DE102019135622A1 - Method for determining operating data of a wind turbine - Google Patents

Abstract

A method for determining at least one operating parameter of a wind energy installation is provided. The determination method comprises creating a digital image of the wind energy installation with regard to the at least one operating parameter and acquiring at least one structure sensor signal which reflects a structural behavior of the wind energy installation. By means of a system model that is made available with regard to the wind energy system, the at least one operating parameter of the wind energy system is determined as a function of the structure sensor signal and the digital image.

Description

TECHNICAL AREA

Embodiments of the present invention generally relate to condition monitoring of a wind energy installation, and in particular relate to the determination or estimation of operating data of a wind energy installation on the basis of a combination of sensor data and a model-based approach.

STATE OF THE ART

Current operating data of such systems are required to control and / or regulate wind energy systems. A wind energy installation comprises a large number of sensors which enable statements to be made, for example, about the vibration behavior of rotor blades, a rotor blade mass, a stiffness of a rotor blade, a pitch angle, a rotor speed and an outside temperature. In addition, the extensive control and monitoring of a wind turbine requires that different operating conditions be taken into account. Due to the conditions associated with the operation of a wind turbine, for example pressure and temperature fluctuations, but also in particular strongly changing wind conditions, as well as the large number of legally prescribed safety measures, the monitoring of a wind turbine is a complex process.

The sensors used to monitor a wind turbine are therefore subject to a large number of boundary conditions. Parasitic influencing variables such as pitch angle, rotor speed and outside temperature are often measured directly or also recorded using auxiliary variables such as proxies. A disadvantage of conventional methods occurs, for example, when one or more of the measured variables such as the pitch angle, the rotor speed and / or the outside temperature or the proxies are not available due to transmission problems - because a slip ring is not working correctly, for example. In such a case, the problem arises that the connected system is not ready for operation.

In particular, natural frequencies or natural modes of a rotor blade, for example, are also important properties, the behavior of which can be measured with vibration sensors. Such natural frequencies or vibration characteristics change with, for example, a rotor blade mass and a stiffness or a spring constant of the rotor blade of a wind turbine. A precise analysis of the natural frequencies or natural modes of a rotor blade provides important information about, for example, its structure and its vibration behavior. Once these natural frequencies or natural modes are known, deviations can indicate a change in the state of the system structure.

Many systems have variable operating parameters or are operated under different environmental conditions. This is the case, among other things, with rotor blades of wind turbines, in which, for example, rotor speed, pitch angle, surface loading, angle of flow, flow velocity, etc. vary at short intervals. In addition to changes to these operating parameters, external influences, i.e. environmental conditions such as wind conditions, temperatures, ice cover, moisture, etc. are also subject to constant variations. The natural frequencies and natural modes of the system are therefore dependent on the operating or environmental conditions, wherein such a dependency can differ for each natural frequency or natural mode. A plurality of sensors is used to monitor the operating states of wind energy installations. For example, pressure measurements to determine a pressure on a rotor blade, strain measurements to measure the bending of a rotor blade, acceleration measurements to measure an acceleration of a rotor blade or other variables can be measured.

Due to the large number and complexity of structural and monitoring sensors, their error-free operation cannot always be guaranteed. If one or more structure and monitoring sensors are not ready for operation, there is a risk that an insufficient number of relevant sensor signals will be available. The case can therefore arise that the limited availability of sensor data, for example also due to bandwidth limitation and / or unavailability of cloud data, cannot guarantee reliable monitoring and / or reliable operation of the wind turbine. In many cases, there is the disadvantage that, due to connection problems, relevant data for determining operating parameters is not available. It is therefore desirable to ensure improved operational readiness and improved status monitoring through a robust determination of operating parameters of the wind energy installation.

SUMMARY OF THE INVENTION

• Erststellen eines digitalen Abbilds der Windenergieanlage in Bezug auf den mindestens einen Betriebsparameter, Erfassen mindestens eines Struktursensorsignals, welches ein Strukturverhalten der Windenergieanlage widerspiegelt, und Ermitteln, mittels eines Anlagenmodells, des mindestens einen Betriebsparameters der Windenergieanlage in Abhängigkeit von dem Struktursensorsignal und dem digitalen Abbild.

According to one embodiment, a method for determining at least one operating parameter of a wind energy installation is provided, with the following steps:

• Creation of a digital image of the wind energy installation in relation to the at least one operating parameter, acquisition of at least one structure sensor signal which is a Reflects the structural behavior of the wind energy installation, and ascertaining, by means of a system model, the at least one operating parameter of the wind energy installation as a function of the structure sensor signal and the digital image.

According to a further embodiment, a computer program product is provided which is stored on a computer-readable storage medium and is adapted to execute a method for determining at least one operating parameter of a wind turbine when running on a computer.

Figure list

• 1 zeigt beispielhaft eine Windenergieanlage, deren Zustand mit einem Verfahren zum Bestimmen mindestens eines Betriebsparameters gemäß hierin beschriebenen Ausführungsformen überwacht werden kann;
• 2 zeigt einen Rotor der in 1 gezeigten Windenergieanlage, von vorne gesehen;
• 3 zeigt schematisch ein Auswerteverfahren zum Bestimmen mindestens eines Betriebsparameters einer Windenergieanlage gemäß hierin beschriebenen Ausführungsformen; und
• 4 zeigt ein Ablaufdiagramm zur Veranschaulichung eines Verfahrens zum Bestimmen mindestens eines Betriebsparameters einer Windenergieanlage gemäß hierin beschriebenen Ausführungsformen.

Exemplary embodiments are shown in the drawings and explained in more detail in the description below. In the drawings show:

• 1 shows, by way of example, a wind energy installation, the state of which can be monitored using a method for determining at least one operating parameter according to embodiments described herein;
• 2 shows a rotor in 1 shown wind turbine, seen from the front;
• 3 shows schematically an evaluation method for determining at least one operating parameter of a wind energy installation according to embodiments described herein; and
• 4th shows a flowchart to illustrate a method for determining at least one operating parameter of a wind energy installation according to embodiments described herein.

In the drawings, the same reference symbols denote the same or functionally identical components or steps.

WAYS OF CARRYING OUT THE INVENTION

In the following, reference is made in more detail to various embodiments of the invention, one or more examples being illustrated in the drawings.

Numerous sensors or measuring systems are installed in a wind turbine, which perform a wide range of tasks. For example, sensors are provided in wind power plants which supply sensor data which allow statements to be made about the structural behavior of various components of the wind power plant. The different sensors, which enable statements about the condition of a wind turbine, record, among other things, but not exclusively, sensor data which, for example, indicate vibration modes of a rotor blade of the wind turbine, an ice coating on at least one rotor blade, natural frequencies of the rotor blade or the tower of the wind turbine, shift of natural frequencies Vibration characteristics of a wind turbine structure, an outside temperature and / or a wind direction in the vicinity of the wind turbine, rotor blade noises during a rotation of the rotor of the wind turbine, and also combinations of the above variables.

The above-mentioned sensor data are transmitted to evaluation and control components of the wind energy installation via different interfaces and / or field buses. Such an interface transmits a signal containing the sensor data to other components of the wind energy installation. The interface can in particular be designed as a SCADA interface (Supervisory Control and Data Acquisition Interface). The method according to the embodiments described herein is also particularly suitable when SCADA failures have to be compensated for. The method according to the embodiments described herein has the advantage that no new sensor units have to be provided and even a reduction in the number of sensor systems can be supported.

1 schematically illustrates a wind turbine 200 as an application example for a method for determining at least one operating parameter of the wind energy installation 200 . Different structure sensors 110 can be attached to the wind turbine, the sensor units 110 by means of light guides 112 with an evaluation unit 114 are operationally connected. Such a structure sensor 110 can for example be a mechanical load sensor. One or more rotor blades 100 the wind turbine 200 can have different sensor positions for the sensor unit 110 include. The sensor unit 110 can in particular be designed as a vibration sensor. It is also possible to use the sensor unit 110 as an acceleration sensor, a strain sensor, a temperature sensor, etc. to form. Here, the sensor unit 110 measure the sensor signals on the basis of electrical or optical effects, ie the sensor can be designed as an electrical or as a fiber optic sensor.

Sensor positions of structure sensors 110 can work on the entire structure of the wind turbine 200 be localized. So includes the structure of the wind turbine 200 A tower 202 , a gondola 203 and one with reference to 2 rotor described below 500 with rotor blades 100 . Mounting positions for the sensor units 110 can thus be used by numerous Components of the wind turbine 200 such as on the tower 202 , the gondola 203 , the rotor hub 205 and other places on the wind turbine 200 be provided. After an evaluation, the evaluation unit 114 a signal to a control unit 204 for controlling and / or regulating the wind turbine 200 to transfer. The wind energy installation is controlled, for example, by changing the pitch angle of at least one rotor blade 100 , where the pitch angle is a rotation of the rotor blade 100 around a longitudinal axis 101 of the rotor blade 100 describes.

The length of the rotor blade 100 extends from a blade flange to a blade tip, with no electrical lines being present between the blade flange and blade tip in an expedient manner. For this reason it is advantageous if the sensor units 110 as optical sensors such as fiber optic sensors. Measurement data received from the sensor units 110 are delivered via the light guide 112 to the evaluation unit 114 transferred in an expedient manner in the rotor hub 205 is provided. Every rotor blade 100 can individually distribute sensors individually 110 exhibit. According to embodiments described herein, which can be combined with other embodiments, at least one sensor unit 110 in every rotor blade 100 made available.

2 shows a rotor 500 the in 1 illustrated wind turbine 200 in greater detail. As shown by way of example, the rotor 500 the wind turbine 203 on the rotor hub 205 attached rotor blades 100 on. According to typical embodiments, the rotor 500 at least two rotor blades, in particular three rotor blades. The rotor rotates when the wind turbine is in operation 500 , ie the hub 205 with the rotor blades 100 around an axis. A generator is driven to generate electricity.

In one or more of the rotor blades 100 is at least one sensor unit 110 provided. The signal from the sensor unit 110 is via the light guide 112 to a distributor 501 transfer. Preferably the manifold is 501 on a blade bulkhead of the rotor blade 100 appropriate. As also in 2 shown is a transmission light guide 503 or an optical fiber from the distributor 501 to the evaluation unit 114 made available. The transmission light guide 503 can for example be guided along a spring or a spiral or by a spring or by a spiral or a corresponding mechanical element, so that when the rotor blade rotates 100 around its longitudinal axis 101 , ie when adjusting the pitch angle, no damage occurs.

3 shows a block diagram of an evaluation process 102 to determine the at least one operating parameter 106 . According to an embodiment, which can be combined with other embodiments described herein, in a plant model 105 different data such as image data 108 and parasitic influencing variables 109 processed to at least one operating parameter 106 concerning those in the 1 and 2 wind turbine described above 200 or the rotor 500 the wind turbine 200 to determine. The plant model 105 concerns at least one wind turbine 200 and can be used for this purpose, operating parameters 106 determine which ones cannot be obtained directly from sensor data due to sensor failures or system errors.

As the entire system of a wind turbine 200 may not be ready for operation if a relevant operating parameter 106 is not available, result from the model-based approach with the help of the system model 105 output data have significant advantages. For example, it is possible to estimate parameters such as pitch angle, rotor speed and outside temperature on the basis of signals recorded with a vibration sensor or other structure sensors. According to a further modification, the at least one is an operating parameter 106 selected from the group consisting of a pitch angle of the rotor blade of the wind turbine, a rotor speed, an ambient temperature of the wind turbine, vibration properties of the rotor blade, a force acting on a tower of the wind turbine, a yaw angle of a nacelle of the wind turbine in relation to the wind direction, a Wind speed, a rotor blade temperature, an azimuth angle and any combination thereof.

To one or more relevant operating parameters 106 from the plant model 105 It is necessary to obtain alternative data sources for the operating parameters 106 to use. This can either be directly measured data such as structural sensor data 104 that are obtained by means of structure sensors such as vibration sensors, or that are obtained from a digital image 103 determined image data 108 act. The directly measured data can be data from structure sensors, ie structure sensor signals 104 act, with structure sensors on different components of the wind turbine 200 may be appropriate. A structural behavior of the wind energy installation determined with the structure sensors 200 can for example be based on vibration measurements that can be carried out with vibration sensors attached to rotor blades. For example, it is possible to measure over Inferring rigidity data, which may have been determined with structure sensors, to a speed of the rotor of the wind energy installation or a temperature on the rotor blade.

According to a further embodiment, which can be combined with other embodiments described herein, the acquisition of the structure sensor data can 104 creating a frequency distribution of one or more data signals from structure sensors and deriving the at least one operating parameter 106 from the frequency distribution. The structure sensor data 104 be selected from the group consisting of vibration modes of the rotor blade of the wind turbine, natural frequencies of the rotor blade or the tower of the wind turbine, shift of natural frequencies, vibration characteristics of a wind turbine structure, an outside temperature and / or a wind direction in the vicinity of the wind turbine, rotor blade noise during a Rotation of the rotor of the wind turbine, expansion, bending moments and any combination thereof.

For example, vibration eigenmodes of an oscillatable system such as a rotor blade 100 of a rotor 500 the wind turbine 200 detected, it is possible to use the at least one structure sensor signal 104 by creating the frequency distribution of the detected natural oscillation modes as a function of a mode frequency. The plant model 105 can also include other data such as data from other wind turbines 200 or their sensor data for determining the at least one operating parameter 106 draw in. Further data related to the system model 105 can also be supplied, for example, can also include external data from meteorological sensors.

The digital image 103 the wind turbine 200 to determine the image data 108 can in particular, alone or in combination with the system model 105 , as a digital twin 113 - digital twin - be trained. The digital twin 113 be provided as a model-based approach, with which the parasitic influencing variables 109 can be determined. According to a further modification, the creation of a digital image 103 the provision of one or more digital twins.

In particular, the determination method presented in relation to different embodiments described herein results in the advantage that a condition monitoring system provided in the wind energy installation is available even if the parasitic influencing variables cannot be supplied via the fieldbus. In this case, parameters such as the pitch angle, the rotor speed and the outside temperature are estimated on the basis of the quantities measured with a structure sensor such as a vibration sensor. The structure sensor signals supplied by the structure sensor are then used as a data source for the operating parameter to be determined. On the one hand, it is possible to use the parasitic influencing variables measured directly from the fieldbus as a data source. On the other hand, there is the possibility of using variables calculated with the so-called digital twin as a data source.

It should be pointed out here that the expression “digital twin” is a digital representation of a physical structure or a system in connection with a wind turbine to be monitored 200 describes. Here, a digital twin can include a computer program which uses known input parameters to determine the operating parameters of its counterpart. Here, the digital twin corresponds to a model of an object to be examined, such as the wind turbine, for example 200 . In addition, it is possible for digital twins to have different wind turbines 200 are assigned, communicate with each other or exchange data.

According to an embodiment, which can be combined with other embodiments described herein, the plant model 105 both with the digital image 103 determined image data 108 as well as parasitic input variables 109 fed. The parasitic input quantities 109 are in a structure signal acquisition 111 provided, which on the one hand structure sensor data 104 and on the other hand fieldbus data 107 can be fed. At least one of the above-mentioned data sources is used to determine the at least one operating parameter 106 the wind turbine 200 used. As in 3 illustrated, the structure signal acquisition can 111 both structure sensor data 104 as well as fieldbus data 107 are supplied, from which, individually or in combination, parasitic influencing variables 109 for further processing in the system model 105 can be determined. According to a further modification, it is possible to use fieldbus data 107 from at least two wind turbines 200 of a wind farm to provide the parasitic data 109 to use.

According to a further embodiment, which can be combined with the embodiments described herein, measured and calculated influencing variables can be combined to produce an output uncertainty for the operating parameter 106 further decrease. The combination of measured and calculated influencing variables can take place for example by means of a Kalman filter. According to a further modification, it is possible to have a combination of creating a digital image 103 and detecting the at least one parasitic influencing variable 109 using the Kalman filter.

According to a further embodiment, which can be combined with other embodiments described herein, the system model can 105 include a machine learning method based on statistical predictions to determine the operating parameter 106 contributes. In particular, the system model 105 include a Bayesian, continuously learning hybrid model. According to a further modification, it is then possible to combine known physical relationships and direct measurements in a hybrid model to create the system model 105 to continuously improve. According to a further modification, the measurement data fed to the system model can be used for continuous and / or automated model improvement. With a cloud-based approach, such training data is pushed into the cloud for further use.

4th shows a flow chart to illustrate a method for determining at least one operating parameter 106 a wind turbine 200 according to embodiments described herein.

In detail, the method includes for determining at least one operating parameter 106 a wind turbine 200 the blocks 301 to 305 . The procedure is in one block 301 started. In a subsequent block 302 a digital image of the wind energy installation is created in relation to the at least one operating parameter. It should be pointed out here that the digital image can also be created corresponding to two or more operating parameters.

Since the wind turbine has structure sensors such as vibration sensors on different components, it is possible in one block 303 at least one structure sensor signal can be detected. This structure sensor signal, which is output from the corresponding structure sensor, corresponds to a structure behavior of the wind energy installation and is used to make a statement with regard to the operating parameters with the aid of the digital image. In the example mentioned, the vibration behavior of the rotor blade will change both in the event of a temperature variation and, for example, in the event of bearing damage.

Thus, the measured vibration behavior of at least one rotor blade can be fed back into the model-based approach and, in one example, can be used to detect ice on the at least one rotor blade. In another example, the detection of the vibration behavior of a rotor blade of the wind energy installation can be used to draw model-based conclusions about other relevant operating parameters. For example, if an acceleration in all three rotor blades of the in 2 shown rotor 500 the wind turbine 200 measured, a total acceleration can be detected by means of the acceleration sensors arranged in the rotor blades and conclusions can be drawn about an imbalance in the rotor.

Eventually it will be in a block 304 which determines at least one operating parameter of the wind turbine based on the structure sensor signal in combination with the digital image. Such a determination takes place by means of the system model, which in combination with the digital image can also be provided as a digital twin. The determination procedure ends in a block 305 .

In particular, it is advantageous that the determination method according to the embodiments described herein can use raw data from vibration sensors in order to be able to utilize such a data source for determining the operating parameters with the aid of the digital twin or the system model. Such an estimation of parasitic influencing variables from vibration sensor data in particular can then advantageously be carried out using a model-based approach. With the method according to embodiments described herein, it is also possible to provide event-based applications and a tracking of slowly progressing changes of different operating parameters. Event-based data can also be used in an expedient manner for training the system model.

In addition, there is the advantage that a data rate that is supplied by functional sensors can be reduced if a digital image is used. This means that high-frequency data transmission from different sensors to a condition monitoring unit does not have to be ensured in every case.

According to one embodiment that can be combined with other embodiments described herein, a computer program product is provided which is stored on a computer-readable storage medium and is adapted to provide a determination of at least one operating parameter of a wind turbine when running on a computer. The computer program product can contain software code sections with which steps according to one or more of the other aspects can be carried out when the computer program product runs on a computing device.

Such a computing device can contain, in addition to one or more memory units, a CPU, signal inputs and signal outputs, and other elements typical of a computing device. A computer program product can contain software code sections with which the steps of the method of the embodiments described here can be carried out at least partially when the computer program product is running on the computing device. Any embodiments of the method can be carried out by a computer program product.

Although the present invention has been described above on the basis of typical exemplary embodiments, it is not restricted thereto, but rather can be modified in many ways. The invention is also not restricted to the possible applications mentioned. It should also be pointed out at this point that the aspects and embodiments described herein can be appropriately combined with one another, and that individual aspects can be omitted where it is sensible and possible within the framework of professional action. Modifications and additions to the aspects described herein are familiar to the person skilled in the art.

Claims (10)

Method for determining at least one operating parameter of a wind turbine, with the steps: Creating a digital image of the wind energy installation in relation to the at least one operating parameter; Detecting at least one structure sensor signal which reflects a structure behavior of the wind energy installation; and Determining, by means of a system model, the at least one operating parameter of the wind energy system as a function of the structure sensor signal and the digital image. Procedure according to Claim 1 , wherein at least one parasitic influencing variable is detected from the structure sensor signal and fieldbus data of the wind energy installation. Procedure according to Claim 1 or 2 , a digital twin of the wind energy installation being created by means of the installation model on the basis of image data of the digital image. Method according to one of the preceding claims, wherein the operating parameter is selected from the group consisting of a pitch angle of the rotor blade of the wind turbine, a rotor speed, an ambient temperature of the wind turbine, vibration properties of the rotor blade, a force acting on a tower of the wind turbine, a yaw angle a nacelle of the wind energy installation in relation to the wind direction, a wind speed, a rotor blade temperature, an azimuth angle and any combination thereof. Method according to one of the preceding claims, wherein data of the digital image are used to determine the at least one operating parameter when at least one parasitic influencing variable is not available. Method according to one of the preceding claims, wherein the structure sensor data are selected from the group consisting of vibration modes of the rotor blade of the wind turbine, natural frequencies of the rotor blade or the tower of the wind turbine, shift of natural frequencies, vibration characteristics of a wind turbine structure, an outside temperature and / or a Wind direction in the vicinity of the wind energy installation, rotor blade noises during a rotation of the rotor of the wind energy installation, expansions, bending moments and any combination thereof. Method according to one of the preceding claims, wherein a combination of creating a digital image and detecting the at least one parasitic influencing variable is carried out by means of a Kalman filter. Method according to one of the preceding claims, further comprising creating a frequency distribution of parasitic influencing variables and deriving the at least one operating parameter from the frequency distribution. Method according to one of the preceding claims, wherein field bus data from at least two wind energy installations of a wind park are used to record the at least one parasitic influencing variable. Computer program product, which is stored on a computer-readable storage medium and is adapted to run on a computer, the method according to one of the Claims 1 to 9 to execute.

Images