Wind monitoring arrangement

Abstract

The invention is directed to a wind monitoring arrangement (100), comprising one or more wind monitoring units the comprise a RF-transceiver (106) configured to provide and receive RF-signals (108), and defining a sensing volume (110), in which a respective target object (112) is located that has a position depending on a wind state (W). A signal quality data determination unit (114) is configured to determine signal quality data (SQ) indicative of a value of a signal quality parameter that is correlatable to the position of the target object. Current target object data (TOD1) is generated by associating current time data to the signal quality value. A processing unit (118) is configured to determine, using the current target object data and a wind-data determination algorithm (120), wind data (WD) indicative of the current wind state in at least the sensing volume, thereby reducing the hardware requirements for wind monitoring.

Description

[0001] Wind monitoring arrangement

[0002] FIELD OF THE INVENTION

[0003] The present invention is directed to a wind monitoring arrangement, to a method for controlling operation of a wind monitoring arrangement and to a computer program.

[0004] BACKGROUND OF THE INVENTION

[0005] Wind gusts in an urban environment create a range of hazards. As the windpressure scales with square of wind speed, most wind damage in a city is caused by wind gusts rather than sustained wind. A wind gust, or gust, is a brief increase in wind speed. Gusts occur basically anytime the wind is blowing. Gusts become more noticeable and impactful as the wind speed increases. This is because the force exerted by the wind on a person or object increases rapidly with increasing wind speed since, the wind pressure roughly scales with square of wind speed. Wind gusts occur because the air is not able to move along the ground at a uniform rate. The gusts are strongly influenced by the local area the wind is blowing into and hence varies strongly even within the same city. Surface friction from vegetation, land, the surrounding buildings and elevation changes will hyper-locally slow the wind in some places more than others. The air flow close to the ground is also influenced more by friction than air flow higher up. This creates a much more turbulent wind flow along the ground which is experienced as gusts. While urban wind has hitherto mainly been studied for individual buildings or individual street canyons, there is a high interest by cities to map & model the wind for larger urban areas. For instance, micro-scale modelling of the urban wind speed is a crucial element for a city to hyper-locally predict the air pollution at critical spots.

[0006] US 2016 / 0113092 Al describes a street lighting fixture containing an accelerometer that is used to detect and characterize acceleration events on the street lighting fixture. The accelerometer data can be used for monitoring wind speeds or detection of winds exceeding certain limits. SUMMARY OF THE INVENTION

[0007] Despite of many good use cases for city-wide observation-based wind speed models, financial and technical constraints are preventing the installation of dedicated wind speed measurement equipment to assess the urban wind speed. There is hence a need to estimate the urban wind speed at many locations in the city without extra hardware investments and minimized cost of cellular connectivity. It would therefore be beneficial to provide a wind monitoring arrangement with lower hardware investment and lower costs of cellular connectivity.

[0008] A first aspect of the present invention is formed by a wind monitoring arrangement. The wind monitoring arrangement of the first aspect of the invention comprises one or more wind monitoring units. Each of the wind monitoring units comprises a radiofrequency (RF) transceiver that is configured to provide and receive radiofrequency signals, wherein the radiofrequency transceiver is arranged defining a sensing volume, in which a respective target object is located, the target object having a position which depends on a current wind state in a vicinity of the target object.

[0009] In the wind monitoring arrangement of the first aspect of the invention, the wind monitoring unit comprises a signal quality data determination unit that is connected to the radiofrequency transceiver and configured to, using the received radiofrequency signals, determine signal quality data indicative of a current signal quality value of a predetermined signal quality parameter of the received radiofrequency signal, the signal quality parameter being correlatable to the position of the target object, and a data provision unit, that is connected to the signal quality data ascertaining unit and configured to ascertain time data indicative of a current time, to associate the current time data to the current signal quality value for generating current target object data and to output the current target object data.

[0010] Further, the wind monitoring arrangement of the first aspect of the invention comprises a processing unit that is configured to receive, from the data provision unit, the respective current target object data and to determine, using the current target object data and using a wind-data determination algorithm, wind data indicative of the current wind state in at least the sensing volume.

[0011] According to the invention, the radiofrequency transceivers of each of the wind monitoring units that form a pair of wind monitoring units define a sensing volume. Sensing volume refers to that volume of space, in which changes in the position of the target object have a measurable influence on the value of a given signal quality parameter of the radiofrequency signal provided by one of the transceivers and received by the other transceiver.

[0012] The signal quality determination unit is connected to the radiofrequency transceiver or receiver, and using the received radiofrequency signal sent by the radio frequency transceiver or transmitter is configured to determine signal quality data that is indicative of a current (or instant) signal quality value of the given signal quality parameter, and thus, correlated to the actual position of the target object that is located in the sensing volume. Thus, the transmission and reception of the radiofrequency signal and the determination of the signal quality data are considered an passive RF-sensing method for inferring a position of the target object. The target object is an object located within the sensing volume, whose current position is influenced by, or depends on, the current wind state.

[0013] The data provision unit ascertains time data indicative of the current point in time in which the signal quality value is determined and associates both time data and the current signal quality value, thereby generating the current target object data, that is correlated to the position of the target object at a given time. This current target object data is provided to the processing unit, which, using the current target object data provided by the one or more data provision units (in principle one for each wind monitoring unit), and using a wind-data determination algorithm, determines wind data that is indicative of the current wind state (the wind state at the time ascertained by the data provision unit and that was associated to the current signal quality value) in the sensing volume. Since the position of the target object can be inferred by its effects on the received radiofrequency signal and the wind state alters the position of the target object in a predictable manner, the processing unit is advantageously configured to determine the wind data that is indicative of the current wind state in the sensing volume, e.g., in the vicinity of the target object by analyzing the current target object data according to the wind-data determination algorithm, without the need of dedicated accelerometers or wind sensors.

[0014] In the following, embodiments of the wind monitoring arrangement of the first aspect of the invention will be described.

[0015] In general, the invention also applies to pairs of wind monitoring units according to the first aspect of the invention, wherein both of them comprise a radiofrequency transceiver, or wherein a first wind monitoring unit comprises a radiofrequency transmitter and a second wind monitoring unit comprises a radiofrequency receiver. Here the sensing volume is a region of space located between the two RF- transceivers (or the transmitter and the receiver). The radiofrequency signal provided by the first wind monitoring unit is received, after interacting with the target object, as a received radiofrequency signal at the radiofrequency transceiver of the second wind monitoring unit.

[0016] In a preferred embodiment, the wind monitoring units, or the pairs of wind monitoring units, are arranged on street luminaires or street lights, in particular, wirelessly controllable street luminaires of an intelligent lighting system. For instance, in an embodiment, the street lighting fixtures include an outdoor lighting controller (OLC) enabling the formation of a smart, responsive network for operating a lighting arrangement, for instance a connected city lighting arrangement (see, e.g. Inter Act city). These lighting fixtures already include the necessary communication hardware and can be advantageously configured to perform wind monitoring according to the present invention without the need of dedicated wind sensors and without an increased need of communication capabilities.

[0017] In an embodiment, the signal quality parameter is a received signal strength indicator (RS SI) indicative of the signal strength of the received radiofrequency signal, which depends on the initial signal strength of the radiofrequency signal provided and on the interactions of the radiofrequency signal with the objects located in the sensing volume. Additionally, or alternatively, the signal quality parameter is a channel state information (CSI) that is indicative of the channel properties of a communication link established between a transmitter and a receiver or between two transceivers. The CSI is indicative of how the signal propagates and may represent a combined effect of scattering, fading, power decay, channel gain, line of sight component, spatial correlation, etc. In other embodiment, the signal quality parameter is any parameter whose instant value can be correlated to a position of the target object within the sensing volume. In other embodiments, the signal quality parameter includes a frequency value, or a frequency shift value.

[0018] In an embodiment, the radio frequency transceivers are configured to provide, as the radiofrequency signals, wireless communication signals in accordance with a predetermined wireless communication protocol. The radiofrequency signals are therefore communication signals in accordance with said wireless communication protocol. The radiofrequency communication signals can be transmitted according to, for example, Zigbee, WiFi, 2G, 3G, 4G, 5G, Bluetooth, Thread, or any other suitable wireless communication protocol. Preferably, the radiofrequency transceivers include a cellular radio and / or a Zigbee radio.

[0019] In another embodiment, the radio frequency transceivers are additionally or alternatively configured to provide, as the radiofrequency signals, radar signals. In this particular embodiment, the radiofrequency transmitter and the radiofrequency receiver maybe integrated in a common wind monitoring unit, or in a common street lighting fixture. The radiofrequency transmitter, configured as a radar transmitter emits a radiofrequency signal in the form of radar waves, for instance WiFi radar waves. The waves interact with the target object, and the signal is reflected back to radiofrequency receiver, configured as a radar sensor. The reflected radiofrequency signal arriving at the radar sensor contains information about the position of the target object. The radar unit can be integrated into a so-called outdoor sensor bundle (OSB) of the street lighting fixture or in the outdoor luminaire controller (OLC) of the streetlighting fixture.

[0020] In an embodiment, the current target object data is indicative of a movement of the target object, wherein the movement is correlated to a presence of wind in a vicinity of the target object. Additionally, or alternatively, the current target object data is indicative of a direction of movement of the target object, wherein the direction of movement is correlated to a direction of the wind in the vicinity of the target object. Additionally, or alternatively, the current target object data is indicative of a displacement of the target object, wherein the displacement amount is correlated to a strength of the wind in the vicinity of the target object. The displacement may relate to a displacement amount and / or to a displacement rate correlated to a speed at which the position of the target object changes in the time domain. For instance, if the position of target object changes slowly, this is indicative a of gradually increasing wind speed / wind strength. In another case, the target object may exhibit a fast change in its position indicative of a gust with a sharper onset.

[0021] In an embodiment the target object is an object that has a first section that has a fixed position with respect to the monitoring units (e.g., a trunk / roots of a tree or bush, a pole of a flag, etc) and a second section whose position with respect to the monitoring units is not fixed and may vary in dependence on the current wind state (e.g., branches or leaves of a tree / bush, fabric of a flag, etc.).

[0022] In an embodiment, the target object is a flag pole with a flag, a traffic sign, a canopy, an awning, etc.

[0023] Preferably, in another embodiment, at least one of the target objects is a tree. In the particular embodiment, the current target object data is indicative of a movement of one or more branches of the tree and / or of a direction of movement of the one or more branches of the tree that is correlated to a direction of the wind in the vicinity of the tree and / or of a displacement amount or displacement rate of the one or more branches of the tree which is correlated to a strength or speed of the wind in the vicinity of the tree. The current target object data, or the current tree data, in the case where the target object is a tree, in addition to the time data indicative of an arrival time stamp of the wind gust also comprises information about wind state (i.e. its influence on the position of the target object) at each street lighting pole. The radiofrequency transceivers (e.g. Zigbee radio, cellular radio, radar sensor, etc.) are used for perform RF-sensing and / or radar sensing to infer from the movement of target object (e.g. the tree branches) that a wind gust has arrived at a specific location in the city. The RF sensing can also be advantageously used to analyze the movement direction of the target object (e.g., the tree branches as indicator for the directionality of the wind gust (e.g. the branch movements indicate a south-south- westerly wind gust).

[0024] Preferably, in an embodiment, the data provision unit comprises a synchronized GPS time clock unit for ascertaining the time data. The GPS time clock unit provides a hyper accurate time clock.

[0025] In a preferred embodiment, the processing unit comprises a machine learning engine configured to run a determination algorithm, in particular a clustering algorithm, as a wind-data determination algorithm. The machine learning engine is configured to use, as input data, the respective current target object data provided by one or more pairs of wind monitoring units, that are indicative of positions of respective target objects at a respective given time for determining the wind data, in particular wind data indicative of a wind gust.

[0026] A clustering algorithm is advantageously used to identify which sets of current target object data, recorded by different pairs or wind monitoring units are associated with the same specific wind gust progressing through space. The clustering algorithm utilizes input features derived from the target object data such as the arrival time of the gusts observed at each monitoring unit, the directionality of the target object displacement (indicative of the gust’s direction), the rate of target object displacement (gust with very fast increasing windspeed vs slowly increasing windspeed) and the gust’s relative strength based on the displacement of the target object. After clustering the observed target object movements into individual wind gusts, the machine learning engine or model infers, for instance based upon the GPS-timestamped RF-sensing data (which is preferably concurrently recorded by multitude pairs of monitoring units in a given area or neighbourhood), wind data indicative of the current wind stare, that includes, for example, data indicative of the wind speed, the duration and the wind direction of an specific gust.

[0027] In another embodiment, wherein the wind monitoring arrangement of any of the preceding claims comprises a plurality of wind monitoring units or of pairs of wind monitoring units, the wind monitoring arrangement further comprises a selection unit that is connected to the processing unit and configured to, using the determined wind data, to determine a subset of wind monitoring units, or of pairs of wind monitoring units, for performing the wind monitoring (i.e., for determining and / or providing the current target object data based on which the wind data is determined) and to control operation of the wind monitoring units such that the provision of the current target object data is solely performed by the determined subset of wind monitoring units or of pairs of wind monitoring units. This control of the selected wind monitoring units enables a reduction of the data exchange and can be performed for a predetermined amount of time or until the wind data indicates, according to a predefined criterion, that the wind gust is no longer detected. At this point, the selection unit can be configured to control operation of the wind monitoring units such that all the wind monitoring units determine and provide the current target object data again.

[0028] As first step, a wind monitoring unit or a pair of wind monitoring units, for instance a pair of street lighting OLCs and / or OSBs are used for passive wind sensing as explained above utilizing the respective arrival time of wind gusts at the target object. The passive RF sensing of the position and / or movement of the target object (e.g., tree branch) located in the sensing volume eliminates the need for dedicated sensors such as the mechanical wind wheel as used by conventional weather stations. The RF sensing also allows identifying a direction of movement of the target object as indicator for the directionality of the current wind gust.

[0029] The selection unit enables the use of a predetermined selection criteria to determine which subset of all available wind monitoring units or pairs of wind monitoring units within the city to select for continuously reporting out the current target object data indicative of the position of the target object and the time at which the signal quality value indicative of the position was determined. Preferably tree branch movements are determined as they exhibit strong physical displacement in the windy conditions than displacement of the poles or traffic signs.

[0030] In an embodiment, the processing unit is configured to retrieve from many wind monitoring units or pairs of wind monitor units in the city the arrival timestamps of the most recent wind gusts. To subsequently limit cellular communication cost, according to an embodiment, the selection unit is configured to dynamically select based on the initial current target object data from many monitoring units a suitable subset of wind monitoring units or of pairs of wind monitoring units to perform the continuous gust monitoring. Preferably, this selection is based on the wind direction in this specific part of the city as determined from the initial current target object data.

[0031] For instance, in an embodiment wherein the wind monitoring arrangement is implemented in an connected InterAct city lighting arrangement, the selection unit may purposefully select a number (e.g., four) InterAct city street lights which are lined up in direct line of sight of each other and where these four street lights are significantly aligned (e.g. less than 45°, more preferably less than 30°, ever more preferably less than 20°) with the current direction the wind. In other words, the four (or any other number) of wind monitoring units on the first street are selected to continuously report their observed current target object data as they are downstream of each other with respect to the direction of the wind, whereas other pairs of available wind monitoring units on a second street where the wind is blowing orthogonal to the orientation of the roadway are not selected.

[0032] If the current wind direction is not blowing parallel to any of the streets in the neighbourhood, the selection unit may be configured to select a subset wind monitoring units or of pairs of wind monitoring units which are located on multiple streets, for instance at an intersection.

[0033] In another embodiment, the pair-selection unit is further configured to ascertain location data indicative of a respective location of the wind monitoring units, and wherein the determination of the subset of wind monitoring units or pairs of wind monitoring unit is further based on the location data. The location data can be determined using a GPS unit and may also include information pertaining to the surroundings of the wind monitoring unit, for instance pertaining to the surrounding terrain or buildings. For instance, the wind state at a luminaire location also depends on the terrain and surrounding building effects (e.g. wind shadows). Hence, when selecting the subset of wind monitoring units for the wind gust detection, the location of the respective wind monitoring units with respect to the built structure and the current wind direction are also taken into account. Similarly, we the selection of the subset of wind monitoring units for the wind gust detection task can further be adjusted based on the expected wind shadows from the buildings and trees given the current prevalent wind direction.

[0034] In yet another embodiment, the selection unit is further configured to ascertain current weather data indicative of a wind direction and / or wind speed of wind at a given location, and to further use the received weather data for the determination of the subset of wind monitoring units or of pairs of wind monitoring units. For instance, the selection unit can be signally connected to a local weather station data (e.g. from the airport) and configured to receive weather data therefrom. The weather data can be indicative of a direction of the wind. For example, the weather data can indicate that the wind in the city or in a specific neighborhood or area is currently coming from a given direction (e.g., north-easterly direction). The selection of the subset of wind monitoring units or of pairs thereof may also be based on information about adjacent buildings (obtained, for instance, from 3D satellite images) and trees located close to the wind monitoring units (obtained, for instance from RF sensing data or from satellite images).

[0035] Most often wind gusts usually come in 2-minute intervals. Hence, given the typical spacing between wind monitoring units, such as those integrated in street lighting poles (~50m), most of the time no overlap is expected between subsequent wind gusts. However, in an embodiment, and to eliminate classification errors associated to unusually closely spaced wind gusts (and / or deal with a highly non-linear wave front of a wind gust; and / or deal with a larger angel between the direct line connecting the two lighting pole pairs and the direction of the wind vector), the clustering algorithm also takes the duration of the gust events into account. For instance, a first wind gust may have lasted 10 seconds while a second wind gust started 5 second after the first wind gust stopped and the second wind gust lasted 20 seconds. Adding the wind gust duration as input feature for the clustering algorithm improves the identification of which of the brief increases of the various wind speed observed by the different pairs of wind monitoring units are related to the very same wind gust progressing through an area or neighborhood. To uniquely assign wind events observed at different pairs of wind monitoring units to a specific wind gust, the maximum absolute wind speed of an observation or the time series wind speed changes within the same gust can also be used.

[0036] This approach hence enables us to accurately report the number of wind gusts and the speed of the wind gusts, for instance for a InterAct City dashboard.

[0037] Preferably, in an embodiment, the determination of the subset of pairs of wind monitoring units is performed based on a orientation of the RF sensing channel (i.e., the relative position of both wind determination units) with respect to the wind direction .

[0038] In principle, it is preferred that the determined subset of wind monitoring units or pairs of wind monitoring units, result in a maximum radiofrequency sensing signal response, i.e. a larger effect of a location change of the target object on the received radiofrequency signal. The orientation of the sensing volume, also referred to as RF sensing zone, between the first and second monitoring unit (the RF sensing zone has typically an American football like shape) can be substantially parallel or substantially perpendicular to the wind direction. Based thereon, the a pair of wind monitoring units can be selected on the same street side, or on different sides of the streets.

[0039] In general, the determination of current target object data is most sensitive for target object movements (e.g. tree branch movements) that are perpendicular to an imaginary connection line between the pair of wind monitoring units. Hence, purposefully selecting the subset of pairs of wind monitoring units so that the RF sensing wireless channel (which is along the imaginary line connecting the wind monitoring units of a pair) is aligned with the wind direction will give in principle the strongest RF sensing signal as the wind moving a tree branch which is located on a tree in the middle between the two streetlights will create the maximal interferences on the RF sensing signal.

[0040] It should be noted, however, that if the branch (or the target object in general) is located very close to the radiofrequency transceiver unit, the strongest change in RF sensing signal, and thus on the signal quality value will be caused if the wind moves the branch away from blocking the transceiver (as a branch in the near field will significantly block the wireless transmission to the second transceiver).

[0041] In a further embodiment the processing unit is configured to carry out a machine learning algorithm based on a random forest regression model to predict the wind speed at a large number (e.g., 1000) of locations in the city, given a boundary condition of wind speed and RF-sensing based wind direction observations at just a subset of the original locations (e.g., 25 locations).

[0042] As the first boundary condition area- wide and / or city-wide wind model, the wind speed measured or otherwise determined by a weather station, (e.g., the airport weather station is used. Additionally a number of wind speed boundary conditions (e.g., 25) from the best pairs of wind monitoring units (also 25) are used. These 25 pairs of wind monitoring units are tasked with determining wind data indicative of the wind state (e.g. direction and / or strength) with the help of RF sensing based monitoring of the movement of the target object as explained above.

[0043] During the training phase of the model, wind data is collected from all (e.g., 1000) pairs of wind monitoring units in the city or area. The random forest model is then trained to use the time-stamped current target object data observed by well-selected 25 pairs of wind-monitoring units (plus the weather data from the weather station) to make an output inference for the wind speed at all the pair locations (e.g. 1000). This way, the algorithm 1 earns during the training phase of the model the relationship between the 25+1 wind data observations (as inputs) and the 1000 determined wind data at each of the 1.000 monitoring location (as outputs).

[0044] Once the model has been trained for this specific city or city, it can infer the wind speed at each of the 1000 locations in the city solely based on observations from a subset of pairs of wind monitoring unit (e.g. 25) across the city. Hence, as only 25 pairs of wind monitoring units need to report their current target object data, the cellular or data connectivity cost of the wind monitoring arrangement for the city or area is greatly reduced while still retaining high spatial granularity.

[0045] In yet another embodiment, after the selection unit has selected the subset of most reliable wind monitoring units or pairs thereof for determining the wind data (e.g. wind speed, and wind direction or angle) in a given area or neighborhood of the city, the inferred wind data is provided to a weather service business (e.g. Google Brain) to train a city-wide weather prediction machine learning algorithm. Similarly, our the inferred wind data can serve as input features for the inferences of an external hyperlocal weather model.

[0046] In an embodiment, the determined wind data can be used to generate a proactive warning ahead of the arrival of a strong wind gust at a location. A knowledge of the wind direction and speed at a given location is then used to forecast the effect of a wind gust at another downwind location. For instance, on a highway section equipped with dynamic road signs, a wind gust alert symbol can be displayed in dependence on the determined wind data, or the speed limit on a road section can be dynamically adjusted in a proactive manner.

[0047] Air pollution from traffic is hyperlocal. For instance, hyperlocal air pollution can lead to the closure of certain areas for traffic when the lack of wind leads to large pollution levels in violation of official health rules.

[0048] Presently regulatory air pollution models have to rely on the modelled hyperlocal urban wind speed and wind direction as input. Typically, the urban wind speed is modelled using either large-scale numerical weather prediction models, physical models (e.g. wind or water tunnels), computational fluid dynamics models or models based on analytical horizontal and vertical extrapolations. However, all of the currently available wind speed modeling options are very cumbersome for the city (e.g. building physical model of a city and testing it in a wind tunnel is extremely expensive) or are very computing intensive. In addition, the models are not very accurate as local wind gusts are influenced by many factors. Researchers for instance find in their urban wind model efforts that the inhomogeneity of a city’s underlying surfaces and the high roughness arising from architectural complexes make the urban boundary layer very different from the easily understood atmospheric boundary layer on homogeneous and flat terrain, especially for the urban canopy layer that extends from the ground to the roofs of buildings. In addition, the local wind gusts even depend on the temperature and horizontal temperature gradients.

[0049] Therefore, in an embodiment the hyperlocal wind data determined by the wind monitoring arrangement of the first aspect of the invention, (e.g. inferred by the RF sensing data from the wind monitoring units streetlights) can be provided as input for a machine learning model (e.g., a third party machine learning model) that can infer the hyperlocal pollution level on a specific street at a later point in time.

[0050] In a further embodiment the hyperlocal wind data is advantageously used to flag local wind nuisances for pedestrians. The identified wind nuisances can be subsequently provided for urban planning purposes.

[0051] Similarly, the determined wind data may be used for assessments of wind loadings on tall structures or to provide accurate estimations of the maximum wind gust speeds at an specific location.

[0052] Additionally, GPS equipped street lights are in practice sometimes inaccurately located on the street map of a dashboard. This is due to GPS artifacts and / or the inherent limitations of GPS in a densely built or tree covered environment. Therefore, in another embodiment, the time stamped current target object data and / or the resulting wind data is advantageously used to identify and / or flag inaccurately performed GPS locations of street lights and present them for consideration to an operator of the lighting system, e.g. the InterAct City street lighting maintenance manager.

[0053] For instance, it is first determined that the current wind direction is aligned with a number of wind monitoring units (e.g. street lights), which according to the InterAct City map are located along a specific road section. Based on the map and the current wind direction, it is determined that the second street light should be downstream of the first street light and the third street light should be downstream of the second street light. It is then checked whether the respective time stamps of the wind-state related current target object data determined by the pairs of wind monitoring units are consistent with their respective locations of these three street lights as recorded on the InterAct City map. If there is an inconsistency, a potential GPS location fix-error is flagged and / or even corrected. For instance, the map location of the second and third street lights can be swapped.

[0054] In another embodiment, besides performing distributed wind gust detection with the OLC’s cellular / Zigbee radio as described above, additionally OSB’s microphone audio data can be used where available. The availability of even a limited number of OSBs capable of performing audio-based wind gust detection in an area will provide valuable ground-truth data for customizing our wind monitoring arrangement in general and the machine-learning model in particular to the specific area or city.

[0055] In prior art wind monitoring, the sonic anemometers used by weather services are typically mounted 4 m to 5 m above surface level, hence roughly the height of streetlight. Street lighting luminaires are hence very well positioned to perform granular wind gust detection in a city. When selecting which subset of wind monitoring units to assign for the wind monitoring, preferably pairs of wind monitoring units with similar mounting height above ground are selected.

[0056] A second aspect of the present invention is formed by a method for controlling operation of a wind monitoring arrangement. The method comprises:

[0057] - providing by a first radiofrequency transceiver (or radiofrequency transmitter) radiofrequency signals,

[0058] - receiving at a second radiofrequency transceiver (or radiofrequency receiver), the provided radiofrequency signal, wherein the first and the second transceiver are arranged defining a sensing volume, in which a respective target object is located, the target object having a position which depends on a current wind state,

[0059] - using the received radiofrequency signals, determining signal quality data indicative of a current signal quality value of a predetermined signal quality parameter of the received radiofrequency signal, the signal quality parameter being correlatable to a position of the target object;

[0060] - ascertaining time data indicative of a current time

[0061] - associating the current time data to the current signal quality to generate current target object data

[0062] - outputting the current target object data,

[0063] - determining using the current target object data and using a wind-data determination algorithm, wind data indicative of the current wind state in at least the sensing volume.

[0064] Thus, the method of the second aspect of the invention shares the advantages of the wind monitoring arrangement of the first aspect.

[0065] In the following, embodiments of the method of the second aspect will discussed.

[0066] In an embodiment, the method further comprises: - providing, as the radiofrequency signals, wireless communication signals in accordance with a predetermined wireless communication protocol; and / or

[0067] - providing, as the radiofrequency signals, radar signals; and / or

[0068] - ascertaining the time data using a synchronized GPS time clock.

[0069] In yet another embodiment, the method additionally or alternatively comprises determining a subset of pairs of wind monitoring units for performing the wind monitoring using:

[0070] - the determined wind data and / or

[0071] - ascertained location data indicative of a respective location of the wind monitoring units, and / or

[0072] -ascertained weather data indicative of a wind direction and / or wind speed of wind at a given location; and controlling operation of the wind monitoring units such that the provision of the current target object data is solely performed by the determined subset of wind monitoring units or pairs of wind monitoring units.

[0073] A third aspect of the present invention is formed by a computer program comprising instructions which, when executed by a wind monitoring arrangement, cause the wind monitoring arrangement to perform the method of the second aspect of the invention.

[0074] It shall be understood that the wind monitoring arrangement of claim 1, the method of claim 12, and the computer program of claim 15, have similar and / or identical preferred embodiments, in particular, as defined in the dependent claims.

[0075] It shall be understood that a preferred embodiment of the present invention can also be any combination of the dependent claims or above embodiments with the respective independent claim.

[0076] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.

[0077] BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In the following drawings:

[0079] Fig. 1 shows a schematic block diagram of an exemplary wind monitoring arrangement according to an embodiment of the invention;

[0080] Fig. 2 shows a schematic block diagram of another exemplary wind monitoring arrangement according to another embodiment of the invention; Fig. 3 shows an implementation of a wind monitoring arrangement according to the invention in a street lighting system for RF -based sensing of the position of a tree;

[0081] Fig. 4 shows a schematic diagram of another exemplary wind monitoring arrangement according to yet another embodiment of the invention;

[0082] Fig. 5 shows a flow diagram of an exemplary method for controlling operation of a wind monitoring arrangement in accordance with the invention.

[0083] DETAILED DESCRIPTION OF EMBODIMENTS

[0084] Fig. 1 shows a schematic block diagram of an exemplary wind monitoring arrangement 100 according to an embodiment of the invention. The exemplary wind monitoring arrangement comprises a wind monitoring unit 101. In the example shown in Fig. 1, the monitoring unit comprises a radiofrequency transceiver 106 that includes a transmitter unit 106. 1 that is configured to provide radiofrequency signals 108 and a receiver unit 106.2 that is configured to receive radiofrequency signals 108b. The transmitted radiofrequency signal 108 is received as a received radiofrequency signal 108b by the radiofrequency receiver unit 106.2 of the wind monitoring unit 101. The radiofrequency transceiver 106 is arranged defining a sensing volume 110. Sensing volume refers to that region of space, in which certain changes within said region have a measurable influence on the transmitted radiofrequency signals 108, i.e. cause a measurable and discernible effect that can be detected on the received radiofrequency signal 108b. In particular, the radio frequency transceiver is configured to provide, as the radiofrequency signals, radar signals. The receiver unit 106.2 thus acts as a radar sensor. A target object 112 is located within the sensing volume and therefore, changes in the position of the target object are correlated to changes in the received radiofrequency signal 108b. In general, the target object 112 has a position which depends on a current wind state W in a vicinity of the target object 112. The target object can be for instance, a pole, such as a flag pole, a canopy, an awning, and / or, preferably, a tree, as it will be discussed below. The wind monitoring unit 101 of comprises a signal quality data determination unit 114 that is connected to the radiofrequency transceiver 106 and configured to, using the received radiofrequency signals 108b, determine signal quality data SQ indicative of a current signal quality value of a predetermined signal quality parameter of the received radiofrequency signal 108b, the signal quality parameter being correlatable to the position of the target object 112. In the case of a radar-based sensor, the signal quality parameter can be a frequency or a frequency shift or any other signal quality parameter that can be used for radar-based sensing of the position of the target object. The wind monitoring unit also comprises a data provision unit 116 that is connected to the signal quality data ascertaining unit 114 and configured to ascertain time data TD indicative of a current time, to associate the current time data to the current signal quality value for generating current target object data TOD and to output the current target object data. In the example shown in Fig. 1, the data provision unit is connected to a synchronized GPS time clock unit 117 for ascertaining the time data TD. Preferably, the current target object data TOD provided at different times as indicated by the time data is indicative of a movement M of the target object that is correlated to a presence of wind W in a vicinity of the target object and / or of a direction of movement D of the target object that is correlated to a direction of the wind WD in the vicinity of the target object and / or of a displacement D*, e.g. displacement amount and / or displacement of the target object which is correlated to a strength of the wind WS in the vicinity of the target object.

[0085] The wind monitoring arrangement 100 also comprises a processing unit 118 that is configured to receive, from the data provision unit 116, the current target object data. In the example of Fig.1, the processing unit is an external unit signally connected (via a wired and / or a wireless connection to the one or more wind monitoring units of the wind monitoring arrangement 100. In particular, the processing unit receives current target object data (i.e. time stamped data indicative of the determined signal quality value and thus correlated to the position of the target object at a given time) TOD1 from the wind monitoring unit 101 and one or more current target object data TODn from additional wind monitoring units (not shown). The processing unit is configured to determine using, the current target object data TOD1, ... , TODn and using a wind-data determination algorithm 120, wind data WD indicative of the current wind state in at least the respective sensing volume 110. The processing unit 118 is an external and independent device in Fig. 1. In alternative examples, the processing unit can be integrated in one of the wind monitoring units.

[0086] In particular, the exemplary processing unit 118 of the wind monitoring arrangement 100 of Fig 1 comprises a machine learning engine 122 configured to run a determination algorithm, in particular a clustering algorithm, using as input data the respective current target object data TOD provided by a plurality of wind monitoring units indicative of positions of respective target objects at a respective given time for determining the wind data, in particular wind data indicative of a wind gust.

[0087] Fig. 2 shows a schematic block diagram of another exemplary wind monitoring arrangement 100b according to another embodiment of the invention. In this embodiment. Features of the wind monitoring arrangement 100b of Fig. 2 that have an identical or similar function as those of the wind monitoring arrangement 100 of Fig. 1 will be referred to using the same reference numbers. The wind-monitoring arrangement 100b comprises a plurality of pairs of wind monitoring units 102, 104. Each wind monitoring unit 102, 105 comprises a respective radiofrequency transceiver 106. Alternatively (not shown) a first wind monitoring unit comprises a radiofrequency transmitter for transmitting the radiofrequency signals and a second wind monitoring unit of the pair of wind monitoring units comprises a radiofrequency receiver for receiving the radiofrequency signals. The sensing volume 110 is located in a space between the respective transceivers 106 or between the transmitter and the receiver. At least one of the wind monitoring units 102, 104 of the pair of wind monitoring, or at least the wind monitoring unit comprising the radiofrequency receiver, comprises a signal quality determination unit 114, and data provision unit 116, optionally in connection with a synchronized GPS time clock unit 117 for ascertaining the time data as explained with reference to Fig. 1

[0088] In the case of the wind monitoring arrangement 100b, the signal quality parameter is, preferably, aa received signal strength indicator (RSSI) indicative of the signal strength of the received radiofrequency signal 108b, which depends on the initial signal strength of the radiofrequency signal 108 provided and on the interactions of the radiofrequency signal 108 with the object 112 located in the sensing volume 110. Additionally, or alternatively, the signal quality parameter SQ can be a channel state information (CSI) that is indicative of the channel properties of a communication link established between a transmitter and a receiver or between two transceivers 106. The CSI is indicative of how the radiofrequency signal 108 propagates and may represent a combined effect of scattering, fading, power decay, channel gain, line of sight component, spatial correlation, etc. In general, the signal quality parameter SQ is any parameter whose instant value can be correlated to a position of the target object 112 within the sensing volume 110.

[0089] Fig. 3 shows an implementation of a wind monitoring arrangement according to the invention in a street lighting system for RF -based sensing of the position of a tree 112b. A pair of adjacent street lighting poles 105 include a wind monitoring unit 102, 104 as explained above with reference to Fig. 2 forming a pair of wind monitoring unit. A tree 112b is arranged between the street lighting poles 105 and at least some branches 113a, 113b, 113c of the tree 112b are located within the sensing volume 110 that is defined by the transceivers of the wind monitoring units 102, 104. Here, the current target object data TOD1 is indicative of a movement M of one or more branches 113a, 113b, 113c of the tree 112b and / or of a direction of movement D of the one or more branches 113a, 113b, 113c of the tree that is correlated to a direction of the wind WD in the vicinity of the tree and / or of a displacement amount and / or rate D* of the one or more branches 113a, 113b, 113c of the tree which is correlated to a strength of the wind WS in the vicinity of the tree 112b.

[0090] Fig. 4 shows a schematic diagram of another exemplary wind monitoring arrangement 100b according to yet another embodiment of the invention. Fig. 4 depicts a city area or neighborhood exemplarily formed by 4 blocks and two perpendicular streets A and B. Street A has a West-East orientation and street B has North-South orientation. On the sidewalks there are a plurality of pairs of street lights, each symbolized by a circle, for instance as part of an InterAct City lighting arrangement, each of which comprises an outdoor light controller OLC (some may also comprise an outdoor sensor bundle (OSB) comprising a radar sensor. The operation of the street lights is typically controlled by a controller unit 119 that is wirelessly connected to the street lights for exchanging status information and control instructions. The OLC are configured to perform a wind monitoring function as explained above. For instance, the wind monitor units can be implemented in the OLC and / or the OSB of the street lights. The wind monitoring unit can be also arranged on a traffic light, traffic sign or any other suitable structure, including a dedicated pole. The wind monitoring units are associated in pairs Pl to P2, each pair forming a sensing volume that mainly includes a tree T (only one is labelled in Fig. 4 for the sake of simplicity). Each of the pairs P1-P8 of generates and provides respective current target object data (TOD1-TDO8) as explained above to the processing unit, which can be integrated in the controller unit of the lighting arrangement.

[0091] Advantageously, the processing unit 118 comprises a selection unit 124 that is connected to the processing unit 118 and configured to, using the determined wind data WD, to determine a subset of wind monitoring units or pairs of wind monitoring units for performing the wind monitoring and to control operation of the wind monitoring units such that the determination and / or the provision of the current target object data is solely performed by the determined subset of wind monitoring units. For instance, if the wind data WD is indicative of a wind gust flowing from East to West, i.e., along street A, the selection unit 124 is configured to select a subset of relevant pairs Pl, P2, P3 and P4 for continuing the determination of the current target object data TOD1-TOD4 and stopping the provision of the current target object data TOD5-TOD8 provided by the pairs P5-P8, which are substantially aligned in a North-South direction along street B. The evaluation of the timestamped target object data, for example from the relevant pairs of wind monitoring units, provides information about the speed and the strength of the wind gust.

[0092] The wind monitoring arrangement determines from the current target object data provided by street light pairs P1-P8 in the area, the arrival timestamps of the most recent wind gusts. To subsequently limit cellular communication cost, the selection unit is advantageously configured to dynamically select -based on the initial wind data from many streetlight pairs- the most suitable subset of streetlights to perform the continuous gust monitoring. This selection is based on the wind direction in this specific part of the city as observed from the initial RF sensing gust monitoring data set, i.e., the current target object data TD1-TD8.

[0093] For instance, four street lights pairs P1-P4 (e.g. InterAct City street lights) are selected which are lined up in direct line of sight of each other and whereas these four streetlights are aligned with the current direction the wind W (as indicated by the arrow flowing along street A). In other words, these four street lights pairs on the first street A are selected to continuously report their determined current target object data indicative of the current target object data that is correlated to the movement of the trees, as they are downstream of each other with respect to the direction of the wind W, which is in the E-W direction. The remaining pairs P5-P8 on the second street B where the wind W is blowing orthogonal to the orientation of the roadway, are not selected.

[0094] If the wind data is indicative of a current wind direction that is not blowing parallel to any of the streets in the neighborhood, streetlights which are located on multiple streets, for instance at an intersection may be selected for the RF-based wind sensing task.

[0095] Preferably the selection unit 124 is further configured to ascertain location data LD indicative of a respective location of the wind monitoring units P1-P8. The determination of the subset of pairs of wind monitoring unit is further based on the location data. The location data can be determined using a GPS unit and may also include information pertaining to the surroundings of the wind monitoring unit, for instance pertaining to the surrounding terrain or buildings. For instance, the wind state at a luminaire location also depends on the terrain and surrounding building effects (e.g. wind shadows). Hence, when selecting the subset of wind monitoring units for the wind gust detection, the location of the respective wind monitoring units with respect to the built structure and the current wind direction are also taken into account. Similarly, we the selection of the subset of wind monitoring units for the wind gust detection task can further be adjusted based on the expected wind shadows from the buildings and trees given the current prevalent wind direction.

[0096] Preferably, in an exemplary wind monitoring arrangement, such as the one depicted in Fig. 4, the selection unit is further configured to ascertain current weather data 126 indicative of a wind direction and / or wind speed of wind at a given location, and to further use the received weather data for the determination of the subset of pairs of wind monitoring units. The weather data 126 can be received from a local weather station 125, such as a weather station located at an airport, where the wind data has a high degree of accuracy.

[0097] Generally, while GPS provides a hyper accurate time clock, the GPS-based location-fix of the wind monitoring unit pole is typically insufficiently accurate to determine the wind speed solely based on the respective GPS time clocks and respective GPS- determined locations of two street lighting poles alone. In addition, the wave front of a wind gust is not progressing as a straight line through the city (the wave front of the time-varying wave field is defined as the set of all points having the same phase). Thus, it is preferred that the wind data is indicative of both a wind speed and a wind direction of the current wind state.

[0098] Thus, Fig. 4 exemplary depicts a wind monitoring arrangement where the OLCs of the street lights are used for passive RF sensing (with either the cellular radio and / or Zigbee radio) to infer from the movement of tree branches that a wind gust has arrived at a specific lighting pole in the area. The RF sensing process based on the determination of the signal quality value is also suitable for analyzing the movement direction of the tree branches as indicator for the directionality of the wind gust (e.g. the observed branch movements may indicate an east-west wind gust). Preferably a clustering algorithm is applied to identify which tree branch movements recorded by different streetlights are associated with the same specific wind gust progressing through the neighborhood. The clustering algorithm utilizes input features derived from the RF sensing data (i.e., the current target object data) such as the arrival time of the gusts observed at each pole, the directionality of the tree branch displacement (indicative of the gust’s direction) and the gust’s relative strength based on the observed displacement of the tree branch. After clustering the observed tree-branch movements into individual wind gusts, a machine learning model infers, based upon the GPS-timestamped RF sensing data (which is concurrently recorded by multitude streetlights in the neighborhood), the wind speed, duration and wind direction of this specific gust. Fig. 5 shows a flow diagram of an exemplary method 500 for controlling operation of a wind monitoring arrangement in accordance with the invention.

[0099] The method comprises, in a step 502, providing, by a radiofrequency transmitter or by a radio frequency transceiver, radiofrequency signals, in particular wireless communication signals in accordance with a wireless communication protocol and / or radar signals. The method also comprises, in a step 504, receiving, at a radiofrequency receiver or at a radiofrequency transceiver, the provided radiofrequency signal, wherein the radiofrequency transmitter and the radiofrequency receiver are arranged defining a sensing volume, in which a respective target object is located, the target object having a position which depends on a current wind state. The received radiofrequency signals are the provided radiofrequency signals after interaction with the environment, in particular with the sensing volume, more in particular with the target object. The method further include, in a step 506, and using the received radiofrequency signals, determining signal quality data indicative of a current signal quality value of a predetermined signal quality parameter of the received radiofrequency signal, the signal quality parameter being correlatable to a position of the target object. The method further includes, in a step 508, ascertaining time data indicative of a current time, in particular by using a synchronized GPS time clock unit. Further, in a step 510, the method includes associating the current time data to the current signal quality to generate current target object data, and, in a step 512, outputting or providing the current target object data, in particular via a wired or wireless communication channel to a processing unit. The method then comprises, in a step 514, determining, using the current target object data and using a wind-data determination algorithm, wind data indicative of the current wind state in at least the sensing volume.

[0100] Optionally, as indicated by the dashes boxes in Fig. 5, the method 500 comprises, in a step 516, determining a subset of wind monitoring units for performing the determination of current target object data using

[0101] - the previously determined wind data and / or

[0102] - ascertained location data indicative of a respective location of the wind monitoring units, and / or

[0103] -ascertained weather data indicative of a wind direction and / or wind speed of wind at a given location; and, in a step 518, controlling operation of the wind monitoring units such that the determination and / or the provision of the current target object data is solely performed by the determined subset of wind monitoring units. In summary, the invention is directed to a wind monitoring arrangement, comprising one or more wind monitoring units that comprise a RF -transceiver configured to provide and receive RF-signals and defining a sensing volume in which a respective target object is located, which has a position depending on a wind state. A signal quality data determination unit is configured to determine signal quality data indicative of a value of a signal quality parameter that is correlatable to the position of the target object. Current target object data is generated by associating current time data to the signal quality value. A processing unit is configured to determine, using the current target object data and a winddata determination algorithm, wind data indicative of the current wind state in at least the sensing volume, thereby reducing the hardware requirements for wind monitoring.

[0104] Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims.

[0105] In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality.

[0106] A single unit or device may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

[0107] A computer program may be stored / distributed on a suitable medium, such as an optical storage medium or a solid-state medium, supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems.

[0108] Any reference signs in the claims should not be construed as limiting the scope.

Claims

CLAIMS:

1. Wind monitoring arrangement (100), comprising onewind monitoring unit(101), wherein the wind monitoring unit comprises:- a radiofrequency transceiver (106) configured to provide and receive radiofrequency signals (108, 108b), wherein the radio frequency transceiver (106) is configured to provide, as the radiofrequency signals, radar signals and wherein the radiofrequency transceiver is arranged for defining a sensing volume (110), in which a respective target object (112) is located, the target object having a position which depends on a current wind state (W) in a vicinity of the target object,- a signal quality data determination unit (114) connected to the radiofrequency transceiver and configured to, using the received radiofrequency signals, determine signal quality data (SQ) indicative of a current signal quality value of a predetermined signal quality parameter of the received radiofrequency signal (108b), the signal quality parameter being correlatable to the position of the target object, whereby the signal quality parameter is a frequency or a frequency shift;- a data provision unit (116), connected to the signal quality data ascertaining unit and configured to ascertain time data (TD) indicative of a current time, to associate the current time data to the current signal quality value for generating current target object data (TOD) and to output the current target object data, wherein the target object has a first section that has a fixed position with respect to the monitoring unit and a second section whose position with respect to the monitoring unit is not fixed and varies in dependence on a current wind state; the wind monitoring arrangement further comprising:- a processing unit (118) configured to receive, from the data provision unit, the respective current target object data (TOD1, TODn) and to determine using the current target object data and using a wind-data determination algorithm (120), wind data (WD) indicative of the current wind state in at least the sensing volume, wherein the current target object data (TOD) is indicative of a movement (M) of the target object that is correlated to a presence of wind (W) in a vicinity of the targetobject and / or of a direction of movement (D) of the target object that is correlated to a direction of the wind (WD) in the vicinity of the target object and / or of a displacement (D*) of the target object which is correlated to a strength of the wind (WS) in the vicinity of the target object.

2. Wind monitoring arrangement (100), comprising one or more pairs of wind monitoring units (102, 104), wherein the radiofrequency transceivers of the wind monitoring units of a pair of wind monitoring units are arranged defining a sensing volume (110) in which a respective target object (112) is located, the target object having a position which depends on a current wind state (W) in a vicinity of the target object, wherein each of the wind monitoring units comprises:- a radiofrequency transceiver (106) configured to provide and receive radiofrequency signals (108), wherein the radio frequency transceiver (106) is configured to provide, as the radiofrequency signals, wireless communication signals in accordance with a predetermined wireless communication protocol;- a signal quality data determination unit (114) connected to the radiofrequency transceiver and configured to, using the received radiofrequency signals, determine signal quality data (SQ) indicative of a current signal quality value of a predetermined signal quality parameter of the received radiofrequency signal (108b), the signal quality parameter being correlatable to the position of the target object;- a data provision unit (116), connected to the signal quality data ascertaining unit and configured to ascertain time data (TD) indicative of a current time, to associate the current time data to the current signal quality value for generating current target object data (TOD) and to output the current target object data, wherein the target object has a first section that has a fixed position with respect to the monitoring unit and a second section whose position with respect to the monitoring unit is not fixed and varies in dependence on a current wind state; the wind monitoring arrangement further comprising:- a processing unit (118) configured to receive, from the data provision unit, the respective current target object data (TOD1, TODn) and to determine using the current target object data and using a wind-data determination algorithm (120), wind data (WD) indicative of the current wind state in at least the sensing volume, wherein the current target object data (TOD) is indicative of a movement (M) of the target object that is correlated to a presence of wind (W) in a vicinity of the targetobject and / or of a direction of movement (D) of the target object that is correlated to a direction of the wind (WD) in the vicinity of the target object and / or of a displacement (D*) of the target object which is correlated to a strength of the wind (WS) in the vicinity of the target object.

3. The wind monitoring arrangement, of any of the preceding claims, wherein the target object (112) is a tree (112b) and wherein the current target object data (TOD) is indicative of a movement (M) of one or more branches (113a, 113b, 113c) of the tree (112b) and / or of a direction of movement (D) of the one or more branches (113a, 113b, 113c) of the tree that is correlated to a direction of the wind (WD) in the vicinity of the tree and / or of a displacement (D*) of the one or more branches (113a, 113b, 113c) of the tree which is correlated to a strength of the wind (WS) in the vicinity of the tree.

4. The wind monitoring arrangement of any of the preceding claims, wherein data provision unit comprises a synchronized GPS time clock unit (117) for ascertaining the time data.

5. The wind monitoring arrangement of any of the preceding claims, wherein the processing unit comprises a machine learning engine (122) configured to run a determination algorithm, in particular a clustering algorithm, using as input data the respective current target object data provided by a plurality of pairs of wind monitoring units indicative of positions of respective target objects at a respective given time for determining the wind data, in particular wind data indicative of a wind gust.

6. The wind monitoring arrangement of any of the preceding claims, comprising a plurality of wind monitoring units or pairs of wind monitoring units(Pl, P2, P3, P4, P5, P6, P7, P8), the wind monitoring arrangement further comprising a selection unit (124) that is connected to the processing unit (118) and configured to, using the determined wind data (WD), to determine a subset of wind monitoring units of pairs of wind monitoring units (Pl, P2, P3, P4) for performing the determination of current target object data and to control operation of the wind monitoring units such that the provision of the current target object data is solely performed by the determined subset of wind monitoring units or pairs of wind monitoring units (Pl, P2, P3, P4).

7. The wind monitoring arrangement of claim 6, wherein the selection unit (124) is further configured to ascertain location data (LD) indicative of a respective location of the wind monitoring units, and wherein the determination of the subset of wind monitoring units is further based on the location data.

8. The wind monitoring arrangement of claim 6 or claim 7, wherein the selection unit is further configured to ascertain current weather data (126) indicative of a wind direction and / or wind speed of wind at a given location, and to further use the received weather data for the determination of the subset of wind monitoring units.

9. The wind monitoring arrangement of any of the preceding claims, wherein the wind monitoring units are arranged on street luminaires, in particular, wirelessly controllable street luminaires of an intelligent lighting system.

10. Method (500) for controlling operation of a wind monitoring arrangement, the method comprising:- providing (502), by a radiofrequency transmitter, radiofrequency signals,- receiving (504) at a radiofrequency receiver, the provided radiofrequency signal, wherein the radiofrequency transmitter and the radiofrequency receiver are arranged defining a sensing volume, in which a respective target object is located, the target object having a position which depends on a current wind state, wherein the target object has a first section that has a fixed position with respect to the monitoring unit and a second section whose position with respect to the monitoring unit is not fixed and varies in dependence on a current wind state,- using the received radiofrequency signals, determining (506) signal quality data indicative of a current signal quality value of a predetermined signal quality parameter of the received radiofrequency signal, the signal quality parameter being correlatable to a position of the target object;- ascertaining (508) time data indicative of a current time- associating (510) the current time data to the current signal quality to generate current target object data;- outputting (512) the current target object data,- determining (514) using the current target object data and using a wind-data determination algorithm, wind data indicative of the current wind state in at least the sensingvolume, wherein the current target object data (TOD) is indicative of a movement (M) of the target object that is correlated to a presence of wind (W) in a vicinity of the target object and / or of a direction of movement (D) of the target object that is correlated to a direction of the wind (WD) in the vicinity of the target object and / or of a displacement (D*) of the target object which is correlated to a strength of the wind (WS) in the vicinity of the target object.

11. The method of claim 10, further comprising providing, as the radiofrequency signals, wireless communication signals in accordance with a predetermined wireless communication protocol; and / or providing, as the radiofrequency signals, radar signals; and / or ascertaining the time data using a synchronized GPS time clock.

12. The method of claim 10 or 11, further comprising determining a subset of wind monitoring units for performing the provision of current target object data using:- the previously determined wind data and / or- ascertained location data indicative of a respective location of the wind monitoring units, and / or-ascertained weather data indicative of a wind direction and / or wind speed of wind at a given location and controlling operation of the wind monitoring units such that the provision of the current target object data wind monitoring is solely performed by the determined subset of wind monitoring units.

13. Computer program comprising instructions which, when executed by a wind monitoring arrangement according to claim 1 or claim 2, cause the wind monitoring arrangement to perform the method of any of the claims 10 to 12.

Images