Method and device for the torque measurement in the drive train of a wind energy facility

Abstract

A method and a device for measuring the torque in the drivetrain (1) of a wind power plant is described, having at least two incremental encoders (7, 8) which are positioned at two different positions on at least one shaft (3) of the drivetrain (1) and which each supply periodic rotational signals, wherein the phases of the rotational signals are evaluated in order to detect a phase shift, and a torque of the shaft (1) is determined from the phase shift. The detected phase shift is corrected as a function of a zero load phase shift (A.sub.Zero), using a rigidity factor K, wherein, in order to determine the zero load phase shift (A.sub.Zero) and the rigidity factor K, in-situ calibration is carried out before and/or between the torque-determining processes. The in-situ calibration is performed at zero load of the wind power plant, i.e. below a rated rotational speed and with a generator torque equal to zero, and at the rated load of the wind power plant, i.e. at the rated rotational speed and with a generator torque greater than zero.

Claims

1. A method for measuring torque in a drive train of a wind energy facility, the method comprising: determining a phase shift and a torque of a first shaft of the drive train, the first shaft including a first incremental encoder and a second incremental encoder positioned at different locations on the first shaft, the first incremental encoder and the second incremental encoder configured to provide a first periodic rotation signal and a second periodic rotation signal, wherein a phase of the first periodic rotation signal and the second periodic rotation signal is evaluated for determining the phase shift, and the torque of the shaft is determined from the phase shift, and wherein the determined phase shift is corrected based on a zero-load phase shift and using a stiffness factor K, wherein an in-situ calibration is carried out before or between the torque determination, for determining the zero-load phase shift and the stiffness factor K, and wherein the in-situ calibration comprises the following steps: measuring a first zero-load signal of the first incremental encoder and a second zero-load signal of the second incremental encoder over a first measuring time period and determining a temporally averaged zero-load phase shift between the first zero-load signal and the second zero-load signal, wherein the wind energy facility during the first measuring time period is operated below a specified speed and a generator torque is equal to zero; and measuring a first nominal load signal of the first incremental encoder and a second nominal load signal of the second incremental encoder over a second measuring time period, determining a temporally averaged nominal load phase shift between the first nominal load signal and the second nominal load signal, and determining the stiffness factor K based on the nominal load phase shift, wherein the wind energy facility is operated at the specified speed during the second measuring time period and the generator torque of the wind energy facility is kept larger than zero.

2. The method according to claim 1, wherein the phase shift is corrected adding or subtracting the zero-load phase shift, or multiplying by the stiffness factor K.

3. The method according to claim 1, wherein the generator torque is kept constant during the second measuring time period.

4. The method according to claim 1, wherein the first measuring time period or the second measuring time period is between 2 minutes and 20 minutes.

5. The method according to claim 1, wherein the first incremental encoder is located on one or another of the first shaft or a second shaft, wherein the second incremental encoder is located on the other of the first shaft or the second shaft, and wherein the first shaft and the second shaft are coupled to one another using a gear.

6. The method according to claim 1, wherein, the wind energy facility is braked so as to reach a speed below the specified speed, and wherein a measurement value of the first incremental encoder and the second incremental encoder is evaluated in a region of a predefined speed, wherein the predefined speed is between 5 and 1 rpm for determining the zero-load phase shift.

7. The method according to claim 6, wherein the averaged zero-load phase shift is determined, by averaging a first zero-load phase shift that is determined after the braking of the wind energy facility and a second zero-load phase shift that is determined after starting up the wind energy facility.

8. The method according to claim 6, wherein the predefined speed is between 1 rpm and 3 rpm.

9. The method according to claim 1, wherein the wind energy facility is started up from a shut-down state for reaching a speed below the specified speed, wherein the measured values of the first incremental encoder and the second incremental encoder evaluated in a range of a predefined speed between 1 and 5 rpm for determining the zero-load phase shift.

10. The method according to claim 1, wherein the first incremental encoder and the second incremental encoder are arranged in a manner such that the measured values of the first incremental encoder and the second incremental encoder given a rotation of the wind energy facility without load have a phase shift with a predefined value, and wherein the first incremental encoder or the second incremental encoder are adjusted to the pre-defined value after their arrangement.

11. The method according to claim 1, wherein the generator torque is averaged over the second measuring time period.

12. The method according to claim 1, wherein the stiffness factor K is determined by dividing by the nominal load phase shift.

13. The method according to claim 1, wherein the first measuring time period or the second measuring time period is between 4 minutes and 15 minutes.

14. The method according to claim 1, wherein the first measuring time period or the second measuring time period is between 5 minutes and 10 minutes.

15. A device for measuring torque in a drive train of a wind energy facility, comprising: a first incremental encoder and a second incremental encoder located at different positions on at least one shaft of the drive train; and an evaluation device connected to the first incremental encoder and the second incremental encoder configured to: determine a phase shift from a measurement signal of the first incremental encoder and the second incremental encoder; determine a torque of the shaft to correct the phase shift based on a zero-load phase shift and using a stiffness factor K; and carry out an in-situ calibration before or between the torque measurement, for determining the zero load phase shift and the stiffness factor K, wherein the evaluation device for carrying out the in-situ calibration is configured to: measure a first zero-load signal of the first incremental encoder and a second zero-load signal of the second incremental encoder over a first measuring time period and to determine a temporally averaged zero-load phase shift between the first zero load signal and the second zero-load signal, wherein the wind energy facility during the first measuring time period is operated below a specified speed and a generator torque is equal to zero measure a first nominal load signal of the first incremental encoder and a second nominal load signal of the second incremental encoder over a second measuring time period; and determine a temporally averaged nominal load phase shift between the first nominal load signal and the second nominal load signal and to determine the stiffness factor K based on the nominal load phase shift, wherein the wind energy facility during the second measuring time period is operated at the specified speed and the generator torque at nominal load is kept greater than zero.

16. The device according to claim 15, wherein the evaluation device comprises: a logic circuit, wherein the logic circuit includes an AND logic unit or an OR logic unit, for determining the phase shift between the measured signals of the first incremental encoder and the second incremental encoder; and an analysing unit configured to determine a mechanical angle twist between the first incremental encoder and the second incremental encoder.

17. The device according to claim 15 further comprising: an electromechanical adjusting device configure to a position of the first incremental encoder or the second incremental encoder in a direction tangential to the shaft or along a circumference of the shaft.

18. The device according to claim 15, wherein the first incremental encoder or the second incremental encoder is based on at least one of a magnetic measuring principle, an optical measuring principle, or an inductive measuring principle.

19. The device according to claim 15, wherein the first incremental encoder or the second incremental encoder have a resolution of at least four impulses per shaft revolution.

20. The device according to claim 15, wherein, a resolution of the first incremental encoder and a resolution of the second incremental encoder are the same or differ by an integer multiple.

Description

(1) Embodiment examples of the invention are hereinafter described by way of figures.

(2) There are shown in:

(3) FIG. 1a a drive train of a wind energy facility in a schematic representation,

(4) FIG. 1b an evaluation device of the wind energy facility of FIG. 1a,

(5) FIG. 2 rotation signals of a first incremental encoder and of a second incremental encoder and a signal of a logic evaluation given rotation without load,

(6) FIG. 3 rotation signals of the first incremental encoder and of the second incremental encoder and the signal of a logic evaluation given a rotation under load of a positively defined torque,

(7) FIG. 4 rotation signals of the first incremental encoder and of the second incremental encoder and the signal of a logic evaluation given a rotation under load of a negatively defined torque,

(8) FIG. 5 an adjusting device for adjusting at least one of the two incremental encoders.

(9) A drive train 1 of a wind energy facility is schematically represented in FIG. 1a. A rotor 2 is arranged on a rotor main shaft 3. The rotor main shaft 3 is an input shaft of a gear 4. The gear 4 comprises a gear output shaft 5 which as a generator input shaft is simultaneously connected to a generator 6. A wind force which acts upon the rotor 2 rotates the rotor and thus drives the rotor main shaft 3. The gear 4 steps up a slow rotation movement of the rotor main shaft 3 into a quicker rotating movement of the gear output shaft 5 for the generation of electricity in the generator 6. A first incremental encoder 7 is arranged on the rotor main shaft 3. A distance between the rotor 2 and the first incremental encoder 7 is smaller than a distance between the first incremental encoder and the gear 4. A second incremental encoder 8 is arranged on the gear output shaft 5. The second incremental encoder 8 can also be arranged on the rotor main shaft 3, for example at a position 8′. The position 8′ has a smaller distance to the gear 4 than to the rotor 2 and both incremental encoders in this case should be arranged at the greatest possible distance. The incremental encoders 7 and 8 are connected to an evaluation device 9 (the connection is not represented in FIG. 1 for a better overview). The incremental encoders 7 and 8 provide output signals in the form of periodic rotation signals. These rotation signals are transferred to the evaluation device 9 via an electric lead connection.

(10) Given normal operation of the wind energy facility, the rotor 2 is driven in accordance with the wind strength and wind direction, so that different torques which change with time act upon the rotor main shaft 3. In order to be able to regulate the wind energy facility in a manner adapted to this, the respective torque M (t) is required. The necessary torque measurement is based on a direct evaluation of the relative twist angle between the two incremental encoders 7, 8. The mechanical angle twist between both encoders leads to a phase shift of the encoder output signals and on superimposing the output signals leads to a measurement signal A(t).sub.meas, with a changing pulse-pause ratio (in the case of rectangular signals of the incremental encoder) as a direct measure for this twisting and therefore for the effective torque (is explained further below in more detail).

(11) FIG. 1b shows the evaluation device 9 of FIG. 1a. The evaluation device comprises a superposition or logic circuit 9′ which links the output signals of the incremental encoders 7, 8 to one another as measurement signals A.sub.1 and A.sub.2. This logic circuit is designed for example as an AND logic unit. An output signal of the logic circuit 9′ is led further to an analysing unit 9″ which determines a mechanical angle twist between the two incremental encoders and determines the torque therefrom. An output signal A.sub.9 of the evaluation device 9 is forwarded to the wind energy facility regulation.

(12) FIG. 2 shows measurement signals or rotation signals A.sub.1 and A.sub.2 of the two incremental encoders 7, 8 on rotating without a load. The rotation signal A.sub.1 of the first incremental encoder 7 is represented in a diagram 10 and the rotation signal A.sub.2 of the second incremental encoder 8 is represented in a diagram 20. The rotation signal A.sub.1 has an impulse amplitude 11 or T.sub.imp with a pulse-pause ratio or duty cycle of 50/50. Herein, the impulse duration T.sub.high 12 is just as long as the pause time T.sub.low 13.

(13) The impulse period is T.sub.imp=60/nx (s), wherein x is the number of the impulses per revolution and n the shaft speed in r.p.m. The duty cycle is TV((n−1)T, nT)=T.sub.high/(T.sub.high+T.sub.low)=0.5.

(14) The rotation signal A.sub.2 has an impulse period 21 which corresponds to the impulse period 11 and likewise has a pulse-pause ratio of 50%. Herein, the pulse duration 22 is just as long as the pause time 23. Furthermore, the pulse duration 22 corresponds to the pulse duration 12 and therefore also the pause time 23 to the pause time 13. The rotation signal A.sub.2 is shifted with respect to the rotation signal A.sub.1 by a quarter of the period 21 or 11. This corresponds to a rotation angle of Δφ.sub.signal=90°.

(15) The measurement signals A.sub.1 and A.sub.2 are transferred to the evaluation device 9. The measurement signals A.sub.1 and A.sub.2 are superimposed in the evaluation device 9, here comprising an AND logic unit 9′. The AND logic unit 9′ thereupon provides an output signal A.sub.logic which is represented in a diagram 30 and which has an impulse period 31 which corresponds to the impulse period 11 and 21. A pulse duration 32 herein corresponds to a quarter of the impulse period 31, whereas a pause time 33 consequently corresponds to three-quarters of the impulse period 31 (pulse-pause ratio 25%).

(16) In order to achieve the aforementioned phase difference or phase shift of preferably half an impulse period (with the pulse-pause ratio of 50% this corresponds to a phase shift of 90%) for the impulse signals of the two incremental encoders 7, 8 in the non-loaded state, a position adjustment between the encoders is carried out since such a relative adjustment of the two incremental encoders 7, 8 cannot be reliably ensured during the assembly.

(17) In order to adjust the position, the incremental encoders 7, 8 are adjusted by way of an adjusting device in a manner such that the desired pulse-pause ratio is achieved. To this end, one of the two encoders, i.e. the incremental encoder 7 is automatically or manually adjusted along a tangential path or one which is adapted to the shaft contour, for example along a threaded rod or spindle. Herein, it can move over the complete theoretical measuring region during the adjusting. The pulse-pause ratio of 50% is specified by way of example in the description, and any arbitrary pulse-pause ratio can be set, as long as a conclusive evaluation of the signals is achieved.

(18) In a further diagram 40, the output signal A.sub.logic is represented in the extreme positions A.sub.min and A.sub.max during the adjustment of the incremental encoders 7, 8 on moving over the complete theoretical measuring region. The dashed line represents the output signal A.sub.min and the unbroken line represents the output signal A.sub.max and herewith the limit of the adjustment.

(19) After the previously described position adjustment, an in-situ calibration method is carried out, in order to determine a zero-load phase shift or a zero-point adjustment of the torque measuring device and a stiffness factor K which takes into account an amplification factor or a gradient of the torque measurement characteristic line of the drive train or a stiffness of the drive train.

(20) Before a calibration of the torque measuring device, the drive train of the wind energy facility should have reached its nominal operating temperature. This can be achieved e.g. by an adequately long operation of the wind energy facility without the use of the torque measuring device. This can be carried out without any restrictions since the additional torque sensorics, although improving the facility behaviour, however are not absolutely necessary for the operation.

(21) In order to determine the zero-load phase shift, the wind energy facility is brought into the idling region, i.e. into non-productive operation, via the facility's control system, given operating conditions below the nominal speed, preferably with weak wind condition. The braking of the facility into the idling speed region can be carried out actively via the pitch adjustment of the rotor blades. In this operational region, the facility is taken off the mains, i.e. the generator is not “subjected to current” and does not therefore develop a torque M.sub.gen(t)=0, wherein this state is a necessary constraint. The blades of the facility are then situated completely in the feathering position, at which the pitch angle is roughly 90°. The rotor therefore no longer develops a positive drive torque, in contrast an aerodynamic braking moment M.sub.rot(t).sup.˜0 is active, but this becomes exponentially smaller with a reducing speed.

(22) After switching off the generator, the measurement values which are continuously provided by the torque measuring device are recorded over the speed in the range of very small speeds (between 3 and 1 rpm). In the evaluation device 9, the recorded measurement values are superimposed and a temporally averaged zero-load phase shift A(t)brake between the output signal of the first incremental encoder 7 and of the second incremental encoder 8 is determined for determining the zero point of the torque measuring device. The determined zero-load phase shift is stored in the evaluation unit as a zero-point of the torque measuring device.

(23) This measurement can also be repeated on starting up the facility given weak wind and in idling operation for determining a zero-load phase shift A(t).sub.high, likewise given a switched-off generator, wherein it is preferably carried out in the same rotor speed range (1-3 rpm). The blades are moved very slowly 1°/min out of the feathering position in the direction of the nominal position by way of the control system, in order to minimise the influence of acceleration moments M.sub.rot(t).sup.˜0. The average of the averaged measurement values of the two trials (course of braking, starting up) is then formed
A.sub.zero=(A(t).sub.brake+A(t).sub.high)/2
and the zero-load phase shift or the zero point of the torque measurement characteristic line which is represented by this is accordingly adapted and is stored in the evaluation device.

(24) For an exact measurement of the zero-load phase shift, it is advantageous for both trials to be carried out in a directly subsequent manner. These trials may be repeated several times, as appropriate.

(25) The determining of the stiffness factor K or of the reinforcement factor or the gradient of the torque measurement characteristic line of the torque measuring device is described hereinafter. This determining is preferably carried out in full-load operation, i.e. in the nominal operation of the facility, in order to obtain an accurate a measurement as possible over the complete torque measuring range. In this operating region, the generator torque is kept constant by the facility control system/facility regulation. The rotor speed is actively regulated to the facility nominal speed via the rotor blade adjustment (pitch). However, due to the relatively sluggish characteristics of the pitch regulation, the rotor speed fluctuates in a range of as a rule ±10% about the nominal speed, so that dynamic braking and acceleration torque components could upset the calibration of the torque measuring device. For this reason, a longer measuring time period of several minutes, preferably up to maximally 10 minutes, is selected. During this measuring time period, the generator torque should not change, i.e. the facility must operate in its nominal power region during the complete measuring time period.

(26) The generator torque corresponds to the air gap moment and this cannot be detected by measuring technology in a direct manner, but is determined numerically from the measured electrical characteristic variables and the generator parameters by the converter which is contained in the wind energy facilities, and is “communicated” to the facility control as signal/information. As is standard, for this the converter requires a generator rotor position signal or at least a generator rotor speed signal. This signal can be provided by one of the incremental encoders 7, 8 or be provided by a standard encoder on the generator.

(27) The torque measuring device with the evaluation device 9 according to FIG. 1b and with an evaluation according to FIG. 2 provides measurement signals with respect to a phase shift Δφ between the two encoder signals given nominal operation. The temporal average of the measurement signals of the torque measuring device which specify the phase shifts, and the values of the certain generator torque result in the stiffness factor
K=M.sub.Cre/Δφ
by way of which hence the second necessary characteristic line point is given for determining the torque measurement characteristic line, this point corresponding to the wind energy facility nominal moment. This measurement may, where necessary, be repeated arbitrarily often. In this case too, the facility should have reached its nominal operating temperature. Optionally, the determining of this value of the characteristic line can also be carried out before determining the zero point.

(28) The output signals or rotation signals A.sub.1′ and A.sub.2′ of the two incremental encoders 7, 8 in normal measuring operation under the load of the wind energy facility given a positive torque direction are represented in FIG. 3. The rotation signals A.sub.1′ of the first incremental encoder 7 are specified in the upper diagram 10′. The rotation signals A.sub.2′ of the second incremental encoder 8 are represented in the middle diagram 20′. In the second measuring time period, the first incremental encoder 7 provides a rotation signal A.sub.1′ with an impulse period 11′ and with a pulse pause ratio of 50/50. Herein, the pulse duration 12′ is just as long as the pause time 13′. The rotation signal A.sub.2′ has an impulse period 21′ which corresponds to the impulse period 11′ and likewise has a pulse-pause ratio of 50%. Furthermore, the pulse duration 22′ corresponds to the pulse duration 12′ and therefore the pause time 23′ to the pause time 13′. The signal A.sub.2 according to FIG. 2 is represented in a dotted manner. The rotation signal A.sub.2′ leads the rotation signal A.sub.1′ by 15% of the impulse period 21 or 11. This corresponds to a rotation angle Δφ.sub.signal of 54°.

(29) The output signals A.sub.1′ and A.sub.2′ are superimposed in the AND logic which provides an output signal A.sub.logic′ with an impulse period 31′, wherein here too the signal A.sub.logic according to FIG. 2 is represented in a dotted manner. The pulse duration 32′ results with
A((n−1)T,nT)=T.sub.high/(T.sub.high+T.sub.low)=0.35
i.e. 35% of the impulse period 31′. The output signal A.sub.logic′ thus contains a phase shift which results under load from the torque which is exerted upon the rotor, as well as a phase shift which already prevails during a zero-load operation of the wind energy facility. The twist angle of the shaft Δφ.sub.shaft which is caused by the load torque can be determined from the output signal A.sub.logic′ with the variables according to FIG. 2 by
Δφ.sub.shaft=(A((n−1)T,nT)−0.25)/0.25.Math.90°/x=36°/x
wherein x is the number of impulses per revolution. The torque can be determined from the torque characteristic line in dependence on the phase shift. Rotation signals A.sub.1″ and A.sub.2″ of the two incremental encoders 7, 8 in measuring operation under load with a negative torque direction are represented in FIG. 4. The torque which acts upon the rotor main shaft corresponds to the torque of FIG. 3, but in FIG. 4, as mentioned, is defined as negatively rotating. The rotation signal A1″ of the first incremental encoder 7 which corresponds to the output signal or rotation signal A.sub.1′ is represented in the diagram 10″ and the rotation signal A.sub.2″ of the second incremental encoder 8 is specified in diagram 20″. The rotation signal A.sub.2″ corresponds to the rotation signal A.sub.2′ but is displaced oppositely to A.sub.2′. The time difference between the signals A.sub.1″ and A.sub.2″ is ΔT=0.35 T.sub.imp which corresponds to an angle Δφ.sub.logic of 126°. The output signals A.sub.1″ and A.sub.2″ are superimposed in the AND logic which provides an output signal A.sub.logic′ with an impulse period 31″, wherein here too the signal A.sub.logic according to FIG. 2 is represented in a dotted manner. The pulse duration 32″ results by
A((n−1)T,nT)=T.sub.high/(T.sub.high+T.sub.low)=0.15
i.e. 15% of the impulse period 31″. The pulse duration 32″ is herein 15% of the impulse period 31″. The output signal A.sub.logic″ represents a phase shift which results from the torque under load said torque being applied onto the rotor, as well as a phase shift which also prevails during a zero-load operation of the wind energy facility. The twist angle of the shaft Δφ.sub.shaft which is caused by the load torque can be determined from the output signal A.sub.logic′ with the variables according to FIG. 2 by
Δφ.sub.shaft=(A((n−1)T,nT)−0.25)/0.25.Math.90°/x=−36°/x
wherein x is the number of impulses per revolution.

(30) In order to measure the torque of the rotor main shaft of the wind energy facility 1, arbitrary output signals A.sub.meas1 and A.sub.meas2 of the two incremental encoders are evaluated as is explained with FIG. 3 and FIG. 4. Herein, the rotation signals of the incremental encoders 7, 8 are superimposed by the AND logic circuit which provides a signal A.sub.measlog which represents the phase shift between the two rotation signals. In order to determine a momentary torque, the signal of the zero-point evaluation A.sub.zero is added to this current measurement signal A.sub.measlog and the result is multiplied by the stiffness factor K: M(t)=A.sub.measlog(t)*K+A.sub.zero*K.

(31) The torque signal is used for example for the regulation of the wind energy facility.

(32) FIG. 5 shows an adjusting device 50 for adjusting an incremental encoder 51 relative to a shaft 52. An encoder disc or an encoder tape is fixedly connected to the shaft 52. The incremental encoder 51 comprises a sensor 51′ for scanning the encoder disc or the encoder tape. The incremental encoder 51 can scan the encoder disc for example in a photoelectrical or magnetic manner. Toothed wheels can also be used as encoder discs.

(33) The adjusting device 50 is represented in FIG. 5(1), whereas a cross section through the shaft 52 and the arrangement of the adjusting device 50 relative to the shaft 52 is shown in FIG. 5(1). The adjusting device 50 comprises a linear drive 53 with a spindle 53′, on which a slide 54 is arranged. The incremental encoder sensor 51′ is fastened to the slide. Furthermore, the adjusting device comprises two guide rails 55, along which the slide is movable. The spindle 53′ of the linear drive 53 can be driven by an electric drive 56 in a manner such that the slide is displaceable along the spindle in a direction R.sub.1. In FIG. 5(2) it is represented that the adjusting device 50 is preferably arranged in a manner such that the direction R.sub.1 runs tangentially to the shaft 52, so that the incremental encoder sensor 51 is adjustable tangentially to the shaft 52. Furthermore, the slide 54 can be rotated about the spindle 53′ so that an adjustment of the sensor 51′ along the direction R.sub.2, adapted to the shaft contour, is possible. The arrow R.sub.3 describes the detection direction of the incremental encoder sensor 51′.

(34) Optionally, an observer supported by a torque measurement value can be used for the facility regulation, i.e. the measured torque measurement value can use a mathematical model of the wind energy facility drive train in the form of a so-called observer for an improved facility model. The observer can mathematically estimate model-based states in the drive train and feed them back to the regulation. The speed signal from the generator speed encoder can be used for “supporting” the observer model, so that the observer does not drift away due to inaccuracies in the modelling and an estimation result which is true to expectation is achieved. The measuring device described above can improve the estimation accuracy of the observer with regard to the accuracy, i.e. with regard to the stationary deviation and dynamics (speed estimation value convergence) by way of the signals of the two incremental encoders and in particular by way of the torque signal. This observer application is optional and not absolutely necessary since the aforementioned control-technological advantages are already achieved solely by the measuring device. The regulation accuracy can be improved once again by the introduction of observer models.