Obtaining a load angle in an electric motor

Abstract

A method of obtaining a load angle in an electric motor having a stator, a salient pole rotor, and an airgap monitoring system is described. The method comprises: determining an airgap profile; determining a magnetic flux angle evolution, determining a magnetic-flux-based reference oscillation signal; determining a phase shift between the periodic oscillation of the airgap profile and the magnetic-flux-based reference oscillation signal; and determining the load angle from the phase shift. Further, a method of operating an electric motor is described. The method comprises obtaining a load angle and controlling the electric motor by a magnetic flux control scheme such that the load angle remains within a predetermined load angle tolerance range. Further, an electric motor comprising: a stator; a salient pole rotor; an airgap monitoring system; and a monitoring- and control system is described.

Description

[0001] OBTAINING A LOAD ANGLE IN AN ELECTRIC MOTOR

[0002] Technical field

[0003] The disclosure relates to a method of obtaining a load angle in an electric motor, particularly a gearless mill drive. Further, the disclosure relates to a method of operating an electric motor and a corresponding electric motor.

[0004] Technical Background

[0005] Usually, electric salient pole synchronous motors comprise a rotor and a stator. An encoder can be used for some applications to obtain the physical rotor position from which a load angle can be derived. Knowing the load angle of the rotor in the stator can be crucial for operating the electric motor. For example, a driving current of the electric motor can be adjusted according to the load angle to improve control accuracy. Monitoring the margin to the pull-out load angle can allow for higher output torque and / or higher efficiency by reducing flux depending on the load angle conditions.

[0006] Mounting an encoder in an electric motor can be difficult, especially in harsh environmental conditions. For example, mounting an encoder in a ring motor, particularly a ring motor with a diameter in between 4m and 13m, can be difficult since there is no shaft on which the encoder can be mounted. Further, the accuracy of an encoder decreases with increasing diameters of the rotor since effects such as thermal expansion or vibration can become more dominant. In addition, the high number of poles of ring motors can amplify mechanical inaccuracies in the electrical reference frame e.g. by a factor of 40 for an 80-pole motor. Hence, there is a need for obtaining a load angle of an electric motor without using an encoder.

[0007] Description of invention

[0008] According to an aspect of the disclosure, an electric motor has a stator, a salient pole rotor and an airgap monitoring system.

[0009] According to a first aspect of the disclosure, a method of obtaining a load angle in an electric motor having a stator, a salient pole rotor, and an airgap monitoring system, comprises: determining an airgap profile by the airgap monitoring system, wherein the airgap profile is based on an airgap width, at at least one stationary angular position of the stator, between the stator and the salient pole rotor over time, and wherein the airgap profile comprises a periodic oscillation having a fundamental frequency due to a periodicity of a pole structure of the electric motor; determining a magnetic flux angle evolution, wherein the magnetic flux angle evolutionis indicative of an angular orientation of a magnetic flux vector over time, whereby the magnetic flux vector represents the magnetic flux induced by the stator in a stationary electromagnetic reference frame of the stator; determining, from the magnetic flux angle evolution, a magnetic-flux-based reference oscillation signal having the fundamental frequency; determining a phase shift between the periodic oscillation of the airgap profile and the magnetic-flux-based reference oscillation signal; and determining the load angle from the phase shift.

[0010] According to a second aspect of the disclosure, a method of operating the electric motor comprises obtaining a load angle of the electric motor according to the first aspect of the disclosure and controlling the motor by a magnetic flux control scheme such that the load angle remains within a predetermined load angle tolerance.

[0011] According to a third aspect of the disclosure, an electric motor comprises a stator, a salient pole rotor, an airgap monitoring system and a monitoring- and control system with a memory unit. The memory unit stores a program code, which when executed by the monitoring- and control system causes the monitoring- and control system to perform a method according to the first and / or second aspect of the disclosure.

[0012] In view of the above, embodiments of the present disclosure aim to provide a method of obtaining a load angle in an electric motor, particularly without using an encoder, a method of operating an electric motor and an electric motor. According to an aspect of the disclosure, the method of obtaining the load angle in the electric motor and the method of operating the electric motor can be executed without the use of an encoder, thus avoiding the use of an encoder and the above-mentioned disadvantages. Further, according to the aspects of the disclosure a load angle of an electric motor can be obtained with high accuracy. For example, for an electric motor with a diameter larger than 4 m, a method according to an aspect of the disclosure can obtain the load angle of the electric motor with a higher accuracy than the accuracy achievable for such a motor with a conventional encoder.

[0013] According to an aspect of the disclosure, the salient pole rotor can comprise poles. The poles can be arranged equally distributed, particularly periodically, around the circumference of the rotor. The airgap profile can be based on the shape of the pole(s) and / or on the rotational speed of the rotor, particularly the rotational speed of the rotor relative to the stator.

[0014] According to an aspect of the disclosure, the oscillation can reflect a periodicity of the rotor’s pole structure, particularly while the rotor is rotating at the rotor’s rotational speed relative to the stator. Thus, the fundamental frequency can reflect the periodicity of the rotor’s pole structure. In an embodiment, the fundamental frequency reflects the fundamental periodicity of the rotor’s pole structure (e.g., in units 1 / rad), multiplied by the rotor’s rotational speed,particularly relative to the stator, (e.g., in units rad / sec). For example, the fundamental frequency can be in time units of 1 / sec.

[0015] According to an aspect of the disclosure, a magnetic flux angle evolution can be determined by determining, e.g. estimating, the inner stator voltage and / or measuring the terminal voltage of the electric motor. For example, effects such as an ohmic winding resistance and / or stray impedances of the electric motor can be estimated enabling to estimate the inner stator voltage from the terminal voltage.

[0016] According to an embodiment of the disclosure, the load angle can be determined dynamically. Determining the load angle dynamically can be understood as determining the load angle during operation of the electric motor, particularly fast enough to allow motor control using the determined load angle. This way, the control may allow for example for corrective action to avoid instability. Determining the load angle of an electric motor dynamically can enable real time adjustments, such as adjusting magnitude and / or angle of the stator flux of the electric motor and / or adjusting the rotor current of the electric motor.

[0017] According to an aspect of the disclosure, the airgap monitoring system can comprise at least one airgap monitoring unit. Each airgap monitoring unit can have a distance sensor for measuring the airgap width between the stator and the salient pole rotor at a respective stationary angular position of the stator.

[0018] According to an aspect of the disclosure, each airgap monitoring unit can be mounted on the stator of the electric motor. Each airgap monitoring unit can measure the airgap width between the stator and the salient pole rotor individually.

[0019] According to an aspect of the disclosure, the airgap monitoring system can comprise either one or a plurality of airgap monitoring units and a phase shift between the periodic oscillation of the airgap profile and the magnetic-flux-based reference oscillation signal can be determined for the airgap profile(s) determined by at least two, particularly individually for each, of the airgap monitoring unit(s).

[0020] According to an aspect of the disclosure, the airgap monitoring system can comprise, at least one, particularly at least two, airgap monitoring unit(s). Typically, the airgap monitoring system can comprise at least 8, preferably 15, airgap monitoring units.

[0021] The plurality of airgap monitoring units can create a redundancy. For example, if some of the airgap monitoring units fail, an airgap profile can still be obtained by the remaining functioning airgap monitoring unit(s). Further, the accuracy of the obtained load angle can be increased by determining a plurality of phase shifts between a plurality of periodic oscillations of the airgap profiles determined by the plurality of airgap monitoring units.According to an aspect of the disclosure, determining the magnetic flux angle evolution can comprise calculating, from a terminal voltage of the electric motor, a magnetic flux vector in the stationary electromagnetic reference frame. Particularly, the magnetic flux vector can be rotating. Particularly, the magnetic flux vector can be calculated from a (measured) terminal voltage and / or from an (estimated) inner stator voltage.

[0022] The inner stator voltage can be estimated based on the physical motor properties and the actual operation point of the motor from the terminal voltage. Particularly, the terminal voltage can be measured and / or known from an electric motor control.

[0023] According to an aspect of the disclosure, the magnetic-flux-based reference oscillation signal can comprise two magnetic-flux-based reference oscillation terms. Typically, a first magnetic-flux-based reference oscillation term of the two magnetic-flux-based reference oscillation terms and a second magnetic-flux-based reference oscillation term of the two magnetic-flux-based reference oscillation terms can be orthogonal, particularly they can be phase shifted with respect to each other by 90°, as is the case with a cos and a sin function sharing the same periodicity. Together, they can span an orthogonal coordinate system, rotating synchronously with the frequency of the magnetic-flux-based reference oscillation signal.

[0024] In an equivalent implementation, the two orthogonal oscillation terms can be represented by a complex-number oscillation signal (e.g., as real and imaginary parts of the complex-number oscillation signal).

[0025] According to an aspect of the disclosure, a periodicity of the magnetic flux angle evolution can correspond to half the periodicity of the pole structure. Determining the magnetic-flux-based reference oscillation signal can include: determining, from the magnetic flux angle evolution, a magnetic flux vector oscillation frequency of the angular orientation of the magnetic flux vector; and generating the two magnetic-flux-based reference oscillation terms as a periodic signal having twice the magnetic flux vector oscillation frequency.

[0026] According to an aspect of the disclosure, determining the phase shift can comprise determining two products by multiplying an airgap profile dependent term with each of the two magnetic-flux-based reference oscillation terms. According to an aspect of the disclosure, determining the phase shift can comprise determining two products by individually multiplying an airgap profile dependent term, the airgap profile dependent term derived by subtracting a compensation term from the airgap profile, with the two components of the magnetic-fluxbased reference oscillation signal.Generally, the airgap profile dependent term may be the airgap profile itself, or a term derived from the airgap profile. In an example, the airgap profile dependent term may be obtained by subtracting a compensation term from the airgap profile.

[0027] Thus, determining the phase shift can comprise determining two products by multiplying the periodic oscillation of the airgap profile, after subtracting a compensation term, with each of the two magnetic-flux-based reference oscillation terms. Thus, each of the two products can be obtained by multiplying the periodic oscillation with a respective one of the two magnetic-fluxbased reference oscillation terms. Thus, a zero-frequency component of the product can be obtained under the condition that both factors of the multiplication have a common frequency component, and the phase shift of these frequencies is different to 90 degrees.

[0028] According to an aspect of the disclosure, determining the phase shift can further comprise low-pass filtering of the two products. Low-pass filtering of the two products can either consist of integration, amplifying low frequency components or by attenuative low pass filters. Particularly, a zero-frequency offset term (DC-term) of the two products can be obtained and / or amplified by this step.

[0029] According to an aspect of the disclosure, determining the phase shift between the periodic oscillation of the airgap profile and the two magnetic-flux-based reference oscillation terms can further comprise performing a closed loop compensation routine. The closed loop compensation routine can compensate a constant DC-term in the airgap profile.

[0030] Advantageously, low-pass filtering the two products by integration can amplify (obtain) the DC-term, which may result in an improved closed loop compensation routine.

[0031] According to an aspect of the disclosure, the closed loop compensation routine can comprise: low-pass filtering the two products by (separately) integrating the two products over time; and subtracting the products of the integrator outputs and the corresponding components of the magnetic-flux-based reference oscillation signal from the airgap profile. For the case that a low pass filter is used, which does not affect the magnitude of a DC-term, the subtraction of the integrator outputs and the corresponding components of the magnetic-flux-based reference oscillation signal from the airgap profile can be trivial since the algorithm could work without feedback loop at the cost of higher ripple which requires longer filtering times.

[0032] Typically, a DC-term, representing the average airgap can be removed by an additional DC-closed loop compensation routine.

[0033] According to an aspect of the disclosure, the phase shift can be determined from compensation terms derived from the closed loop compensation routine. Particularly, the phase shift can be determined from the amplitudes of compensation terms derived from the closed loopcompensation routine Particularly, the phase shift can be determined from the ratio of the amplitudes of two compensation terms derived from the closed loop compensation routine. According to an aspect of the disclosure, an electric motor calibration can be used to determine the load angle as a function of the phase shift between the periodic oscillation of the airgap profile and the magnetic-flux-based reference oscillation signal.

[0034] The electric motor calibration can depend on one or multiple factors, particularly on the rotor position during activation and / or the mounting position(s) of the airgap monitoring unit(s) of the airgap monitoring system. The electric motor calibration can depend on the temperature and / or the thermal expansion of the electric motor. Furthermore, load angle dependent saturation of the rotor poles can influence the dependency between mechanical rotor position and electrical load angle. For example, the electric motor calibration can be determined by operating the electric motor at a known load angle e.g. 0°, or at a plurality of load angles and / or at different conditions such as different stator temperatures. Thereby, a calibration curve outputting the load angle as a function of the phase shift and optionally the temperature and / or other parameters (such as saturation effects, particularly saturation effects of the magnetic flux at the tip of the poles) can be obtained by any known calibration technique. During the method of obtaining the load angle, a temperature dependent correction signal and / or a correction signal depending on other parameters related to the actual operation point of the motor can be added to the load angle depending on the actual temperature and / or operation point.

[0035] According to an aspect of the disclosure, the electric motor can be a ring type electric motor. Typically, a ring type electric motor comprises a ring-shaped stator and a ring-shaped salient pole rotor.

[0036] According to an aspect of the disclosure, the electric motor can belong to a gearless mill drive. According to an aspect of the disclosure, the electric motor can have a diameter of more than 1 m. In some embodiments the electric motor can have a diameter of more than 4 m. In some embodiments the electric motor can have a diameter between 4 m and 13 m. Typically, the diameter of an electric motor can be determined from the outer circumference of the salient pole rotor, from the inner diameter of the stator and / or from the outer diameter of the stator. According to an aspect of the disclosure, the airgap can be filled with any fluid e.g. air and / or oil.

[0037] According to an aspect of the disclosure, the method of obtaining a load angle in an electric motor without an encoder can be specifically beneficial for low-speed and / or high torque electric motors, e.g. a rotor speed lower or equal to 18 rpm and / or a torque higher or equal to 3000 kNm, particularly for a rotor speed ranging from 8 rpm to 18 rpm and / or a torque rangingfrom 3000 kNm to 31000 kNm. The electric motor can be a ring-type motor with a ring-shaped stator and a ring-shaped salient pole rotor. These characteristics can be particularly beneficial for a gearless mill drive.

[0038] According to an aspect of the disclosure, the electric motor may be controlled by a magnetic flux control scheme using the load angle. For example, the control scheme may comprise comparing the determined load angle with a predetermined load angle tolerance range (which may be defined by an upper threshold and / or a lower threshold) and taking corrective action for keeping the load angle within the tolerance range. Thereby, it is ensured that the load angle remains within the predetermined load angle tolerance range.

[0039] According to an aspect of the disclosure, the predetermined load angle tolerance range can be dynamically defined according to the operation and / or the physical properties of the electric motor.

[0040] Advantageously, the magnetic flux induced by the salient pole rotor and / or stator of the electric motor can be reduced, resulting in improved energy efficiency of the electric motor. For control methods based on the stator flux, particularly which don’t comprise of any feedback of the mechanical rotor position (e.g. encoder), information about the load angle can improve torque-, speed-, and / or position control accuracy. By monitoring the load angle, the control algorithm may operate the motor closer to the stability limit and achieve higher torque.

[0041] Brief description of the Figures

[0042] In the following, examples according to aspects of the disclosure are described in more detail with reference to the drawings. Therein:

[0043] Figure 1 a schematic illustration of an electric motor;

[0044] Figure 2 a diagram illustrating the method of obtaining a load angle;

[0045] Figure 3a - c three graphs representing an airgap profile and a compensation term; the two magnetic-flux-based reference oscillation terms; and a simulated and an obtained load angle; Figure 4 an illustration of an implementation of the method for obtaining a load angle.

[0046] Detailed description of the figures

[0047] Figure 1 illustrates an electric motor comprising a stator 110 and a salient pole rotor 120. The electric motor can be a ring type electric motor comprising a ring-shaped stator 110 and a ringshaped salient pole rotor 120.The stator 110 can comprise stator windings 111 to induce a magnetic field. The salient pole rotor 120 can comprise a plurality of poles 121. The plurality of poles 121 can be arranged periodically on the salient pole rotor 120.

[0048] Further, the electric motor comprises an airgap monitoring system to determine an airgap profile. The airgap monitoring system can comprise an airgap monitoring unit 113 to determine the airgap profile. The airgap profile measured by the airgap monitoring unit 113 is based on an airgap width 115, at a stationary angular position of the stator 110, between the stator 110 and the salient pole rotor 120. The airgap monitoring system can comprise a plurality of airgap monitoring units 113 to determine a plurality of airgap profiles. Each airgap profile measured by the airgap monitoring units 113 may for example be based on an airgap width, at a respective stationary angular position of the stator 110, between the stator 110 and the salient pole rotor 120.

[0049] The airgap profile may be a combined (e.g., average) value of the airgap widths obtained by the airgap monitoring units 113, and in this sense may be based on, or indicative of, the airgaps at the stationary positions associated with the airgap monitoring units 113. It might be necessary to account for the phase relations between the airgap profiles obtained by different airgap monitoring units 113, e.g. by mounting the airgap monitoring units 113 on the stator 110 at positions having the same phase with respect to the rotor. It might be necessary to do a phase alignment, since a phase shift e.g. of 180° could double the fundamental frequency in the averaged signal.

[0050] Alternatively, a plurality of the airgap monitoring units 113 may be used for obtaining a plurality of individual airgap profiles, each one obtained by a respective one of the airgap monitoring units 113. These individual airgap profiles may be processed separately in the subsequent step(s) as described below. The load angle may be derived by circularly averaging the individually calculated load angles. This approach has the advantage of being insensitive to phase shifts between the measurements from the different airgap monitoring units 113.

[0051] The electric motor can comprise a monitoring system with a memory unit. The memory unit can store a program code, which when executed by the monitoring system causes the monitoring system to perform a method of obtaining a load angle and / or operating the electric motor according to any aspect of the disclosure.

[0052] Figure 2 shows a diagram illustrating the method of obtaining a load angle according to an aspect of the disclosure. The method of obtaining a load angle comprises; determining 201 an airgap profile; determining 203 a magnetic flux angle evolution; determining 205 a magnetic-flux-based reference oscillation signal; determining 207 a phase shift between the periodicoscillation of the airgap profile and the magnetic-flux-based reference oscillation signal; and determining 209 the load angle from the phase shift.

[0053] Determining 201 an airgap profile by the airgap monitoring system can comprise measuring, by at least one airgap monitoring unit of the airgap monitoring system, an airgap between the stator and the salient pole rotor at at least one angular position of the stator.

[0054] Determining 203 a magnetic flux angle evolution can comprise determining a magnetic flux vector. The magnetic flux vector can be determined based on the inner terminal voltage of the electric motor.

[0055] Determining 207 a phase shift between the periodic oscillation of the airgap profile and the two magnetic-flux-based reference oscillation terms of the magnetic-flux-based reference oscillation signal can be performed for a single airgap profile determined by the airgap monitoring system, for an averaged airgap profile determined by at least two airgap monitoring units of the airgap monitoring system and / or for a plurality of airgap profiles determined by a plurality of airgap monitoring units of the airgap monitoring system.

[0056] Determining 209 a load angle from the phase shift can be performed for a phase shift of a single airgap profile determined from the airgap monitoring system, for a mean phase shift of a plurality of airgap profiles determined from the airgap monitoring system and / or for a plurality of phase shifts of a plurality of airgap profiles determined from the airgap monitoring system. If a plurality of phase shifts of a plurality of airgap profiles are determined, a mean load angle can be determined from the plurality of determined load angles.

[0057] Figure 3a shows a graph of a simulation with time on the x-axis. An airgap profile 301 as a function of the airgap (in units of length e.g. mm) over time is displayed. A compensation term 325 is plotted over time. During the starting process of the method (starting at 0 sec) the compensation term 325 converges to approximate the airgap profile 301 as a sinusoidal wave with a superimposed DC-offset. The compensation term 325 comprises two orthogonal and sinusoidal components with pole passing frequency and a DC-offset. The sum of both orthogonal and sinusoidal components results to a sinusoidal oscillation with the same phase and magnitude as the fundamental component in the airgap profile 301. The DC-offset of the compensation term 325 has the same magnitude as the DC offset in the airgap profile 301. As a result of folding sinusoidal signals with a, particularly non-sinusoidal, airgap profile dependent term 309, higher frequency components, are also present in the compensation term 325, but according to the characteristics of the low pass filter and / or integrators these frequency components are attenuated.At one second the load angle gets linearly increased from 0° to a value of 90° at a rate of 40° / 0.2 sec. The compensation term 325 converges to approximate the airgap profile 301 during and / or after the change in the load angle.

[0058] Figure 3b shows a graph with time on the x-axis. The first and the second component of the magnetic stator flux vector 303.1, 303.2 are plotted over time with per-unit scaling on the y-axis (1=100% flux). At one second the load angle is constantly changed from 0° to 90° at a rate of 40° / 0.2 sec, not affecting the components of the magnetic stator flux vector 303.1, 303.2 since the load angle is simulated as an increasing lag in the mechanical rotor position while the components of the magnetic stator flux vector 303.1, 303.2 remains unaffected.

[0059] Figure 3c shows a graph with time on the x-axis. A simulated load angle 327 and an obtained load angle 323 (in degree) is plotted as a function of time. During the starting process of the method (starting at 0 sec) the obtained load angle 323 converges to approximate the simulated load angle 327. At one second the simulated load angle 327 is constantly changed from 0° to 90° at a rate of 40° / 0.2 sec. The obtained load angle 323 converges to approximate the simulated load angle 327 during and / or after the change in the simulated load angle 327.

[0060] Figure 4 shows an illustration of an implementation of the method of obtaining a load angle. The method comprises determining 201 an airgap profile 301.

[0061] The method comprises determining 203 a magnetic flux angle evolution, which can comprise the following steps:

[0062] The inner stator voltage can be estimated based on the measured terminal voltages and the expected voltage drop across the stator winding resistance. The estimated inner stator voltage can be transferred into an orthogonal, stationary reference frame.

[0063]

[0064] <

[0065] Here, uR, us, uTcan correspond to the terminal voltages at the three phases of the stator. iR, is, iTcan correspond to the current flowing through the three phases of the stator windings. Rscan correspond to the ohmic resistance of the stator winding.

[0066] The magnetic flux vector 303 can be determined by

[0067] (fc) = / (“«)«

[0068] The angular orientation of the magnetic flux vector 303 0 can be determined by

[0069]

[0070] Determining the frequency for a magnetic flux angle evolution can be based on the derivation of the magnetic flux vector angle over time:> Vsi~ 2ndt

[0071] The method comprises determining 205 a magnetic-flux-based reference oscillation signal 305, which can comprise the following steps:

[0072] From the magnetic flux angel evolution, the frequency for the calculation of the magnetic-fluxbased reference oscillation signal 305 can be calculated by

[0073] f pp 2 • f pSt

[0074] The two magnetic-flux-based reference oscillation terms can be determined by

[0075]

[0076] The method comprises determining 219 a phase shift 319 between the periodic oscillation of the airgap profile 301 and the magnetic-flux-based reference oscillation signal 305, which can comprise the following steps:

[0077] A closed loop compensation routine can comprise:

[0078] 1. Calculating an airgap profile dependent term 309 by subtracting a compensation term 325 from the airgap profile 301;

[0079] 2. Folding 231, 233 the airgap profile dependent term 309 with the magnetic-flux-based reference oscillation terms by multiplication;

[0080] 3. Integrating 211, 213 and / or low pass filtering of the result of the multiplication (step 2); 4. Folding 215, 217 the signal at the output of the integrators, the magnitude of the first and the second fundamental frequency component 311, 313, (step 3) with the magnetic- flux-based reference oscillation terms by multiplication;

[0081] 5. Integrating 221 and / or low pass filtering the airgap profile dependent term 309;

[0082] 6. Determination of a compensation signal 325 by addition of the first and second component of the compensation term 315, 317 (result of the two multiplications in step 4) and the DC component of the compensation term 321 (result of the integration from step 5).

[0083] For example, the closed loop compensation can be performed by:

[0084]

[0085] wherein the sum can be converted to an integral over the time t for sufficiently small time intervals k (Comp [

[0086]

[0087] The first component of the compensation term 315, the second component of the compensation term 317 and the DC component of the compensation term 321 for the closed loop compensation routine can be calculated by:

[0088]

[0089] The airgap pr ofile 301 x can be described by:

[0090]

[0091] wherein and / ^can be, particularly approximated by, the magnitude of the fundamental frequency 311, 313. The phase shift 319 can be determined from the, particularly approximated, magnitudes of the fundamental frequency 311, 313:

[0092]

[0093] If a plurality of N>1 airgap profiles are used to determine 209 the load angle, the phase shift 319 can be calculated individually, or a combined phase shift 319 can be obtained by combining the individual contributions using, for example, a circular mean according to:

[0094]

[0095] The load angle 323 may be determined 209 from the phase shift 319. In embodiments, the load angle 323 may be set as being equal to or approximated by the determined phase shift 319. In other embodiments, a calibration function may be used for determining 209 the load angle 323 as a function of the phase shift(s) 319. This calibration function receives the phase shift(s) 319 and particularly the magnetic-flux-based reference oscillation signal 305 as an input and obtains the load angle 323. The calibration function may be initially calibrated, e.g., upon energizing of the motor, by running the motor in one or more calibration runs with known load angle(s) and fitting the calibration function accordingly in a known manner.

[0096] This method is particularly useful for ring-type motors, for which typically no straightforward measurement of the load angle is available but for which typically at least one airgap sensor is already installed for monitoring the airgap. In this manner, the load angle becomes available without the need to add significant sensor hardware.List of reference signs

[0097] 110 stator

[0098] 111 stator windings

[0099] 113 airgap monitoring unit

[0100] 115 airgap width

[0101] 120 salient pole rotor

[0102] 121 pole

[0103] 201 determining an airgap profile

[0104] 203 determining a magnetic flux angle evolution

[0105] 205 determining a magnetic-flux-based reference oscillation signal 207 determining a phase shift

[0106] 209 determining a load angle

[0107] 211 integration

[0108] 213 integration

[0109] 215 folding

[0110] 217 folding

[0111] 219 determining a phase shift

[0112] 221 low pass filtering and / or integration of 309

[0113] 231 folding

[0114] 233 folding

[0115] 301 airgap profile

[0116] 303 magnetic flux vector

[0117] 305 magnetic-flux-based reference oscillation signal

[0118] 309 airgap profile dependent term

[0119] 311 magnitude of first fundamental frequency component 313 magnitude of second fundamental frequency component 315 first component of the compensation term

[0120] 317 second component of the compensation term

[0121] 319 phase shift

[0122] 321 DC component of the compensation term

[0123] 323 obtained load angle

[0124] 325 compensation term

[0125] 327 simulated load angle

Claims

Claims1. A method of obtaining a load angle (323) in an electric motor having a stator (110), a salient pole rotor (120), and an airgap monitoring system, the method comprising:determining (201) an airgap profile (301) by the airgap monitoring system, wherein the airgap profile (301) is based on an airgap width (115), at at least one stationary angular position of the stator (110), between the stator (110) and the salient pole rotor (120) over time, andwherein the airgap profile (301) comprises a periodic oscillation having a fundamental frequency due to a periodicity of a pole structure of the electric motor; determining (203) a magnetic flux angle evolution,wherein the magnetic flux angle evolution is indicative of an angular orientation of a magnetic flux vector (303) over time, whereby the magnetic flux vector (303) represents the magnetic flux induced by the stator (110) in a stationary electromagnetic reference frame of the stator (110);determining (205), from the magnetic flux angle evolution, a magnetic-flux-based reference oscillation signal (305) having the fundamental frequency;determining (219) a phase shift (319) between the periodic oscillation of the airgap profile (301) and the magnetic-flux-based reference oscillation signal (305); and determining (209) the load angle (323) from the phase shift (319).

2. The method according to any previous claim, wherein the load angle (323) is determined (209) dynamically.

3. The method according to any previous claim, wherein the airgap monitoring system comprises at least one airgap monitoring unit (113) having a distance sensor for measuring the airgap width (115) between the stator (110) and the salient pole rotor (120) at a respective stationary angular position of the stator (110).

4. The method according to the previous claim, having a plurality of airgap monitoring units (113), and wherein a phase shift (319) between the periodic oscillation of the airgap profile (301) and the magnetic-flux-based reference oscillation signal (305) is determined (207) for the airgap profiles (301) determined (201) by at least two of the airgap monitoring units (113).

5. The method according to any previous claim, wherein determining (203) the magnetic flux angle evolution comprises calculating, from a terminal voltage of the electric motor, a magnetic flux vector in the stationary electromagnetic reference frame.

6. The method according to any previous claim, wherein the magnetic-flux-based reference oscillation signal (305) comprises two magnetic-flux-based reference oscillation terms.

7. The method according to the previous claim, wherein a periodicity of the magnetic flux angle evolution corresponds to half the periodicity of the pole structure, and wherein determining (205) the magnetic-flux-based reference oscillation signal (305) includesdetermining, from the magnetic flux angle evolution, a magnetic flux vector oscillation frequency of the angular orientation of the magnetic flux vector (303); and generating the two magnetic-flux-based reference oscillation terms as a periodic signal having twice the magnetic flux vector oscillation frequency.

8. The method according to any previous claim, wherein determining (219) the phase shift comprises determining two products by multiplying an airgap profile dependent term (309) with each of the two magnetic-flux-based reference oscillation terms.

9. The method according to any of the two previous claims, wherein determining (219) the phase shift (319) further comprises low-pass filtering (211, 213) the two products.

10. The method according to any of the three previous claims, wherein determining (219) the phase shift (319) between the periodic oscillation of the airgap profile (301) and the two magnetic-flux-based reference oscillation terms further comprises performing a closed loop compensation routine, wherein the closed loop compensation routine compensates a constant DC-term in each of the two products.

11. The method according to the previous claim, wherein the phase shift (319) is determined (219) from the compensation terms (311, 313) derived from the closed loop compensation routine.

12. The method according to any previous claim, wherein the electric motor is a ring type electric motor.

13. The method according to any previous claim, wherein the electric motor belongs to a gearless mill drive.

14. A method of operating an electric motor, the method comprising obtaining a load angle by the method according to any previous claim, the method of operating the motor further comprising:controlling the electric motor by a magnetic flux control scheme using the load angle (323), in particular such that the load angle (323) remains within a predetermined load angle tolerance range.

15. An electric motor, the electric motor comprising:a stator (110);a salient pole rotor (120);an airgap monitoring system; anda monitoring- and control system with a memory unit;wherein the memory unit stores a program code, which when executed by the monitoring- and control system causes the monitoring- and control system to perform a method according to any of claims 1 to 14.

Images