered time course or time interval. The time-dependent individual states can, for example, in the case of binary states, include the definition of state zero and state one for a plurality of time points. For example, state one can be assigned whenever the analog measurement signal is above zero, and state zero can be assigned if the analog signal is below zero. Other assignments may be possible. For example, a properly functioning sensor is expected to provide a measurement signal whose state changes in a substantially periodic manner, at least over a relatively short time span. A faulty sensor can, for example, provide a measurement signal that corresponds to only state one or state zero after a certain time point, which indicates a faulty state pattern when the measurement states of other sensors change. Therefore, evaluating the digitized measurement signal can involve detecting unexpected individual state patterns and / or combined state patterns.

[0026] Any sensor providing a measurement signal whose digitized signal causes the combined state pattern to be irregular or unexpected can be deselected. Thus, effective fault detection can be provided.

[0027] According to an embodiment of the present invention, evaluating the digitized measurement signal includes: combining two or more of the individual status patterns to create a combined status pattern, which is associated with the corresponding sensor group; for the combined status pattern: evaluating the combined status pattern to detect a time point beyond which irregularity or unexpected behavior occurs; indicating a fault; wherein, an irregularity is particularly detected if at least one of the following is true: the combined status pattern deviates from the expected status pattern or the deviation exceeds a threshold; there is an unexpected change in the status pattern over time or the unexpected change exceeds a threshold.

[0028] The digitized measurement signals from all sensors can be formed (combined) to create a combined state pattern. For example, if the individual sensor states are 1, 0 and 1, the formed (combined) state will be 101. A plurality of combined states form a combined state pattern. In other embodiments, less than all measurement signals from all sensors are combined to create a combined state pattern associated with a particular set of measurement sensors. For each sensor set, an expected state pattern can be predicted, and this predicted expected combined state pattern can be compared with the actually achieved combined state pattern of the measurement sensor set. If the (combined) state pattern varies over time or the time variation exceeds a threshold or is unexpected, an irregularity can also be detected without actually comparing it to the expected (combined) state pattern or deriving a difference from the expected (combined) state pattern. Thus, a plurality of different options are provided for detecting faulty sensors.

[0029] According to an embodiment of the invention, evaluating the combined state pattern comprises comparing the combined state pattern with a predetermined reference state pattern for a properly functioning set of sensors to determine a deviation; detecting a fault depending on the deviation; and / or switching selection based on fault detection.

[0030] A reference state pattern, in particular a reference combined state pattern, can be predetermined, for example, based on the installation position(s) of the sensor(s), or in particular the relative angular offset of the installation positions of the sensors under consideration. The method can, for example, involve evaluating the combined state pattern to detect which of the sensors is faulty or to detect that all sensors are healthy and functioning properly. In the selection step, any faulty sensors can be deselected.

[0031] Generally speaking, selection can involve actively selecting any functioning or healthy sensors or deselecting any non-functioning or faulty sensors. Thus, selection can be limited by deselecting specific sensors.

[0032] According to an embodiment of the invention, a sensor under consideration is identified as faulty if a combined state pattern associated with a sensor group comprising the sensor under consideration exhibits irregularities, wherein the measurement signal of the faulty sensor is deselected.

[0033] This allows reliable identification of faulty sensors. Some of the embodiments described above utilize or employ digitized measurement signals and evaluate or analyze them. In other embodiments, analog measurement signals may be analyzed or evaluated, either alone or in combination with the evaluation or analysis of digitized measurement signals according to embodiments of the present invention.

[0034] According to an embodiment of the present invention, evaluating the analog measurement signals comprises: forming several measurement signal groups, each measurement signal group comprising one or more measurement signals originating from one or more sensors; for each measurement signal group, determining the rotor position and / or the rate of change of the rotor position as a grouped rotor position and / or a grouped rotor speed using the one or more measurement signals; assessing the grouped rotor position and / or the grouped rotor speed for at least one irregularity and / or unexpected behavior in order to select at least one measurement signal associated with at least one sensor; determining the rotor position and / or the rate of change of the rotor position based on the selected at least one measurement signal.

[0035] The rotor position and / or the rate of change of the rotor position can be determined using one, two, or three measurement signals. The rate of change of the rotor position can be derived based on the rotor position by performing a differential quotient or a derivative with respect to time. For any properly functioning sensor, the rate of change of the derived rotor position (which is equivalent to the rotor speed) is expected to change only by a slow amount. Therefore, sudden changes in the rotor speed are not expected. However, when considering groups of measurement signals, the rate of change of the derived rotor position changes rapidly, which indicates that one of the sensors included in the group generating the combined measurement signal is faulty. The affected sensor can be identified by evaluating several groups of measurement signals and their associated rotor position and / or rotor speed or rate of change of the rotor speed.

[0036] According to an embodiment of the invention, evaluating the analog measurement signals comprises: comparing the time courses of several groups of rotor positions and / or groups of rotor speeds with one another; deselecting those one or more sensors in each associated group for which irregularities and / or unexpected behavior are observed (for example, which may be an unexpected rate of change of the calculated rotor speed or an unexpected change of the state pattern formed by the digitized measurement signals); determining the rotor position and / or the rate of change of the rotor position based on at least one non-deselected measurement signal.

[0037] This allows reliable identification of faulty sensor(s). The evaluation of the measurement signal can, for example, include checking / verifying the temporal progression of at least one irregularity, in particular a change in the oscillation behavior, an absence or faulty measurement signal, or an amplitude below a threshold value. Other criteria can also be applied.

[0038] According to an embodiment of the present invention, the method comprises: when a fault is detected in one sensor: dynamically switching one or more selected measurement signals to exclude the faulty sensor measurement signal; and determining the rotor position based on the switched one or more selected measurement signals.

[0039] Thereby, dynamic monitoring may be provided, thereby enabling smooth continuation of operation of the generator in the event of a potential fault in one or more sensors.

[0040] According to an embodiment of the invention, three sensors are installed, wherein zero or one or both sensors are faulty or become faulty, or wherein two sensors are installed, wherein zero or one sensor is faulty or becomes faulty, wherein the Hall sensors are installed at arbitrarily different circumferential positions on the stator, in particular spaced apart by electrical angles different than substantially 90° or 120°.

[0041] In other embodiments, more sensors may be installed. These sensors may be installed with known or derivable circumferential angular offsets. The known offsets may, for example, be utilized to derive an expected state pattern or an expected relative measurement signal of the sensor.

[0042] According to an embodiment of the present invention, determining the electrical angular position includes: determining a quadrature signal based on selected measurement signals of two or three or more sensors, in particular using the angular offset of the installation locations of the two sensors; and determining the electrical angular position based on the two quadrature signals.

[0043] The quadrature signal may be a signal corresponding to a hypothetical sensor mounted with a 90° electrical angular offset. The quadrature signal may be determined using a mathematical equation, implemented in electronic circuitry or computer code. Determining the quadrature signal may advantageously enable determination of the electrical angular position according to conventionally known equations.

[0044] According to embodiments of the present invention, at least one of the following applies: determining the electrical angular position comprises normalizing a selected measurement signal; determining the electrical angular position using an open loop or a closed loop; determining the electrical angular position comprises determining and adding an angular offset, in particular based on the stator segment type and / or stator teeth for mounting the sensor and / or the mounting position and / or the sensor position within the air duct.

[0045] Normalizing the selected measurement signal may involve subtracting the mean value from the corresponding measurement signal as determined over a certain time span and dividing the result by the amplitude to produce a signal oscillating around zero and having an amplitude of 1. Open-loop and closed-loop evaluation or determination may be performed in a conventionally known manner.

[0046] Because the sensor isn't necessarily mounted at the "zero" position, an angular offset may need to be determined. According to embodiments of the present invention, the angular offset can be predicted based on the motor design and sensor location. The final angle can then be calculated as the sum of the rotor position, as determined based on the measurement signal, and the angular offset. The offset can be predicted by adjusting for: 1) the selection of stator segment type, 2) the stator teeth used for sensor mounting, 3) the air duct location for skew effects, and 4) the sensor location within the air duct (e.g., left or right end). The offset can then be calculated as the sum of several individual offset values.

[0047] The sensors can be installed in the air ducts of the stator segment teeth. Three sensors or two sensors can be installed at any phase shift, as long as the phase shift is not zero. This allows for flexible selection of sensor installation locations to minimize restrictions or facilitate production.

[0048] According to an embodiment of the present invention, a method for controlling a permanent magnet generator comprising a stator and a rotor having a plurality of mounted permanent magnets is provided, the permanent magnet generator being in particular a permanent magnet generator of a wind turbine, the method comprising: performing the method for determining the electrical angular position of the rotor of the generator according to one of the aforementioned embodiments; and controlling the permanent magnet generator based on the determined electrical angular position of the rotor.

[0049] The control method may perform vector control, which requires the rotor electrical angular position as input.The method may be performed or executed, for example, by a wind turbine controller or a controller of a generator.

[0050] According to an embodiment of the invention, the method is performed during at least one of: during startup; during normal operation at partial or full load to generate electricity; during shutdown; during maintenance involving rotor movement.

[0051] The method is applicable to substantially all operating modes of a typical electrical machine or in particular a typical wind turbine generator.

[0052] It should be understood that the features disclosed, described, explained or adopted for the method of determining the electrical angular position of the rotor, alone or in any combination, are also applicable, alone or in any combination, to the device for determining the electrical angular position of the rotor according to an embodiment of the present invention, and vice versa.

[0053] According to an embodiment of the present invention, a device is provided for determining the electrical angular position of a rotor of a generator comprising a stator and a rotor having a plurality of mounted permanent magnets, the generator being in particular a generator of a wind turbine, the device comprising: an input part adapted to allow reception and / or receiving time-correlated measurement signals from a first and a second analog Hall sensor, the first and second Hall sensors being mounted at different circumferential positions on the stator to detect a magnetic flux at least partially due to one or more of the permanent magnets; a processor coupled to the input part and adapted to: evaluate the received measurement signals; select at least one measurement signal based on the evaluation so as to exclude measurement signals of any faulty sensor; determine the electrical angular position based on the at least one selected measurement signal, the device being in particular further adapted to control the permanent magnet generator based on the determined electrical angular position of the rotor; the device in particular comprising a first Hall sensor and a second Hall sensor.

[0054] The device may be, for example, a module of a wind turbine controller.

[0055] According to an embodiment of the present invention, a generator system, in particular a generator system of a wind turbine, is provided, comprising: a generator comprising a stator and a rotor having a plurality of mounted permanent magnets; and a device according to the aforementioned embodiment connected to control the generator.

[0056] The aspects defined above and further aspects of the invention are apparent from the examples of embodiment to be described hereinafter and are explained with reference to the examples of embodiment.The invention will be described in more detail hereinafter with reference to examples of embodiment but to which the invention is not limited. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 Schematically illustrates an electric machine according to an embodiment of the present invention, the electric machine comprising means for determining the electrical angular position of a rotor in a schematic manner;

[0058] Figure 2 Schematically illustrates a functional diagram of a method for angle calculation adopted according to an embodiment of the present invention;

[0059] Figure 3 Schematically illustrates a stator with installed Hall sensors employed according to an embodiment of the present invention;

[0060] Figure 4 and Figure 5 A graph representing a measurement signal of a Hall sensor is illustrated, including potential evaluation curves or patterns;

[0061] Figures 6 to 11 illustrates a measurement signal from a Hall sensor and the corresponding evaluation steps performed according to an embodiment of the present invention;

[0062] Figure 12 and Figure 13 illustrates simulated measurement signals and potential evaluation methods employed according to an embodiment of the present invention; and

[0063] Figure 14 、 Figure 15 、 Figure 16 The results of the rotor position determination are shown in comparison with a reference angle. DETAILED DESCRIPTION

[0064] exist Figure 1 The generator system 1, schematically illustrated as a functional diagram in FIG, comprises a stator 2 and a rotor 3 (together forming a generator 1 a), the rotor being rotatably supported relative to the stator 2 and comprising a plurality of permanent magnets 4. The rotor is configured to rotate about the stator 2 about an axis of rotation 5. The rotor position is thereby defined by an electrical rotor position 6 (θ) that specifies or represents the electrical position of the rotor 3.

[0065] The generator 1a further comprises a device 7 for determining the electrical angular position 6 (θ) of the rotor of the generator 1 according to an embodiment of the present invention. The device 7 comprises an input section 8 adapted to allow for receiving and / or receiving time-correlated measurement signals 9a, 9b, 9c from first and second analog Hall sensors (labeled with reference numerals 10a, 10b, 10c, respectively). The measurement signals from the different Hall sensors 10a, 10b, 10c are denoted as 9a, 9b, 9c.

[0066] Thus, the first analog Hall sensor 10 a , the second analog Hall sensor 10 b , and the third analog Hall sensor 10 c are mounted at different circumferential positions at the stator 2 to detect magnetic flux at least partially due to one or more of the permanent magnets 4 .

[0067] The device 7 further comprises a processor 11, which is coupled to the input portion 8 and is adapted to evaluate the received measurement signals, in particular in a fault detection module 12, to select at least one measurement signal based on the evaluation and in particular in a selection module 13, so as to exclude measurement signals of any faulty sensor. The processor 11 is further configured to determine the electrical angular position based on the at least one selected measurement signal, in particular using an angle calculation module 14.

[0068] In the illustrated embodiment, the angle determination module 14 determines a plurality of rotor angular positions 15a, 15b, 15c, 15d based on a plurality of different combinations of the received measurement signals 9a, 9b, and 9c. The fault detection module 12 provides information 16 about any potentially faulty sensors or information about healthy or properly functioning sensors to the selection module 13 as input. Based on the information 16 about healthy or faulty sensors, the selection module 13 selects one of the determined plurality of rotor angular positions 15a, 15b, 15c, 15d to obtain a final motor rotor position angle 17 (e.g., also taking into account the angular offset). The final rotor position 17 is supplied to the motor control module 18. The motor control module 18 supplies a control signal 19 to the generator 1a, which includes the stator 2 and the rotor 3, and optionally a converter, based on the angle 17.

[0069] Device 7 allows for fault-tolerant operation of generator system 1. Sensors 10a, 10b, 10c, or potentially more sensors, can be dynamically reconfigured. That is, if a sensor failure is detected, it is removed from service, and two or more other sensors are used to continuously control the motor in a virtually seamless, fail-safe manner. In this sense, device 7 provides a fault-tolerant solution for motor control using analog Hall sensors.

[0070] In the angle calculation module 14 some parallel processing may be scheduled by using eg all three measurement signals 9a, 9b, 9c from the sensors 10a, 10b, 10c.

[0071] In other embodiments, the angle calculation may use three signals (from multiple sensors offset by 120°) to derive an αβ signal for angle calculation. Thus, the αβ signal represents a quadrature signal that may be derived, for example, according to the following formula.

[0072] If three Hall sensors are circumferentially staggered by 120° in a set mounted in the stator, the quadrature signals can be derived as follows:

[0073]

[0074] In other embodiments, the angle calculation may use two signals (from multiple sensors offset by 120°) to derive the αβ signals used for angle calculation. When, for example, only two sensors A and B are used, the quadrature signals (αβ signals) may be calculated as follows:

[0075]

[0076] Figure 2 Schematically illustrates Figure 1 , a functional diagram or scheme of an embodiment of the angle calculation module 14 is shown in FIG. A mapping or transformation module to αβ (denoted by reference numeral 20) receives the measurement signals 9a, 9b, 9c. This mapping or transformation module determines, from two or three of those signals, quadrature measurement signals Sα, Sβ, as explained above. Quadrature signals 21a, 21b (alternatively labeled or represented as Sα, Sβ) are supplied to either an open-loop method 22 or a closed-loop method 23. These modules 22, 23 are capable of determining a (raw or preliminary) rotor position, denoted by reference numeral 17a. An angular offset determination module 24 determines an angular offset 25, for example, based on the configuration of the motor and, in particular, the mounting positions of the Hall sensors 10a, 10b, 10c. An angular offset 25 is added to the output of module 22 or 23 (i.e., the raw rotor position 17a). The result of this addition, using an adding element 26, is the final rotor position 17, which is supplied to the motor control 18.

[0077] Once the two quadrature signals are obtained as described above, the angle can be calculated by using existing techniques such as open-loop calculation or PLL closed-loop calculation, as shown in FIG. Figure 2 The original rotor electrical angle 17a can be calculated, for example, according to the following equation:

[0078]

[0079] Figure 3A portion of a stator 2 of an electric machine according to an embodiment of the present invention is schematically illustrated. The stator includes an air duct 27 and a winding assembly 28. Furthermore, the stator 2 is equipped with three Hall sensors 10a, 10b, and 10c. These Hall sensors 10a, 10b, and 10c are mounted with an electrical angle offset (also referred to as δ), which can be 120°, 90°, or a value different from 90° and also different from 120° in electrical phase angle. The Hall sensors 10a, 10b, and 10c are suitable for use according to embodiments of the present invention.

[0080] Figure 4 and Figure 5 Three Hall sensors (such as Figure 3 Measured signals 9a, 9b, 9c of Hall sensors 10a, 10b, 10c) mounted on the stator 2 are shown in the graph, with time as the abscissa and the signal amplitude as the ordinate. Figure 4 The (reference) curve 30 in FIG. 1 represents a combined state pattern that is calculated by combining the individual state patterns of the measurement signals 9 a, 9 b, 9 c of the sensors. The individual state patterns can be derived, for example, from the analog measurement signals 9 a, 9 b, 9 c, with state 1 being assigned to time spans in which the analog signal is above zero and state zero being assigned to time spans when the analog signal is below zero. Figure 4 The expected state pattern 30 is shown, which is expected when all sensors are healthy. When the sensors are healthy, the expected state pattern 30 corresponds to the sequence 546231 (the vertical axis on the right hand side indicates the ordinate, which indicates the values ​​in the state pattern 30).

[0081] Figure 5 The diagram shows a scenario in which the sensor signal 9b of the sensor 10b degrades (e.g., shorts to a low fault) at and beyond the time point 31. Curve 32 indicates the combined state pattern for those three sensors 10a, 10b, 10c, according to which the sensor 10b is damaged at the time point 31. It can be understood that after the time point 31, the corresponding state pattern 32 exhibits irregularities because it differs from the state pattern 32 shown in FIG. Figure 4 Expected state pattern 30 is illustrated in . After time point 31 , the state pattern assumes profile 640 and it is detected that a fault has occurred at time point 31 .

[0082] When the sensor signal 9b is positive, a fault occurs (for example, a short circuit to low fault), and in this case the fault is detected immediately. If the fault occurs when the sensor signal should be negative, up to 180 degrees of electrical rotation may be required before the fault is asserted.

[0083] According to embodiments of the present invention, dual or triple sensor solutions are provided, with arbitrary phase shifts between the sensor installation locations. Two Hall sensors can be installed, with an angular displacement of 120° or any value of δ, as discussed earlier. The angle calculation techniques based on quadrature αβ signals can still be used. These αβ signals can be derived from the two sensor measurements, as explained below.

[0084]

[0085] Obviously, if the displacement δ is zero or too small, the mapping will not work.

[0086] Figure 6 、 Figure 7 、 Figure 8 The sensor measurement signal and the resulting signal are illustrated when the sensor 10b experiences a fault (eg, a short circuit to low fault) when the corresponding measurement signal is positive, and the corresponding Figure 9 、 Figure 10 、 Figure 11 A similar signal and the resulting signal are illustrated in the case of a fault in the sensor 10b when the measurement signal is negative.

[0087] Figure 6 Curve 33 in the diagram illustrates the true angle. Curve 38 indicates the Figure 8 The motor is controlled based on the control angle selected from one of the four signals in the control angle. When the sensor signal 9b of the sensor 10b is positive, as shown in FIG. Figure 7 It can be seen that a failure occurs at time point 31.

[0088] Then, Figure 8 The electrical angle when the measurement signals 9a, 9b, 9c are used for determination is illustrated as curve 34, curve 35 illustrates the electrical rotor position when the measurement signals 9a, 9b are used, curve 36 represents the electrical rotor position when the measurement signals 9a, 9c are used, and curve 37 illustrates the rotor position when the measurement signals 9b, 9c are utilized for determination.

[0089] Select the module (e.g. Figure 1 The selection module 13 in ) can for example regard the determined angles as being derived from different combinations of the measurement signals, for example as being derived from different combinations of the measurement signals, which are used to determine the curves 34, 35, 36, 37:

[0090]

[0091] As can be appreciated, according to an embodiment of the present invention, when all measurement signals 9a, 9b, 9c are healthy, Figure 1The selection module 13 shown in FIG selects the electrical angle from the curve 34; at the moment of the fault occurring at time 31, the fault is immediately detected and then only the measurement signals 9a, 9c for angle determination that generate the curve 36 will be selected as the electrical angle for control, which is Figure 6 The actual angle 33 shown in FIG. 3 matches perfectly.

[0092] Therefore, when the sensor 10b is faulty when the corresponding measurement signal is positive, the fault can be immediately identified and the angle used for control can be mapped from the three-sensor solution to the two-sensor solution, and there will be no error in the angle.

[0093] In such Figure 9 、 Figure 10 、 Figure 11 In the scenario illustrated in , the same sensor failure occurs in measurement signal 10b when the corresponding measurement signal is negative. Figure 9 In FIG. 3 , curve 33 illustrates the actual angle, and curve 38 illustrates the controlled angle.

[0094] The fault in the second measurement signal 9b occurs at the time point 31. However, the detection of the state pattern 32 exhibiting irregular behavior is detected only at a time point 39 later than the time point 31 at which the fault occurred. Figure 11 The curves 34, 35, 36, 37 in FIG. 3 illustrate the rotor electrical angle when all three measurement signals or the measurement signals 9a, 9b or the measurement signals 9a, 9c or the measurement signals 9b, 9c, respectively, are utilized for the angle determination. Figure 9 、 Figure 10 、 Figure 11 In the scenario illustrated in FIG, a fault occurring at time 31 can be detected with some delay, and the control angle may have an error within 180° of the rotor electrical motion until detection / correction occurs. Before the fault is detected, control angle 38 should be an angle selected from curve 34. During the fault detection period from 31 to 39, curve 34 is still used as control angle 38, but with an angle error. After fault detection, the selection for control angle 38 is switched to curve 36, which is determined based on the two healthy sensor measurement signals 9a, 9c and does not have an error. The maximum control angle error will be approximately 30°, which is still relatively low to allow continuous operation of the motor control.

[0095] However, there is still potential for further improvement in fault detection and response within the maximum 180° window of rotor motion in terms of electrical angle.

[0096] Figure 12 and Figure 13The true rotor position is shown as curve 33, the rotor position calculated using all three measurement signals is shown as curve 34, the rotor position calculated using measurement signals 9a, 9b is shown as curve 35, the rotor position calculated using measurement signals 9a, 9c is shown as curve 36, and the rotor position calculated using measurement signals 9b, 9c is shown as curve 37 in a coordinate system having an abscissa indicating time and having coordinates indicating angle and speed, respectively. Figure 13 In FIG. 3 , curve 34 ′ indicates the speed using all three measurement signals, curve 35 ′ indicates the speed using measurement signals 9 a , 9 b , curve 36 ′ indicates the speed using measurement signals 9 a , 9 c , and curve 37 ′ indicates the speed when measurement signals 9 b , 9 c are used.

[0097] Therefore, according to an embodiment of the present invention, during the period of fault detection, an angle may be calculated by using the filtered velocity, and this angle will be used for control until the sensor reconstruction has been completed. In other embodiments, additionally or alternatively, the control angle may be selected by rationalizing the four formed Hall sensor angles, and one of the simple algorithms is used to detect any sudden rate change of the calculated angle (e.g., the calculated velocity), such as Figure 12 、 Figure 13 As shown in the figure.

[0098] Once in motion, the motor control can be contained even when using a single sensor signal, i.e. the system will tolerate a double sensor failure in a three sensor solution or a single sensor failure in a two sensor solution. Thus, continuous operation in a fault tolerant mode is provided. This can be achieved by using a 90° phase shifter and then using a PLL observer (e.g. Figure 2 This can be simply done by capturing the peak value of the signal (as shown in the figure). For example, given a healthy signal Sx and its peak value Speak, the angle will be calculated by the following formula:

[0099]

[0100] The offset angle may be associated with a health sensor.

[0101] Figure 14 、 Figure 15 、 Figure 16 The diagram is shown in a coordinate system having an abscissa indicating time and an ordinate indicating: Figure 14 Sensor signals 40 and 41 in Figure 15 The resulting quadrature signals 42 and 43, Figure 16Graphs against angles in FIG. 4 (where 44 indicates the angle determined from the two Hall sensor signals, and 45 indicates the angle observed by the EMF observer).

[0102] Will Figure 16 The determined angle 44 in is compared with a curve 45 indicating the electrical angle as determined by the EMF observer. Figures 14 to 16 The curves shown in Figure 1 were obtained in generating mode, with two sensors mounted on the stator, offset by 60°. The two sensor signals are mapped into orthogonal αβ signals. The angle can be calculated using a simple open-loop method.

[0103] The angle calculated from the Hall sensor signals 40 and 41 is shown as Figure 16 44 in the figure. However, it should be noted that there is some fluctuation in the angle calculated for the Hall sensors, because the two Hall sensor measurement signals have not been normalized. However, when normalization is performed, the quality can be further improved. Furthermore, when the motor is in motion, harmonic reduction (i.e., harmonic reduction at 6f) may be possible by using some existing techniques. When three Hall sensors are installed with arbitrary phase shifts between them, the following can be obtained: Figures 14 to 16 Similar results are shown in .

[0104] Embodiments of the present invention may provide the following embodiments or advantages:

[0105] - Application of analog Hall sensors for wind turbine control during shutdown, low speed and high speed for potentially all operating modes,

[0106] - The sensors used are inexpensive, avoid moving parts and do not require mechanical coupling to the motor shaft and are well suited for stators and direct-drive generators.

[0107] - The analog Hall sensors provide the absolute electrical angle of the motor and are thus ready for use or control.

[0108] -Adapts to flexible sensor positions, thus suitable for solutions with two or three sensors

[0109] - Derived from two or three signal feedbacks with arbitrary phase shifts

[0110] - Fault-tolerant operation for single component failure in a dual analog sensor solution or up to two component failures in a three-sensor solution

[0111] -A simple technique for fault detection in dynamic selection of control angles

[0112] - A technique for determining angular offset based on sensor position without requiring rotation testing.

[0113] It should be noted that the term "comprising" does not exclude other elements or steps, and "a" or "an" does not exclude a plurality. Moreover, elements described in connection with different embodiments may be combined. It should also be noted that reference signs in the claims should not be construed as limiting the scope of the claims.

Claims

1. A method for determining the electrical angular position (17) of a rotor of a generator (1a) comprising a stator (2) and a rotor (3) with a plurality of mounted permanent magnets (4), the generator being in particular a generator of a wind turbine, the method comprising: allowing to receive and / or receiving at least time-correlated measurement signals (9a, 9b, 9c) from a first analog Hall sensor (10a), a second analog Hall sensor (10b) and optionally a third analog Hall sensor (10c), said Hall sensors being mounted at different circumferential positions at said stator (2) to detect a magnetic flux at least partially due to one or more of said permanent magnets; evaluating the received measurement signals (9a, 9b, 9c); selecting at least one measurement signal based on the evaluation in order in particular to exclude measurement signals of any faulty sensors; The electrical angular position (17a, 17) is determined based on at least one selected measurement signal.

2. The method according to the preceding claim, wherein Evaluating the received measurement signals (9a, 9b, 9c) comprises: For each sensor (10a, 10b, 10c), the received measurement signal is digitized in order to derive a time-dependent individual state pattern, in particular a binary state pattern; The digitized measurement signals are evaluated, in particular in combination, in order to select at least one non-faulty sensor and / or to deselect at least one faulty sensor.

3. The method according to the preceding claim, wherein Evaluating the digitized measurement signal includes: combining two or more of the individual state patterns to create a combined state pattern (32) associated with a corresponding sensor group; For the combined state mode: evaluating the combined state pattern (32) to detect a point in time (31) beyond which irregularities or unexpected behavior occur; Indicates a fault; Therein, an irregularity is specifically detected if at least one of the following holds true: The deviation of the combined state pattern from the expected state pattern (30) exceeds a threshold value; A temporal variation of the state pattern exceeds a threshold.

4. The method according to the preceding claim, wherein Evaluating the combined state pattern (32) includes: comparing the combined state pattern to a predetermined reference state pattern (30) for a properly functioning set of sensors to determine a deviation; Detecting a fault based on the deviation; and / or The selection is switched based on a result of the deviation or fault detection.

5. The method according to any one of the preceding claims 3 or 4, wherein If all combined state patterns associated with the sensor group containing the sensor under consideration exhibit irregularities, the sensor under consideration is identified as faulty, In this case, the measurement signal of the defective sensor is deselected.

6. The method according to any one of the preceding claims, wherein Evaluating the analog measurement signal comprises: forming several measurement signal groups, each measurement signal group comprising one or more measurement signals originating from one or more sensors; for each measurement signal group, determining the rotor position and / or the rate of change of the rotor position as a grouped rotor position (34, 35, 36, 37) and / or a grouped rotor speed (34', 35', 36', 37') using the one or more measurement signals; evaluating the set of rotor positions and / or the set of rotor speeds for at least one irregularity and / or unexpected behavior in order to select at least one measurement signal associated with at least one sensor; The rotor position and / or the rate of change of the rotor position is determined based on the selected at least one measurement signal.

7. The method according to the preceding claim, wherein Evaluating the analog measurement signal comprises: comparing the time courses of several groups of rotor positions (34, 35, 36, 37) and / or groups of rotor speeds (34', 35', 36', 37') with one another; deselecting those one or more sensors in each associated group for which irregularities and / or unexpected behavior are observed; The rotor position and / or the rate of change of the rotor position is determined based on the non-deselected at least one measurement signal.

8. The method according to any one of the preceding claims, comprising: When a fault is detected in a sensor: dynamically switching the selected one or more measurement signals to exclude the faulty sensor measurement signal; The rotor position is determined based on the switched selected one or more measurement signals.

9. The method according to any one of the preceding claims, in, Three sensors (10a, 10b, 10c) are installed, of which zero, one, or two sensors are or become faulty, or wherein two sensors (10a, 10b) are installed, wherein zero or one of the sensors is or becomes faulty, and / or Therein, the Hall sensors are mounted at any different circumferential positions on the stator, in particular spaced apart by 90° or 120° or by electrical angles different from values ​​of substantially 90° or 120°.

10. The method according to any one of the preceding claims, wherein Determining the electrical angle position includes: Determining a quadrature signal (S_α, S_β) from selected measurement signals of two or three or more sensors, in particular using an angular offset of the installation locations of the two sensors; The electrical angular position is determined based on the two quadrature signals.

11. The method according to any one of the preceding claims, wherein At least one of the following applies: Determining the electrical angular position includes normalizing the selected measurement signal; Determine the electrical angular position using an open loop or a closed loop; Determining the electrical angular position comprises determining and adding an angular offset, in particular based on the stator segment type and / or stator tooth for mounting the sensor and / or the mounting position and / or the sensor position in the air duct.

12. A method of controlling a generator (1a) comprising a stator (2) and a rotor (3) having a plurality of mounted permanent magnets (4), the generator being in particular a generator of a wind turbine, the method comprising: Execution of a method for determining the electrical angular position (17) of a rotor of a generator according to any one of the preceding claims; The permanent magnet generator (1a) is controlled based on the determined rotor electrical angular position (17).

13. The method according to the preceding claim, which is performed during at least one of the following: During startup; During normal operation at partial or full load to generate electricity; During downtime; During maintenance involving rotor movement.

14. A device (7) for determining the electrical angular position of a rotor of a generator comprising a stator and a rotor with a plurality of mounted permanent magnets, the generator being in particular a generator of a wind turbine, the device comprising: an input portion (8) adapted to allow receipt and / or receive time-correlated measurement signals (9a, 9b, 9c) from a first analog Hall sensor (10a) and a second analog Hall sensor (10b), the first and second analog Hall sensors being mounted at different circumferential positions at the stator (2) to detect a magnetic flux at least partially due to one or more of the permanent magnets; a processor (11) coupled to the input portion and adapted to: evaluating the received measurement signals (9a, 9b, 9c); selecting at least one measurement signal based on the evaluation so as to exclude measurement signals of any faulty sensors; determining the electrical angular position (17) based on at least one selected measurement signal, the device being in particular further adapted to control the permanent magnet generator based on the determined electrical angular position of the rotor; The device particularly comprises the first Hall sensor and the second Hall sensor.

15. A generator system (1), in particular a generator system of a wind turbine, comprising: stator (2); a rotor (3) having a plurality of mounted permanent magnets (4); Device (7) according to the preceding claim, connected to control the generator.

Images