CN110630448A - Soft measurement method for wind shear effect - Google Patents
Soft measurement method for wind shear effect Download PDFInfo
- Publication number
- CN110630448A CN110630448A CN201910857475.5A CN201910857475A CN110630448A CN 110630448 A CN110630448 A CN 110630448A CN 201910857475 A CN201910857475 A CN 201910857475A CN 110630448 A CN110630448 A CN 110630448A
- Authority
- CN
- China
- Prior art keywords
- wind
- torque
- soft measurement
- shear effect
- wind shear
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
- F03D—WIND MOTORS
- F03D17/00—Monitoring or testing of wind motors, e.g. diagnostics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
- F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
- F05B2270/00—Control
- F05B2270/30—Control parameters, e.g. input parameters
- F05B2270/304—Spool rotational speed
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
- F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
- F05B2270/00—Control
- F05B2270/30—Control parameters, e.g. input parameters
- F05B2270/337—Electrical grid status parameters, e.g. voltage, frequency or power demand
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
- F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
- F05B2270/00—Control
- F05B2270/70—Type of control algorithm
- F05B2270/709—Type of control algorithm with neural networks
Abstract
The patent provides a soft measurement method for a wind shear effect, and belongs to the technical field of wind power generation. The invention consists of a phase-locked frequency multiplier (10), an electromagnetic torque soft measurement (20), a torque difference soft measurement (30) and a constraint calculation (40). Firstly (10) outputting frequency multiplication sine and cosine components to (20) and (30); then (20) and (30) adopt an online learning neural network based on the self-adaptive noise cancellation principle to respectively pass through the output current i of the wind driven generatorsAnd the actually measured rotating speed omega of the wind wheel obtains an electromagnetic torque and a torque difference estimated value, and finally, soft measurement of mechanical torque is realized through (40). Because no filtering link exists, the method has good real-time performance, and can be used for not only accurately measuring the wind shear effect torque ripple, but also performing online compensation control on the wind shear effect torque ripple.
Description
Technical Field
A wind shear effect soft measurement method relates to mechanical torque soft measurement and belongs to the technical field of wind power generation.
Background
Wind power generation utilizes wind power to drive blades to rotate, and mechanical energy of the generator is converted into electric energy. At high altitudes of the wind farm, the wind speed will vary with vertical height:
whereinRespectively is height Zn,Z0The average wind speed, α, is related to the surface structure and the atmosphere stability. Therefore, when the blades rotate, the wind speeds of all points of the blades are different, a wind shear effect is generated, and the blades are subjected to micro-metaanalysis by using a phyllotactic theory method to obtain the mechanical torque of the n blades of the wind turbine:
wherein A, B is the correlation constant. From the above formula, it can be seen that if the wind changes with the height, the mechanical torque generated by the wind turbine is composed of a stable component and a ripple component, which are expressed asWherein, the stable component is derived from the average wind speed, the ripple component is derived from the wind shear effect, the pulsation frequency is n omega, and omega is the actually measured rotating speed of the wind turbine. This phenomenon not only reduces the performance of the wind turbine and the tower life, but also produces power pulsations in the power generation.
Therefore, the wind shear effect torque ripple can be accurately measured, the damage of the wind shear effect torque ripple can be scientifically evaluated, and a basis can be provided for compensation control of the wind shear effect torque ripple. However, at present, no effective scheme exists for measuring the periodic mechanical torque ripple, and the periodic mechanical torque ripple cannot be directly measured by additionally arranging a sensor; mechanical torque pulsation can not be obtained from the above formula by measuring wind speed changes at different heights and combining a phyllotactic theory; and the ripple component frequency is low, and the general filtering method has large error and time delay.
Disclosure of Invention
In order to measure the mechanical torque pulsation and further ensure the safe and reliable operation of the wind power generation device, the invention provides a soft measurement method for the wind shear effect.
The principle of the wind shear effect soft measurement method provided by the invention is that an artificial neural network is learned on line based on the self-adaptive noise cancellation principle, and iterative learning is carried out by combining a reasonable artificial neural network topological structure and parameters, so that the online real-time measurement of mechanical torque pulsation is realized. As shown in fig. 1, the soft measurement method of the present invention comprises: the phase-locked frequency multiplication circuit outputs frequency multiplication sine and cosine components as the input of an artificial neural network, outputs the optimal weight value by combining a reference item and performing linear combination after iterative learning, and the whole system has strong self-adaptive capacity, is easy to realize and has high efficiency.
The invention relates to a wind shear soft measurement process, which comprises the following steps:
the fan rotating speed sensor equipment inputs the real-time rotating speed omega of the wind turbine to the phase-locked frequency multiplier (10);
as shown in the attached figure 2, the input signal passes through a phase-locked circuit (11) to obtain a sine signal sin ω t, and the sine signal sin ω t and a cosine signal cosn ω t are obtained after frequency multiplication (12) and phase shift (13), wherein n is the number of wind wheel blades.
An artificial neural network (21) in the electromagnetic torque soft measurement (20) inputs { sin n ω t, cosn ω t, 1} with a corresponding weight of { w }, as shown in fig. 3(a)1s,w1c,w1Reference item is the output current i of the wind driven generatorsOn-line learning error amountAnd reference item isSatisfies the formula:whereinRespectively output current i for wind power generatorsA stable component estimation term and a pulsating component estimation term; the learning of the neural network adopts a least mean square algorithm, the weight is adjusted by using error quantity, and the formula is as follows:
wherein eta is the learning rate (eta is more than 0 and less than or equal to 1). Over a number of iterations, the weights approach the optimum value and are employed by a linear combination (22). As shown in FIG. 3(b), (22) the optimal weights in the input set { sinn ω t, cosn ω t, 1}, and (21) and the scaling factor K in FIG. 5p4Linearly combined and outputThe formula is as follows:
wherein the content of the first and second substances,an electromagnetic torque estimation term,Is an electromagnetic torque ripple estimation term based onIn the above formulaWhereinIs a permanent magnet flux linkage, npThe number of pole pairs of the generator.
The torque soft measurement (30) inputs { sinn ω t, cosn ω t, 1} with a weight of { w }2s,w2c,w2}, outputAs shown in fig. 4; the reference term is torque difference delta T, and error amount e is learned onlineΔTSatisfies the formula with the reference term Δ T: is a torque difference estimation term; according toAnd FIG. 5, determining the parameter T3WhereinAnd J is moment of inertia. The learning of the neural network adopts a least mean square algorithm, the weight is adjusted by using error quantity, and the formula is as follows:
wherein eta is the learning rate (eta is more than 0 and less than or equal to 1), and each weight approaches the optimal value after a plurality of iterations. By the formula:
As shown in fig. 4, linear combination (40) input to (20) outputAnd (30) outputPerforming constraint calculation and outputAnd obtaining the optimal soft measurement value of the mechanical torque and the pulsation thereof, and measuring the wind shear effect.
Drawings
The patent is described in further detail below with reference to the figures and the detailed description.
FIG. 1 is an overall flow chart of soft measurement of wind shear effect torque ripple.
Fig. 2 is a flow chart of the internal phase-locked frequency multiplier of fig. 1.
Fig. 3(a) is a schematic diagram of an electromagnetic torque soft measurement artificial neural network.
Fig. 3(b) is a flow chart of the linear combination of the electromagnetic torque soft measurement.
FIG. 4 is a schematic diagram of a soft torque difference measurement
Fig. 5 is a block diagram of fan speed control.
Detailed Description
The invention discloses a soft measurement method for a wind shear effect, and introduces the basic composition and the measurement process of the method. The basic components comprise a standard phase-locked frequency multiplier, electromagnetic torque soft measurement, torque difference soft measurement and constraint calculation, all the parts are connected as shown in figure 1, and a direct-drive permanent magnet synchronous power generation system is taken as an example to introduce a specific flow of wind shear effect soft measurement of the system.
The output of the actually measured rotating speed sensor of the wind power generation system is connected to the phase-locked frequency multiplier and the torque difference soft measurement, and the output of the stator current sensor is connected to the electromagnetic torque soft measurement. And giving an initial weight value, and accessing sine and cosine components output by the phase-locked frequency multiplier into electromagnetic torque soft measurement and torque difference soft measurement. And the online learning neural network in the electromagnetic torque soft measurement and the torque difference soft measurement outputs the optimal estimated values of the electromagnetic torque and the torque difference after a plurality of times of iterative learning.
The output of the electromagnetic torque soft measurement and the torque difference soft measurement is accessed to the constraint calculation, the equation calculation is carried out, and the optimal estimated value of the mechanical torque is output, the pulsation of the optimal estimated value consists of sine and cosine components, and the weight values respectively correspond to the peak values of the sine component and the cosine component.
The practical significance of the invention is as follows: the wind shear effect soft measurement method can be applied to a wind power generation system to realize real-time mechanical torque pulsation measurement, an artificial neural network learning algorithm is applied to the technical method, the topological structure is simple, complex and redundant lead lines are not needed, and high-accuracy and high-efficiency online soft measurement can be realized without using excessive equipment. The method has important significance for ensuring the safe and reliable operation of the wind power generation system.
Claims (5)
1. A wind shear effect soft measurement method is characterized in that: comprises a phase-locked frequency multiplier (10), an electromagnetic torque soft measurement (20), a torque difference soft measurement (30) and a constraint calculation (40);(10) outputting frequency multiplication sine signal and cosine signal components to (20), (30); (20) and (30) respectively passing through the output current i of the wind driven generatorsAnd the electromagnetic torque and the torque difference are calculated with the actually measured rotating speed omega of the wind wheel, and then the soft measurement of the wind shear effect is realized through (40).
2. The device (10) of claim 1, wherein: comprises a phase lock (11), a frequency multiplication (12) and a phase shift (13); inputting actually measured wind wheel rotating speed omega, obtaining sin omega t through phase locking (11), obtaining sin n omega t through frequency doubling (12), and finally obtaining cosn omega t through phase shifting (13), wherein n is the number of wind wheel blades.
3. The method as set forth in claim 1 (20), characterized in that: the method comprises an online learning neural network (21) based on the self-adaptive noise cancellation principle and a linear combination (22); (21) the input set is { sin n ω t, cos n ω t, 1} and the corresponding weight is { w }1s,w1c,w1On-line learning of error amount }And reference item isSatisfy the formulaWhereinRespectively output current i for wind power generatorsStable component estimation item, pulsation component estimation item, optimal weight { w after iterative learning1s,w1c,w1Participating in linear calculation; (22) the input set is { sin n ω t, cos n ω t, 1}, and the output set isSatisfies the formula:
4. The (30) of claim 1, wherein: the online learning neural network based on the self-adaptive noise cancellation principle has an input set of { sin n ω t, cos n ω t, 1} and a corresponding weight of { w }2s,w2c,w2Put the set of outputs asThe on-line learning error amount is obtained by subtracting the output amount from the parameter item delta TWherein Δ T and ω satisfy the formula:j is moment of inertia.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910857475.5A CN110630448A (en) | 2019-09-11 | 2019-09-11 | Soft measurement method for wind shear effect |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910857475.5A CN110630448A (en) | 2019-09-11 | 2019-09-11 | Soft measurement method for wind shear effect |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110630448A true CN110630448A (en) | 2019-12-31 |
Family
ID=68972550
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910857475.5A Pending CN110630448A (en) | 2019-09-11 | 2019-09-11 | Soft measurement method for wind shear effect |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110630448A (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2009026930A2 (en) * | 2007-08-31 | 2009-03-05 | Vestas Wind Systems A/S | Method for controlling at least one adjustment mechanism of a wind turbine, a wind turbine and a wind park |
US20140097618A1 (en) * | 2012-10-09 | 2014-04-10 | Alstom Renovables Espana, S.L. | Method of operating a variable speed wind turbine |
US20160115941A1 (en) * | 2014-10-27 | 2016-04-28 | General Electric Company | System and method for adaptive rotor imbalance control |
CN108518305A (en) * | 2018-03-27 | 2018-09-11 | 沈阳工业大学自控技术有限公司 | A kind of Wind turbines control method and system |
CN109596255A (en) * | 2017-09-25 | 2019-04-09 | 森维安有限责任公司 | Measure device, wind-driven generator operation method and the wind-driven generator of wind-driven generator torque |
CN109936163A (en) * | 2019-03-27 | 2019-06-25 | 中南大学 | A kind of suppressing method of ripple caused by wind shear effect |
-
2019
- 2019-09-11 CN CN201910857475.5A patent/CN110630448A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2009026930A2 (en) * | 2007-08-31 | 2009-03-05 | Vestas Wind Systems A/S | Method for controlling at least one adjustment mechanism of a wind turbine, a wind turbine and a wind park |
US20140097618A1 (en) * | 2012-10-09 | 2014-04-10 | Alstom Renovables Espana, S.L. | Method of operating a variable speed wind turbine |
US20160115941A1 (en) * | 2014-10-27 | 2016-04-28 | General Electric Company | System and method for adaptive rotor imbalance control |
CN109596255A (en) * | 2017-09-25 | 2019-04-09 | 森维安有限责任公司 | Measure device, wind-driven generator operation method and the wind-driven generator of wind-driven generator torque |
CN108518305A (en) * | 2018-03-27 | 2018-09-11 | 沈阳工业大学自控技术有限公司 | A kind of Wind turbines control method and system |
CN109936163A (en) * | 2019-03-27 | 2019-06-25 | 中南大学 | A kind of suppressing method of ripple caused by wind shear effect |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103410659B (en) | Effective wind speed obtaining method of wind generating set based on High-Gain observer | |
CN107061158B (en) | A kind of prediction of low wind speed leeward power generator and tracking and controlling method | |
Ciampichetti et al. | Sliding mode control of permanent magnet synchronous generators for wind turbines | |
CN111293686A (en) | ARMAX system identification-based real-time evaluation method for inertia of power system | |
Mishra et al. | Comprehensive review on maximum power point tracking techniques: wind energy | |
CN101106279A (en) | Merged network and power adjustment system for direct drive permanent magnetic synchronization generator and its method | |
CN107341299A (en) | The blower fan Reliability Modeling that meter and running environment influence | |
Azad et al. | Parameter estimation of doubly fed induction generator driven by wind turbine | |
CN104675629A (en) | Maximum wind energy capturing method of variable-speed wind generating sets | |
Sarkar et al. | A study of MPPT schemes in PMSG based wind turbine system | |
Yan et al. | Torque estimation and control of PMSM based on deep learning | |
Afrasiabi et al. | Ensemble kalman filter based dynamic state estimation of PMSG-based wind turbine | |
Corradini et al. | Fault-tolerant sensorless control of wind turbines achieving efficiency maximization in the presence of electrical faults | |
Thiringer et al. | Control of a variable-speed pitchregulated wind turbine | |
CN109599889A (en) | DFIG low voltage traversing control method, system under unbalance voltage based on fuzzy active disturbance rejection | |
Syskakis et al. | MPPT for small wind turbines: Zero-oscillation sensorless strategy | |
Tian et al. | A Gaussian RBF network based wind speed estimation algorithm for maximum power point tracking | |
He et al. | Grey prediction pi control of direct drive permanent magnet synchronous wind turbine | |
CN110630448A (en) | Soft measurement method for wind shear effect | |
Gong et al. | Wind speed and rotor position sensorless control for direct-drive pmg wind turbines | |
CN105958886A (en) | Online estimating device and method of observing impeller fatigue service life real-timely based torques | |
El Aimani | Modeling and control structures for variable speed wind turbine | |
Fu et al. | MPPT control based fuzzy for wind energy generating system | |
Hocine et al. | A hybrid sensorless control of PMSG wind-power generator with frequency signal injection method and extended Kalman filter | |
CN114465280A (en) | Dynamic equivalent modeling method for new energy grid-connected system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20191231 |