Optical analysis system for determining weather data

Abstract

The invention relates to an optical analysis system (100) for determining weather data from the atmosphere, comprising a plurality of optical measuring devices (110), a central processing device (140) and an optical output device (160). The plurality of optical measuring devices are designed to analyze at least partially different spatial measurement regions (112), wherein the optical measuring devices each have at least one sensor unit (115), a control unit (120) and a communication unit (123). The control unit is designed to receive sensor data (119) from the sensor unit, to apply pattern recognition based on predetermined training patterns for noise suppression to the sensor data, and to provide correspondingly processed sensor data (122). The communication unit is designed to output the processed sensor data as a processed sensor signal (125). The central processing device is designed to receive the associated processed sensor signal of all optical measuring devices from the plurality of optical measuring devices for the at least partially different spatial measurement regions and, on the basis of a cross-validation of the corresponding processed sensor data, to apply pattern recognition (144) based on predetermined overall training patterns (142) for further noise suppression to the received processed sensor signals and thereby to determine and output current atmospheric data (146).

Description

[0001] Optical analysis system for determining weather data

[0002] Description

[0003] The invention relates to an optical analysis system for determining weather data from the atmosphere. According to a further aspect, the invention relates to a method for determining weather data from the atmosphere.

[0004] In principle, sensor units and optical measuring devices are known that detect optical properties of the atmosphere via the reflection of emitted light. A well-known example of this is LIDAR systems, which essentially emit monochrome laser beams perpendicularly into the atmosphere to measure winds and chemical gas compositions in upper atmospheric layers.

[0005] Also known is the combination of optical measuring devices to enable a particularly large measurement range for determining weather data from the atmosphere through their spatial distribution. This allows the measured sensor signals to be compared and noise components in the respective sensor signals to be filtered out. Such data can be used, for example, for weather forecasting and / or for simulating wind dynamics in the analyzed atmospheric layers.

[0006] The object of the present invention is to provide an improved optical analysis system, in particular a particularly reliable and especially insensitive optical analysis system.

[0007] According to the invention, an optical analysis system for determining weather data from the atmosphere is proposed to solve this problem, comprising a plurality of optical measuring devices for analyzing at least partially different spatial measurement areas, wherein the optical measuring devices each have

[0008] - at least one sensor unit designed to receive an optical signal and provide corresponding sensor data,

[0009] - a control unit arranged and configured to receive the sensor data, apply pattern recognition based on predetermined training patterns to the sensor data for noise reduction, and provide appropriately processed sensor data; - a communication unit configured to output the processed sensor data as a processed sensor signal;

[0010] - a central processing device that is arranged and designed to receive the respective processed sensor signal of all optical measuring devices from the majority of optical measuring devices for the at least partially different spatial measurement ranges and, based on a cross-validation of the corresponding processed sensor data, to apply a pattern recognition based on predetermined overall training patterns to the received processed sensor signals for further noise suppression and thereby to determine and output current atmospheric data,

[0011] - an optical output device that is arranged and configured to receive the current atmospheric data and provide it as an optical output via a user interface.

[0012] Within the scope of the invention, it was recognized that while existing optical analysis systems already employ various noise reduction methods, these methods often encounter problems if individual sensor units fail and / or provide corrupted sensor data. Against this background, the optical analysis system according to the invention proposes providing an additional level of sensor data aggregation via the central processing unit and obtaining particularly reliable current atmospheric data through cross-validation of sensor data already processed in the measuring instruments. This allows for the reliable determination of current atmospheric data even in the presence of interference. Such interference can include, for example, malfunctions of individual sensor units, environmental influences such as rain, clouds, and the like, or obstructions such as leaves, birds, and the like.Against this background, the optical analysis system according to the invention allows a particularly reliable determination of current atmospheric data even during the occurrence of such disturbances.

[0013] Advantageously, in the optical analysis system according to the invention, the control units of the measuring instruments and the central processing unit form at least two levels of noise reduction and merging of received data. Such structured data acquisition also enables a particularly advantageous learning process for learning from past data.

[0014] The predetermined training patterns and / or the predetermined overall training patterns can be determined before operation of the optical analysis system, for example empirically, and / or determined during operation of the optical analysis system from past sensor data.

[0015] The central processing unit according to the invention is preferably a separate device from the measuring instruments, which is connected to the majority of measuring instruments either via cable or wirelessly. Alternatively, the central processing unit can also be located in a common housing with a measuring instrument or similarly associated with a measuring instrument. The devices of the analysis system and the units of a respective measuring instrument can be arranged at least partially adjacent to one another, with these components differing from each other at least at a software level.

[0016] In addition to the optical signal, the sensor unit can also receive other signals and provide them as sensor data. For example, temperature, air pressure, humidity, and / or similar parameters can be determined. The optical signal allows for the acquisition of sensor data relating to a spatially distant measurement area. For instance, in so-called LiDAR systems, a laser beam backscattered by the atmosphere can provide reliable atmospheric measurement data, which can then be processed by the respective control unit and the central processing device for a final determination of the current atmospheric data.

[0017] The measuring range of a given measuring device is roughly determined by the spatial arrangement of the device and the orientation of its corresponding sensor unit (at least one). In principle, different sensor units can have different measuring ranges, which the corresponding control unit takes into account during pattern recognition for noise suppression.

[0018] The sensor unit can be used to detect one or more measured quantities. In this sense, a single sensor unit can detect multiple measured quantities, or different sensor units can be used within an optical measuring device to detect different measured quantities. Alternatively, the optical measuring device can detect only one measured quantity.

[0019] Cross-validation, as defined in the invention, is an analysis of sensor data from different sensors for similar measurement ranges. This cross-validation can, for example, correlate sensor data with respect to their measured amplitudes and signal-to-noise ratio, thereby aiding in the acquisition of higher-level characteristics of atmospheric data. Various analytical methods are known for such cross-validation of data, such as the use of a suitable correlation function, so the analytical details will not be discussed in detail here. The central processing device can be a computer, a smartphone, a tablet, and / or the like. In particular, the central processing device can preferably access additional computing resources via a suitable network connection.

[0020] The optical output according to the invention can be located at a distance from the other devices of the optical analysis system. The central processing unit can preferably communicate wirelessly with the optical output device, for example via a network connection. Alternatively or additionally, optical output can be provided directly at the central processing unit.

[0021] Atmospheric data imply, at least in the sense that atmospheric characteristics allow direct inferences about the weather. Therefore, weather data and atmospheric data can be understood as synonyms in the following.

[0022] Preferred embodiments of the optical analysis system according to the invention are described below.

[0023] In a particularly preferred embodiment, the central processing unit and / or the control unit of at least one optical measuring device is configured to perform defragmentation of received time-series-based data. In this embodiment, a significant amount of data can be advantageously reduced, thereby accelerating the processing of this data and / or avoiding redundant data. Furthermore, the defragmentation can enable rapid access to particularly relevant portions of the time-series-based data. Preferably, all data acquired via the optical analysis system are time-series-based, since each measured quantity is characterized by a corresponding measurement point in time at which it was acquired.In a further particularly preferred embodiment, at least one optical measuring device is configured to determine two different measured variables via two different sensor units for providing the corresponding sensor data for a spatial measurement range. Different measured variables allow dependencies between these variables to be used for evaluating the existing correlation. For example, interfering influences can be detected particularly quickly and suppressed for further processing of the sensor data. Examples of such different measured variables are wind speed, temperature, spectral color, humidity, and / or the like. Such sensors are also known in the context of atmospheric measurement, so possible structures for such sensors will not be described in detail below.

[0024] In an advantageous embodiment of the preceding design, the control unit of a corresponding measuring device is configured to evaluate the sensor data from various sensor units of the measuring device in a correlated manner for pattern recognition. In this embodiment, the physical dependencies between different measured variables are advantageously used to enable a reliable, automated analysis of the current state of the atmosphere. For example, based on temperature and wind dynamics, a short-term temperature trend can be indicated, which allows for the automated evaluation of future data in this regard, thereby reliably suppressing inherent signal noise in the analyzed data.

[0025] In a further advantageous embodiment, at least one optical measuring device comprises a disjoint auxiliary sensor unit, which is arranged and configured to have a spatial measuring range that differs at least partially from that of the at least one sensor unit of the optical measuring device. Such an extension of the analyzed measuring range can advantageously enable the evaluation of dependent measured variables and thus lead to a particularly reliable evaluation of the current state of the atmosphere. Such a disjoint auxiliary sensor unit can, for example, be formed by a terrestrial ultrasonic sensor.

[0026] In an advantageous embodiment of the preceding design, the control unit of the corresponding optical measuring device is further developed to perform pattern recognition based on additional sensor data from the auxiliary sensor unit. In this embodiment, the improved recognition performance of the corresponding measuring device is utilized to enable particularly reliable detection of patterns contained in the signal and, consequently, particularly reliable noise suppression. For example, the additional sensor data can also be used for further pattern recognition only in the case of suspected interference. In this example, a particularly reliable verification of acquired data can be performed automatically by physical correlation with the additional sensor data.

[0027] In a further particularly preferred embodiment, at least one control unit of one of the optical measuring devices is configured to determine weight sets for the weighted evaluation of the sensor data for pattern recognition via a neural network based on past sensor data. In this embodiment, the control unit of at least one optical measuring device can advantageously learn from previously acquired sensor data in order to determine the weight sets, for example in the form of weighting factors, based on learned patterns and the signal-to-noise ratios indicated therein, and / or the like, and to apply them for future pattern recognition. In this embodiment, the analysis of the optical analysis system according to the invention can therefore be particularly advantageously improved over time by means of automatically adapting weight sets. Possible structures of such neural networks are generally known.Here, changes are made to weighting factors based on training data and / or previously acquired sensor data in order to learn from the corresponding results which adjustments to weighting factors improve the outcome and subsequently adjust the weighting factors accordingly. An improvement in the result can be achieved, for example, through a manual feedback process during a training phase of the control unit.

[0028] In a particularly advantageous embodiment of the preceding design, at least two optical measuring devices from the plurality of optical measuring devices are configured to exchange specific weight sets and / or data relating to a completed pattern recognition via their respective communication units. In this embodiment, the inventive structure of the optical analysis system can be used particularly advantageously to enable automated learning between the measuring devices. It can thus be expected that the weight sets for similar measuring devices in a similar environment should hardly differ or at least exhibit similar dependencies of various parameters on the expected patterns and / or the expected noise.

[0029] In a further embodiment, the central processing unit is further designed to receive sensor data from at least one sensor unit in addition to the received processed sensor signals, and to determine the current atmospheric data based on this sensor data. Using sensor data directly from the corresponding sensor unit can lead to a further improvement in noise suppression when combining processed sensor data from different measurement ranges. In particular, sensor data for an intersection of different measurement ranges can simplify the combination of the corresponding processed sensor data for these adjacent measurement ranges. Alternatively or additionally, the direct use of sensor data from a sensor unit can enable a plausibility check to automatically avoid systematic errors in pattern recognition.

[0030] In a particularly preferred embodiment, the central processing unit is further configured to use a neural network to weight the received, processed sensor data based on the analyzed spatial measurement areas in order to determine the current atmospheric data. This weighting preferably complements the determination of weight sets within the control unit of each measuring device. The central processing unit can also be configured to receive the weight sets of a respective control unit and use them for pattern recognition. For example, the temporal evolution of the control unit's weight sets can enable particularly fast learning of the neural network in the central processing unit.

[0031] In another embodiment, the central processing unit is configured to perform and output an estimate of future atmospheric data based on received and / or processed sensor data. In this embodiment, pattern recognition is advantageously used to infer a future atmospheric state in the analyzed measurement areas based on the current state of the atmosphere within those measurement areas. For example, the physical relationship between measured quantities can be exploited to estimate their development. Thus, an increase in wind speed and a rise in temperature can indicate the probability of an imminent storm.

[0032] In a further preferred embodiment, the optical output device is configured to graphically display the current atmospheric data through a dynamically changing representation of at least two spatial dimensions. In this embodiment, the determined atmospheric data are advantageously reduced to optical representations that are easily recognizable by a user of the optical analysis system. Information relating to three spatial dimensions, for example as a corresponding diagram, can also be displayed within the two spatial dimensions. The display is preferably dynamic in that it adapts itself over the duration of the output according to the changing atmospheric data. For example, the development of wind speeds in different atmospheric layers can be displayed essentially in real time.

[0033] In a further advantageous embodiment, the optical output device of the analysis system is configured to display the current atmospheric data for a predetermined geographical area using a graphic symbol from a predetermined plurality of graphic symbols. The use of predetermined graphic symbols makes visualizing the atmospheric data results particularly simple and intuitive. For example, a cloud can be used for a cloudy measurement area, an arrow for wind direction, a raindrop for a rainy measurement area, and / or the like. Preferably, the appropriate graphic symbol can be automatically determined based on the received atmospheric data and displayed on a corresponding map.User-specific settings, such as the resolution of the graphic display, the placement of graphic symbols, color contrast, scaling, transparency, a color scheme, and / or the like, can be selected, particularly preferably based on such user-specific settings. A corresponding setting can then be automatically determined at later times, preferably in a context-specific manner. Such adaptive support for data visualization can be provided by known software applications, especially those based on neural networks.

[0034] In a further advantageous embodiment, at least one sensor unit is formed by a LiDAR sensor unit. Such LiDAR sensor units for atmospheric analysis are generally known and can therefore be advantageously integrated into the optical analysis system according to the invention with minimal development effort. In particular, several identical, known LiDAR sensor units can be used for a sensor field as a measuring device within the meaning of the invention or at least as a sensor unit of such a measuring device. A reliable determination of atmospheric data can be carried out at different altitudes and thus in different layers of the atmosphere using such a LiDAR sensor unit. Such a LiDAR sensor unit can, for example, have seven measurement channels: LiDAR measurement in three different spatial directions and color spectral measurements in four different directions.

[0035] In principle, it follows from the above that the sensor units can also differ from LIDAR sensor units, and that in addition to optical sensors, other sensors can also be provided on the measuring device according to the invention. In particular, an ultrasonic sensor, a temperature sensor, an air pressure sensor, and / or the like can be provided to supply the sensor data. Against this background, the optical analysis system according to the invention is not limited to a specific type of sensor and / or to a specific quantity to be measured within the surrounding atmosphere.

[0036] According to a further aspect of the invention, a method for determining weather data from the atmosphere is proposed to solve the aforementioned problem. The method comprises the following steps:

[0037] - Dividing an atmospheric area into a plurality of spatial measurement areas;

[0038] - Providing sensor data from the respective spatial measurement areas to a control unit assigned to the spatial measurement area from a plurality of control units;

[0039] - Applying pattern recognition based on predetermined training patterns to the provided sensor data for noise reduction and providing corresponding processed sensor data; - Outputting the processed sensor data assigned to the respective spatial measurement area as a processed sensor signal;

[0040] - Receiving the processed sensor signals by a central processing device;

[0041] - Determining and outputting current atmospheric data by applying pattern recognition based on predetermined overall training patterns for further noise reduction to the received processed sensor signals; and

[0042] - Providing a visual output of current atmospheric data.

[0043] The method according to a further aspect of the invention uses the optical analysis system according to the invention and thus also exhibits the corresponding advantages of the embodiments of the optical analysis system. In particular, the method according to the invention enables a particularly reliable and robust determination of atmospheric data against interference. Noise suppression is carried out in several layers, including by means of appropriate pattern recognition, whereby the measurement ranges in which the corresponding sensor data were acquired and further data processing steps for the final determination of the atmospheric data can be taken into account.

[0044] In principle, it follows from the above that the sensor units can preferably be at least partially formed by LIDAR sensor units, but that other sensor units can also be used for the optical analysis system. The process steps are preferably carried out at least partially on a computer, a smartphone, a tablet, or another hardware component. The process steps of the method according to the invention are preferably carried out in the specified sequence. In particular, sensor data is provided first, then the processed sensor data is provided by the decentralized control units in the respective measuring areas, and finally the atmospheric data is determined in the central processing device. Thus, the method according to the invention provides a multi-layered and therefore particularly robust and reliable determination of the atmospheric data.

[0045] It follows from the above that the inventive method can be supplemented by additional steps corresponding to the analogous embodiments of the optical analysis system described above. For example, defragmentation of received and / or processed sensor data can be advantageously performed.

[0046] The invention will now be explained in more detail with reference to advantageous embodiments schematically illustrated in the figures. These show, in detail:

[0047] Fig. 1 shows a schematic representation of a first embodiment of an optical analysis system according to the invention;

[0048] Fig. 2 shows a schematic representation of a second embodiment of the optical analysis system according to the invention;

[0049] Fig. 3 shows an optical output of a third embodiment of the optical analysis system according to the invention; and Fig. 4 shows a flowchart of an embodiment of a method according to a further aspect of the invention.

[0050] Fig. 1 shows a schematic representation of a first embodiment of an optical analysis system 100 according to the invention. The optical analysis system 100 is designed for determining weather data from the atmosphere. For this purpose, it comprises a plurality of optical measuring devices 110, a central processing unit 140, and an optical output unit 160.

[0051] The majority of optical measuring devices 110 are designed to analyze at least partially distinct spatial measurement areas 112, wherein these measurement areas 112 are located above the respective measuring device. Preferably, the measuring devices are arranged at a predetermined location in the ground area and detect backscatter of laser light emitted into the atmosphere, similar to a LiDAR system. A sensor unit 115 of the optical measuring device 110 forms a corresponding LiDAR module with a laser and an optical detector for receiving an optical signal 117 and providing corresponding sensor data 119. The sensor data 119 reach a respective control unit 120 of the optical measuring device 110 via cable.The control unit 120 is configured to apply pattern recognition based on predetermined training patterns to the sensor data 119 for noise reduction and to provide the correspondingly processed sensor data 122. The training patterns exhibit, among other things, known movement characteristics of air currents and temperature developments. Furthermore, correlations between different sensor data 119 from the sensor unit 115 are detected, and the intrinsic noise component within the sensor data 119 is reduced based on the correlated data. The processed sensor data 122 are output as a processed sensor signal 125 via a communication unit 123 of the respective optical measuring device 110. The output of the processed sensor signal 125 is wireless via a radio connection to the central processing unit 140.The radio connection is preferably a long-wave radio connection for a particularly long range. In an alternative or supplementary embodiment, the respective measuring device is connected to the central processing unit 140 via a cable. The central processing unit 140 is configured to receive the processed sensor signals 125 from all optical measuring devices 110 of the optical analysis system 100. These processed sensor signals 125 can be received sequentially or in parallel. As a result, the central processing unit 140 has access to the respective processed sensor data 122 for at least partially different spatial measurement ranges 112.The central processing unit 140 is designed to apply pattern recognition 144, based on predetermined overall training patterns 142, to the received processed sensor signals 125 for further noise reduction, based on cross-validation of the corresponding processed sensor data 122, and thereby to determine and output current atmospheric data 146. The cross-validation is performed via a common pattern recognition of processed sensor data 122 from adjacent measurement areas 112. For example, temperature trends, the formation of winds, rain fronts, and / or the like can be correlated and evaluated. This allows for particularly reliable noise reduction within the received data. Thus, data evaluation based on predetermined training data can be performed in at least two stages within the optical analysis system 100.First, the local sensor data 119 of the respective measuring device 110 are evaluated, and then the correspondingly processed sensor data 122 are evaluated centrally in the central processing device 140. This allows errors in a sensor unit 115 and even the failure of a complete measuring device 110 to be automatically compensated for by the corresponding data from the other measuring devices 110 and / or sensor units 115.

[0052] To simplify and / or accelerate the evaluation of the sensor data 119 and / or the processed sensor data 122, the respective control unit 120 and / or the central processing device 140 is designed to perform a defragmentation of received time-series-based data. In this process, data is reordered and data that is no longer needed is deleted to enable more efficient processing of this data.

[0053] The optical output device 160 is arranged and configured to receive the current atmospheric data 146 and to provide it as an optical output 165 via a user interface 162. In this case, the user interface 162 is formed by a touch display, which indicates the direction of movement of air masses by means of corresponding arrows as an optical output 165. In an alternative embodiment, the user interface comprises a display and additionally a rotary knob, a switch, buttons, and / or the like.

[0054] Figure 1 shows five optical measuring devices 110 by way of example. In alternative embodiments, more than five optical measuring devices 110, such as more than ten optical measuring devices 110, are provided for the optical analysis system 100 according to the invention. The optical measuring devices 110 of the optical analysis system 100 are all identical. In an embodiment not shown, different optical measuring units with different sensor units, in particular with different numbers of sensor units, are provided.

[0055] The respective optical measuring device 110 can have various sensor units 115. At least one measuring channel is formed by an optical sensor that detects light from the atmosphere and can thereby collect data from higher atmospheric layers. Different sensor units 115 allow different measured quantities to be determined for the investigated spatial measurement range 112. This makes it possible to utilize physical relationships between such measured quantities, such as temperature and wind direction, in order to determine atmospheric data 146 automatically and with high reliability.

[0056] Fig. 2 shows a schematic representation of a second embodiment of the optical analysis system 200 according to the invention. The optical analysis system 200 is designed similarly to the optical analysis system 100 from Fig. 1. Unlike the measuring devices 110 from Fig. 1, the four optical measuring devices 210 shown in Fig. 2 each have a disjoint additional sensor unit 230, which is arranged and configured to have a spatial measuring range 112 that is at least partially different from that of the at least one sensor unit 215 of the optical measuring device 210. The additional disjoint sensor unit 230 is arranged directly on the respective optical measuring device 210. This allows the respective control unit 220 of the optical measuring device to access more sensor data from at least partially different measuring spaces 112 in order to determine the processed sensor signal 225.This enables particularly reliable pattern recognition and therefore particularly effective noise reduction.

[0057] In the illustrated embodiments, the control units 220 of the optical measuring devices 210 are also configured to determine weight sets for the weighted evaluation of sensor data for pattern recognition via a neural network based on past sensor data. These weight sets can be exchanged between different measuring devices 210 via the communication unit 223. This enables a joint learning process between the optical measuring devices 210, so that particularly precise weight sets are available for weighting received data for the control units 220.

[0058] The central processing unit 240 is further configured to receive, in addition to the received processed sensor signals 225, at least one set of sensor data from a sensor unit 215 and to determine the current atmospheric data 246 based on this data. The central processing unit 240 is further configured to use a neural network to weight the received processed sensor data 222 based on the analyzed spatial measurement ranges in order to determine the current atmospheric data. This additional weighting is compared with the weight sets from the control units 220 and enables an additional automatic learning step, thus ensuring a particularly reliable determination of the atmospheric data 246.

[0059] Furthermore, over the service life of the optical analysis system 200, the central processing device 240 automatically detects the development of the atmospheric data 246, so that an estimate of future atmospheric data can be carried out based on past atmospheric data and output via the optical output device 260.

[0060] Unlike in Fig. 1, in the optical analysis system 200 an optical measuring device 210 is directly connected to the central processing device 240 and thereby outputs its processed sensor signal and / or sensor data directly to the central processing device.

[0061] In the illustrated embodiment, the optical output device 260 displays the atmospheric data 246 via corresponding graphic symbols 250, which support an intuitive and quick understanding of the atmospheric data 246.

[0062] Fig. 3 shows an optical output 365 of a third embodiment of the optical analysis system 300 according to the invention.

[0063] The 365-day optical output displays a map that uses graphic symbols to show 250 different weather phenomena for various regions. This reduces atmospheric data to 300 pieces of information relevant to the user of the optical analysis system. This relevant information can be understood particularly quickly and intuitively. The assignment of the 250 graphic symbols to the atmospheric data can be based, for example, on a predefined rule that determines the graphic symbol to be displayed from weather data such as wind direction, wind speed, temperature, humidity, air pressure, and / or the like. The 365-day optical output can dynamically change according to the current weather situation. User-specific adjustments to the displayed area are also possible.In particular, information about different atmospheric layers at different altitudes can be displayed differently. This allows for a three-dimensional representation of atmospheric data.

[0064] Preferably, the optical output unit is additionally designed to store user-specific settings, such as resolution, scaling, color scheme, transparency, contrast, placement of graphic symbols 250 and / or the like, and to learn from preferred user-specific settings for a specific output situation in the future.

[0065] The optical output system 300 according to the invention, in combination with the intuitive graphical output 365 from this exemplary embodiment, results in a particularly user-friendly and reliable output of weather-specific atmospheric data, which can be displayed essentially in real time. This display can also be supplemented by a short-term weather forecast based on current atmospheric data and on past atmospheric data.

[0066] Fig. 4 shows a flowchart of an embodiment of a method 400 according to a further aspect of the invention.

[0067] Method 400 is designed to determine weather data from the atmosphere. The method comprises the steps described below.

[0068] A first step 410 involves dividing an atmospheric area into a plurality of spatial measurement areas. A next step 420 involves providing sensor data from the respective spatial measurement areas to a control unit assigned to the spatial measurement area from a plurality of control units.

[0069] A further step 430 involves applying pattern recognition based on predetermined training patterns for noise reduction to the provided sensor data and providing corresponding processed sensor data.

[0070] A subsequent step 440 involves outputting the processed sensor data assigned to the respective spatial measurement area as a processed sensor signal.

[0071] A next step 450 involves receiving the processed sensor signals through a central processing device.

[0072] A further step 460 involves determining and outputting current atmospheric data by applying pattern recognition based on predetermined overall training patterns to further reduce noise on the received processed sensor signals.

[0073] A final step, 470, involves providing a visual output of the current atmospheric data.

[0074] Preferably, the process steps are carried out in the specified order. In particular, the sensor data for the relevant measurement ranges are first determined, and then, in a first data processing step 430, the processed sensor data are made available. Subsequently, in a second data processing step 460, this processed sensor data is evaluated to determine the current atmospheric data. It follows from the above that the specified steps can also be supplemented by further process steps. In particular, additional defragmentation of data, such as the sensor data, can accelerate the process. Thus, the output of the current atmospheric data can preferably be performed essentially in real time.

[0075] Additionally, weighting sets for particularly effective noise reduction and / or pattern recognition can be determined within both data processing steps 430 and 460. For example, certain sensor data can be considered as having particularly low noise levels and therefore being especially reliable.

[0076] Although the exemplary embodiments relate to the application of LIDAR sensors in the optical analysis system, it follows directly from the above that the invention is not limited to such LIDAR applications. In particular, other optical sensors can also be used for evaluating the current state of the atmosphere in the relevant measurement range and for providing this sensor data.

[0077]

[0078] 100, 200, 300 optical analysis system; 110, 210 optical measuring device

[0079] 112 measuring range

[0080] 115, 215 Sensor unit

[0081] 117 optical signal

[0082] 119 sensor data

[0083] 120, 220 control unit

[0084] 122, 222 processed sensor data 123, 223 communication unit 125, 225 processed sensor signal 140, 240 central processing unit 142 total training pattern

[0085] 144 Pattern Recognition

[0086] 146, 246 current atmospheric data 160, 260 optical output device 162 user interface

[0087] 165, 365 optical output

[0088] 230 disjoint auxiliary sensor unit 250 graphic symbol

[0089] 400 procedures

[0090] 410, 420, 430, 440, 450, Steps of the procedure 460, 470

Claims

Patent claims 1. Optical analysis system (100) for determining weather data from the atmosphere, with - a plurality of optical measuring instruments (110) for analyzing at least partially different spatial measurement ranges (112), wherein the optical measuring instruments (110) each have - at least one sensor unit (115) configured to receive an optical signal (117) and to provide corresponding sensor data (119), - a control unit (120) arranged and configured to receive the sensor data (119), to apply a pattern recognition based on predetermined training patterns to the sensor data (119) for noise reduction and to provide appropriately processed sensor data (122), - a communication unit (123) configured to output the processed sensor data (122) as a processed sensor signal (125), - a central processing device (140) arranged and configured to receive the respective processed sensor signal (125) of all optical measuring devices (110) from the plurality of optical measuring devices (110) for the at least partially different spatial measurement ranges (112) and, based on a cross-validation of the corresponding processed sensor data (122), to apply a pattern recognition (144) based on predetermined overall training patterns (142) to the received processed sensor signals (125) for further noise reduction and thereby to determine and output current atmospheric data (146), - an optical output device (160) arranged and configured to receive the current atmospheric data (146) and to provide it as an optical output (165) via a user interface (162).

2. Optical analysis system (100) according to claim 1, wherein the central processing device (140) and / or the control unit (120) of at least one optical measuring device (110) is configured to perform a defragmentation of received time series-based data.

3. Optical analysis system (100) according to claim 1 or 2, wherein at least one optical measuring device (110) is configured to determine two different measured quantities via two different sensor units (115) for providing the corresponding sensor data (119) of a spatial measuring range (112).

4. Optical analysis system (100) according to claim 3, wherein the control unit (120) of a corresponding measuring device (110) is configured to evaluate the sensor data (119) of different sensor units (115) of the measuring device (110) in a correlated manner for pattern recognition.

5. Optical analysis system (100) according to at least one of the preceding claims, wherein at least one optical measuring device (110) has a disjoint additional sensor unit (230) which is arranged and configured to have a spatial measuring range (112) that is at least partially different from the at least one sensor unit (115) of the optical measuring device (110).

6. Optical analysis system (100) according to claim 5, wherein the control unit (120) of the corresponding optical measuring device (110) is further configured to perform pattern recognition based on additional sensor data (119) of the additional sensor unit (230).

7. Optical analysis system (100) according to at least one of the preceding claims, wherein at least one control unit (120) of one of the optical measuring devices (110) is configured to determine weight sets for the weighted evaluation of the sensor data (119) for pattern recognition via a neural network based on past sensor data.

8. Optical analysis system (100) according to claim 7, wherein at least two optical measuring devices (110) are configured from the plurality of optical measuring devices (110) to exchange certain weight sets and / or data relating to a completed pattern recognition via the respective communication unit (125).

9. Optical analysis system (100) according to at least one of the preceding claims, wherein the central processing device (140) is further configured to receive sensor data (119) from at least one sensor unit (115) in addition to the received processed sensor signals (125) and to determine the current atmospheric data (146) based on these sensor data (119).

10. Optical analysis system (100) according to at least one of the preceding claims, wherein the central processing device (140) is further designed to perform a weighting of the received processed sensor data (122) via a neural network depending on the spatial measurement areas (112) analyzed in order to determine the current atmospheric data (146).

11. Optical analysis system (100) according to at least one of the preceding claims, wherein the central processing device (140) is configured to perform and output an estimate of future atmospheric data (146) based on received sensor data (119) and / or processed sensor data (122).

12. Optical analysis system (100) according to at least one of the preceding claims, wherein the optical output device (160) is configured to graphically output the current atmospheric data (146) by means of a dynamically changing representation of at least two spatial dimensions.

13. Optical analysis system (100) according to at least one of the preceding claims, wherein the optical output device (160) is configured to represent the current atmospheric data (146) for a predetermined geographical area by means of a graphic symbol (250) from a predetermined plurality of graphic symbols (250).

14. Optical analysis system (100) according to at least one of the preceding claims, wherein at least one sensor unit (115) is formed by a LIDAR sensor unit.

15. Method (400) for determining weather data from the atmosphere, comprising the steps - Dividing an atmospheric area into a plurality of spatial measurement areas (112); - Providing sensor data (119) from the respective spatial measurement ranges (112) for a control unit (120) assigned to the spatial measurement range (112) from a plurality of control units (120); - Applying pattern recognition based on predetermined training patterns to the provided sensor data (119) for noise reduction and providing corresponding processed sensor data (122); - Outputting the processed sensor data (122) assigned to the respective spatial measurement area (112) as a processed sensor signal (125); - Receiving the processed sensor signals (125) by a central processing unit (140); - Determining and outputting current atmospheric data (146) by applying pattern recognition (144) based on predetermined overall training patterns (142) for further noise reduction to the received processed sensor signals (125); and - Providing an optical output (165) of the current atmospheric data (146).

Images