A cloud measurement system and a cloud measurement method

Abstract

An embodiment of this specification provides a cloud measurement system and a cloud measurement method. The system includes a laser device, a radar device, and a processor. The processor is configured to determine a first measurement device and obtain first measurement data according to a visibility value and a precipitation intensity distribution value determined from weather data, so as to determine first cloud layer data and estimate a first measurement accuracy; when the first measurement accuracy meets a first preset condition, generate target cloud layer data based on the first measurement data; when the first measurement accuracy meets a second preset condition, determine a second measurement device to obtain second measurement data and generate target cloud layer data; when the first measurement accuracy meets a third preset condition, obtain second measurement data through the laser device and the radar device simultaneously, and generate target cloud layer data; wherein, the first preset condition, the second preset condition, and the third preset condition are mutually exclusive, and the efficiency and accuracy of cloud measurement can be improved based on this system.

Description

Technical Field

[0001] This specification relates to the field of ground meteorological observation, and particularly to a cloud measurement system and a cloud measurement method. Background Art

[0002] Currently, cloud observation mainly relies on two methods: satellite remote sensing and ground-based observation. Satellite remote sensing can obtain cloud data globally, but it has limitations in spatial resolution and the ability to observe the cloud base and multi-layer clouds, and cannot meet the needs of atmospheric science research.

[0003] Therefore, it is necessary to provide a cloud measurement system and a cloud measurement method with less influence from weather to improve the efficiency and accuracy of cloud measurement. Summary of the Invention

[0004] One or more embodiments of this specification provide a cloud measurement system, including a laser device, a radar device, and a processor, and the processor is configured to: execute the following cloud measurement method.

[0005] One or more embodiments of this specification provide a cloud measurement method, and the method includes: determining a first measurement device according to the visibility value and the precipitation intensity distribution value determined from weather data, and obtaining first measurement data to determine first cloud layer data and estimate the first measurement accuracy; when the first measurement accuracy meets a first preset condition, generating target cloud layer data based on the first measurement data; when the first measurement accuracy meets a second preset condition, determining a second measurement device to obtain second measurement data and generate the target cloud layer data; when the first measurement accuracy meets a third preset condition, obtaining the second measurement data through the laser device and the radar device simultaneously, and generating the target cloud layer data; wherein, the first preset condition, the second preset condition, and the third preset condition are mutually exclusive.

[0006] Advantageous Effects: The cloud measurement system and the cloud measurement method of this specification can judge the accuracy of the first measurement data by estimating the first measurement accuracy. According to the different accuracies of the first measurement data, the target cloud layer data is determined through various methods subsequently. It can improve the accuracy and automation degree of cloud measurement, and can reasonably allocate the use of measurement devices and computing power resources, which is beneficial to improving the operation efficiency of the cloud measurement system and avoiding the generation of redundant data. Brief Description of the Drawings

[0007] This specification will be further described in the form of exemplary embodiments, and these exemplary embodiments will be described in detail through the drawings. These embodiments are not restrictive. In these embodiments, the same numbers represent the same structures, where:

[0008] Figure 1 is an exemplary module diagram of a cloud measurement system shown according to some embodiments of this specification;

[0009] Figure 2 is an exemplary flowchart of a cloud measurement method shown in some embodiments of this specification;

[0010] Figure 3 is an exemplary flowchart of determining the generation of target cloud data shown in some embodiments of this specification;

[0011] Figure 4 is a schematic diagram of determining target cloud data shown in some embodiments of this specification. Detailed implementation manners

[0012] To more clearly illustrate the technical solutions of the embodiments of this specification, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some examples or embodiments of this specification. For those of ordinary skill in the art, without creative efforts, this specification can also be applied to other similar scenarios according to these drawings. Unless obvious from the language context or otherwise stated, the same reference numerals in the drawings represent the same structure or operation.

[0013] It should be understood that the "system", "device", "unit" and / or "module" used herein is a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the described words can be replaced by other expressions.

[0014] Figure 1 is an exemplary module diagram of a cloud measurement system shown in some embodiments of this specification.

[0015] In some embodiments, as Figure 1 shown, the cloud measurement system 100 may include a laser device 110, a radar device 120, and a processor 130.

[0016] In some embodiments, the laser device 110, the radar device 120, and the processor 130 are communicatively linked.

[0017] In some embodiments, the laser device may be a laser ceilometer, which can emit and receive laser signals.

[0018] In some embodiments, the radar device may be a cloud measurement radar, which can emit and receive microwave signals.

[0019] In some embodiments, the processor can process the data and / or information obtained from the laser device 110 and the radar device 120. The processor can execute program instructions based on these data, information, and / or processing results to perform one or more functions described in this application. For more descriptions of the functions of the processor, seeFigures 2 - 4 The corresponding content.

[0020] It should be noted that the above description of the cloud measurement system and its components is only for convenience of description and does not limit this specification to the scope of the examples given. It can be understood that for those skilled in the art, after understanding the principle of the system, various modules may be arbitrarily combined without departing from this principle, or a subsystem may be formed and connected to other modules. In some embodiments, Figure 1 The laser device 110, radar device 120, and processor 130 disclosed in [reference] can be different modules in a system, or a module can implement the functions of two or more of the above modules. For example, each module can share a storage module, or each module can have its own storage module respectively. Such variations are all within the protection scope of this specification.

[0021] Figure 2 is an exemplary flowchart of the process corresponding to the cloud measurement method shown in some embodiments of this specification. As Figure 2 shown, the process corresponding to the cloud measurement method includes the following steps. In some embodiments, the process corresponding to the cloud measurement method can be executed by the processor 130.

[0022] Step 210, determine the first measurement device according to the visibility value and precipitation intensity distribution value determined from the weather data, and obtain the first measurement data to determine the first cloud layer data and estimate the first measurement accuracy.

[0023] The weather data can include temperature, humidity, wind speed, light intensity, air pressure, precipitation, etc.

[0024] In some embodiments, the processor can obtain weather data through a collection device, auxiliary device, etc. Among them, the collection device can include sensors for measuring temperature, humidity, wind speed, light intensity, air pressure, etc. The auxiliary device can include a rain gauge, visibility meter, etc.

[0025] The visibility value refers to the maximum distance at which a person's eyes can clearly identify a target under the current meteorological conditions and is used to characterize the atmospheric transparency. In some embodiments, the visibility value can be determined by a visibility meter.

[0026] The precipitation intensity distribution value can characterize the distribution of precipitation in different regions.

[0027] In some embodiments, the processor can divide the area within the measurement range of the cloud measurement system into multiple regions, and the shape and size of the regions are preset values and can be set according to actual needs. In some embodiments, the precipitation can be determined by a rain gauge.

[0028] The first measurement device refers to a measurement device used to obtain first measurement data. The first measurement device is one of a laser device or a radar device.

[0029] In some embodiments, the processor may normalize the visibility value and the precipitation intensity distribution value, convert the visibility value and the precipitation intensity into dimensionless numerical values, and then calculate the weighted value. Among them, normalization may include Min-Max normalization, etc.

[0030] In some embodiments, the processor may calculate the weighted value through a first preset algorithm. The first preset algorithm may include formula (1):

[0031] W = a×(1 - V)+b×P (1)

[0032] Where, W is the weighted value, V is the visibility value after normalization processing, P is the average value of the precipitation intensity distribution value after normalization processing, and a and b are weight coefficients respectively. In some embodiments, a and b are preset values and can be set according to actual needs.

[0033] Exemplarily, under the current meteorological conditions, the visibility value is 3 km, and the average value of the precipitation intensity distribution value is 20 mm / h. The visibility range is 0 - 10 km, and the precipitation intensity range is 0 - 50 mm / h. After Min-Max normalization processing, it is 0.3, and is = 0.4. The preset weight a is 0.4, b is 0.6, and the calculated W is 0.52.

[0034] In some embodiments, the processor may determine the first measurement device based on the weighted value.

[0035] Exemplarily, when the weighted value is less than 0.5, a laser device is selected as the first measurement device. When the weighted value is greater than or equal to 0.5, a radar device is selected as the first measurement device.

[0036] The first measurement data refers to the data collected by the measurement device for determining the first cloud data.

[0037] In some embodiments, when using a laser device to obtain the first measurement data, the emission system of the laser device may send laser signals to multiple cloud positions, and the reception system of the laser device may receive the laser signals reflected from multiple cloud positions. The laser device may convert the captured optical signal into an electrical signal through an optoelectronic conversion system. The processor may process the electrical signal to determine the exact moments of the emission and reception of the laser signal and calculate the time difference between the two.

[0038] The first measurement data may include the original data about cloud information collected by the laser device. For example, the time difference between the emission and reception of the laser signal, the intensity of the reflected signal, etc.

[0039] A cloud point refers to a point on a cloud. Cloud points can include the cloud bottom point, the cloud top point, the cloud boundary point, etc.

[0040] The cloud bottom point refers to at least one point on the lower surface of the cloud.

[0041] The cloud top point refers to at least one point on the upper surface of the cloud. In some embodiments, multiple cloud bottom points correspond one-to-one with multiple cloud top points, and the corresponding cloud bottom points and cloud top points are located in the same vertical direction.

[0042] The cloud boundary point refers to at least one point on the boundary of the cloud.

[0043] In some embodiments, the cloud point is a preset value and can be determined according to actual needs. In some embodiments, the cloud point can be represented in various ways. For example, it can be represented using longitude and latitude or in three-dimensional coordinates. In some embodiments, the processor can construct a three-dimensional space coordinate system based on the cloud measurement system, and the cloud point can be represented by three-dimensional coordinates in the three-dimensional space coordinate system.

[0044] In some embodiments, when using a radar device to obtain first measurement data, the transmitting system of the radar device can send microwave signals to multiple cloud points, and the receiving system of the radar device can receive the echo signals reflected by the multiple cloud points. The radar device can convert the received echo signals into electrical signals through a microwave signal conversion system. The processor can process the electrical signals to determine the exact moments of the emission and reception of the laser signals and calculate the time difference between the two.

[0045] The first measurement data can include the original data collected by the radar device regarding cloud information. For example, the time difference between the emission and reception of the microwave signal, the intensity of the echo signal, the distribution characteristics of the echo signal, etc.

[0046] The distribution characteristics of the echo signal refer to that the radar can obtain a distribution image of the internal structure of the cloud by sampling the echo signals at different times and different heights. For example, the distribution of strong echo areas and weak echo areas at the cloud top and inside the cloud.

[0047] The intensity of the echo signal refers to the power of the echo signal.

[0048] The first cloud data refers to data related to the cloud. In some embodiments, the first cloud data includes cloud parameters, cloud change trends, cloud development ranges, etc.

[0049] The cloud parameters refer to parameters related to the cloud point. In some embodiments, the cloud parameters can include at least one of a cloud height set, a cloud thickness set, a cloud area, a coverage range, a water content set at different positions in the cloud, etc.

[0050] The cloud height set refers to the set of the heights of the bottoms of multiple clouds. In some embodiments, the processor may determine the time difference between the emission of a laser or microwave signal to the corresponding cloud bottom point and the reception, calculate the product of the time difference and the speed of the laser or microwave signal, determine the distance between the cloud bottom point and the cloud measurement system, and calculate the height of the cloud bottom point in the vertical direction based on this distance through trigonometric functions. In some embodiments, the processor determines the set of the heights of multiple cloud bottom points in the vertical direction as the cloud height set.

[0051] In some embodiments, the processor may calculate the average height of multiple cloud bottom points.

[0052] The cloud thickness set refers to the set of the cloud thicknesses corresponding to multiple cloud points. In some embodiments, the processor may calculate the difference between the height of the top of the cloud and the height of the corresponding cloud bottom point, and determine the set of multiple differences as the cloud thickness set. Calculating the height of the cloud top is similar to calculating the height of the cloud bottom.

[0053] In some embodiments, the processor may calculate the average thickness of the cloud thicknesses corresponding to multiple cloud bottom points.

[0054] The cloud area refers to the area of the projection of the cloud on the horizontal plane. In some embodiments, the processor may determine the projection points of multiple cloud boundary points on the same horizontal plane, calculate the area of the contour formed by surrounding the multiple projection points, and confirm this area as the cloud area. The horizontal plane is a preset plane and can be set according to actual needs.

[0055] The coverage range refers to the spatial range covered by the cloud. In some embodiments, the coverage range may include the heights of multiple cloud boundary points and the cloud area.

[0056] The water content set at different positions in the cloud refers to the set of the water contents of multiple points in the cloud. In some embodiments, the water content can be directly detected by a first measuring device.

[0057] The cloud change trend refers to the trend of the change of cloud parameters over a past time period with respect to time.

[0058] In some embodiments, the processor may calculate the average value of at least one cloud parameter (such as cloud height) among the cloud parameters corresponding to multiple cloud bottom points at different time points, calculate the change amount of the average values of two adjacent time points, and calculate the ratio of the change amount to the time interval between two adjacent time points. In some embodiments, the processor may calculate the average value of multiple ratios corresponding to this cloud parameter over a period of time, and determine this average value as the average change rate of this cloud parameter.

[0059] In some embodiments, the processor may determine a cloud layer change trend based on the average change rate corresponding to any cloud layer parameter. When the average change rate is greater than 0, the cloud layer change trend indicates that the clouds are gradually strengthening, which may indicate that the clouds are rising or expanding. When the average change rate is less than 0, the cloud layer change trend indicates that the clouds are gradually decreasing, which may indicate that the clouds are descending or dissipating. When the average change rate is 0, the cloud layer change trend indicates that the clouds are remaining stable.

[0060] The cloud development range refers to the predicted cloud parameters in the future time period.

[0061] In some embodiments, the processor can determine the cloud development range based on the average rate of change. In some embodiments, the processor can determine at least one cloud parameter (e.g., cloud height) and the corresponding average rate of change at the current moment. The processor can then determine the time interval between the current moment and the future time point, calculate the product of the time interval and the average rate of change, and then sum this product with the current moment's cloud parameter to determine the cloud development range. The future time point is a preset value that can be set as needed.

[0062] The first measurement accuracy is used to characterize the measurement accuracy of the first measuring device.

[0063] In some embodiments, the processor may construct a first data table based on historical data. The first data table includes historical first cloud layer data, historical visibility values, historical precipitation intensity distribution values, historical first measurement equipment, historical first measurement accuracy, and their corresponding relationships. In some embodiments, the first data table may include the following table:

[0064]

[0065] The processor can determine the same or similar historical first cloud layer data, historical visibility value, historical precipitation intensity distribution value, and historical first measurement device based on the current first cloud layer data, visibility value, precipitation intensity distribution value, and first measurement device by querying the first data table, and determine the corresponding historical first measurement accuracy as the current first measurement accuracy.

[0066] In some embodiments, the processor is configured to estimate a first measurement accuracy based on the first cloud layer data and the cloud layer image.

[0067] Cloud layer images refer to images related to the cloud layer, such as images of the cloud layer taken from the ground or the sky. In some embodiments, the cloud layer images can be acquired by an imaging device, such as a camera or a remote sensing satellite.

[0068] In some embodiments, the processor may determine the first measurement accuracy based on the first cloud layer data and the cloud layer image by using an accuracy estimation model.

[0069] The accuracy prediction model refers to a model used to determine the first measurement accuracy. In some embodiments, the accuracy prediction model is a machine learning model. For example, a neural networks (NN) model, etc.

[0070] In some embodiments, the processor may train the accuracy prediction model based on the first sample data set.

[0071] The first sample data set includes first training samples and their corresponding first labels.

[0072] In some embodiments, the first training samples include sample first cloud data and sample cloud images. The first label is the measurement accuracy corresponding to each first training sample.

[0073] In some embodiments, the processor may determine the first training samples through historical data. The processor may select sample first cloud data and corresponding sample cloud images including different weather conditions and different geographical regions to improve the generalization ability of the model. In some embodiments, the processor may compare the ratio of the sample first cloud data in the historical data to the first cloud data independently verified through experiments and use this ratio as the first label.

[0074] In some embodiments, the processor may perform multiple rounds of iteration. At least one round of iteration includes: selecting one or more first training samples from the first sample data set, inputting the one or more first training samples into the initial accuracy prediction model to obtain model prediction outputs corresponding to the one or more first training samples; according to the model prediction outputs corresponding to the one or more first training samples and the first labels of the one or more first training samples, substituting them into the formula of a predefined loss function to calculate the value of the loss function; according to the value of the loss function, updating the model parameters in the initial accuracy prediction model in reverse; this step can be performed using various methods. For example, it can be updated based on the gradient descent method. When the iteration end condition is met, the iteration ends, and the trained accuracy prediction model is obtained.

[0075] Using the accuracy prediction model to determine the first measurement accuracy can improve the accuracy and efficiency of determining the first measurement accuracy. When determining the first measurement accuracy, considering the cloud image can predict the first measurement accuracy in combination with the actual cloud image, which is beneficial to further improving the accuracy of determining the first measurement accuracy.

[0076] Step 221, when the first measurement accuracy meets the first preset condition, generate target cloud data based on the first measurement data.

[0077] In some embodiments, the first preset condition, the second preset condition, and the third preset condition are related to a first threshold and a second threshold. Among them, the first preset condition may include that the first measurement accuracy is greater than the first threshold. The first threshold is a preset value and can be set according to actual needs.

[0078] The target cloud data refers to the data related to clouds that needs to be determined. In some embodiments, the content included in the target cloud data is the same as that of the first cloud data, but the accuracy of the target cloud data is higher than that of the first cloud data. For specific descriptions, reference can be made to the relevant descriptions in step 210.

[0079] In some embodiments, the processor may determine the target cloud data based on the first measurement data through a preset algorithm.

[0080] In some embodiments, the processor may perform filtering processing on the first measurement data obtained by the laser device through Kalman filtering, filter out the interference data in the first measurement data, and calculate the cloud height and / or cloud thickness corresponding to multiple cloud bottoms based on the processed first measurement data, which can improve the accuracy of the determined cloud height and / or cloud thickness.

[0081] In some embodiments, the processor may process the first measurement data obtained by the radar device through a Doppler frequency shift extraction algorithm, filter out the interference data in the first measurement data, and calculate the cloud height and / or cloud thickness corresponding to multiple cloud bottoms based on the processed first measurement data.

[0082] In some embodiments, the processor may, through a spatial interpolation algorithm, based on the first measurement data, determine the cloud height or cloud thickness corresponding to other points in the cloud. Other points refer to the points in the cloud other than the cloud points, and the other points are preset points and can be set according to actual needs. The processor may construct a spatial distribution model of the cloud at different time points based on the cloud points and the cloud height or cloud thickness corresponding to other points at different time points. The processor may compare the spatial distribution model at a certain time point with the spatial distribution model at the previous time point to determine the cloud change trend. The spatial interpolation algorithm is a preset algorithm and may include inverse distance weighted interpolation, etc.

[0083] In some embodiments, the processor may use GIS software (such as ArcGIS) to process the first measurement data through trend surface analysis, simulate the spatial model of the cloud, and determine the cloud development range based on the spatial model.

[0084] Step 222, when the first measurement accuracy meets the second preset condition, determine the second measurement device to obtain the second measurement data and generate the target cloud data.

[0085] In some embodiments, the second preset condition may include that the first measurement accuracy is less than the first threshold and greater than the second threshold. The second threshold is a preset value and can be set according to actual needs. The first threshold is greater than the second threshold.

[0086] The second measurement device refers to a measurement device used to obtain second measurement data. In some embodiments, the second measurement device is the same as the first measurement device.

[0087] The second measurement data refers to the data collected by the acquisition device for determining the target cloud layer data. In some embodiments, the second measurement data is similar to the first measurement data. By performing a second measurement using the same measurement device, it is beneficial to reduce random errors and improve the accuracy of the second measurement data.

[0088] In some embodiments, the processor may determine the target cloud layer data based on the second measurement data through a preset algorithm. In some embodiments, the processor determines the target cloud layer data based on the second measurement data, which is similar to the processor determining the target cloud layer data based on the first measurement data. For more content regarding determining the target cloud layer data based on the second measurement data, reference may be made to the relevant description of determining the target cloud layer data based on the first measurement data in step 221.

[0089] Step 223, when the first measurement accuracy meets the third preset condition, obtain the second measurement data simultaneously through the laser device and the radar device, and generate the target cloud layer data.

[0090] The third preset condition refers to the condition for determining whether the first cloud layer data meets the usage requirements. In some embodiments, the third preset condition may include that the first measurement accuracy is less than the second threshold. In some embodiments, the first preset condition, the second preset condition, and the third preset condition are mutually exclusive.

[0091] In some embodiments, the first threshold and the second threshold may be determined according to the morphological complexity of the cloud layer to be measured.

[0092] The cloud layer to be measured refers to the cloud layer that needs to be measured.

[0093] The morphological complexity is used to characterize the complexity of the cloud layer to be measured. The morphological complexity can be a specific value, and the larger the value, the higher the complexity. In some embodiments, the morphological complexity is determined by the fractal dimension, compactness, and standard deviation of the contour curvature obtained by identifying the cloud image. For more content regarding the cloud image, reference may be made to the relevant description in step 210.

[0094] The fractal dimension is used to characterize the self-similarity of the shape (e.g., cloud image) and the complexity of the edge. The fractal dimension can be a specific value, and the larger the value, the higher the complexity.

[0095] In some embodiments, the processor may binarize the cloud image and use the box-counting method to calculate the fractal dimension.

[0096] Compactness is used to characterize the degree to which a shape approaches a circle.

[0097] In some embodiments, the processor may determine the contour of the cloud based on the cloud boundary points, calculate the side length and area of the contour, and the processor may calculate the ratio of the square of the side length to four times the area and confirm this ratio as the compactness.

[0098] The standard deviation of the contour curvature is used to characterize the complexity of the contour of a shape.

[0099] In some embodiments, the processor may determine the contour of the cloud based on the cloud boundary points, calculate the curvatures of multiple cloud boundary points, and statistically calculate the standard deviation of the curvatures of multiple cloud boundary points.

[0100] In some embodiments, the processor may calculate the sum or product of the fractal dimension, compactness, and standard deviation of the contour curvature and confirm this sum or product as the morphological complexity.

[0101] In some embodiments, the first threshold and the second threshold may be positively correlated with the morphological complexity. When the morphological complexity of the cloud to be measured is large, it can be determined that there is a more complex physical structure inside the cloud, and the processor is more inclined to determine to use the laser device and the radar device jointly to obtain the second measurement data, and when using a single measurement device to obtain the second measurement data, the accuracy requirement for the single measurement device is higher.

[0102] Determining the first threshold and the second threshold based on the morphological complexity of the cloud can dynamically adjust the first threshold and the second threshold according to the actual situation of the cloud, so that the first preset condition, the second preset condition, and the third preset condition can be dynamically adjusted, which is beneficial to improving the accuracy of the conditions for determining the first measurement accuracy subsequently, and then the measurement device and computing power resources can be more reasonably allocated in combination with the actual cloud.

[0103] In some embodiments, the processor may calculate the target cloud data through a weighted fusion algorithm based on the second measurement data respectively obtained by the laser device and the radar device. In some embodiments, the weighted fusion algorithm may include formula (2):

[0104] M = k1×J+(1 - k1)×L (2)

[0105] Where M is the target cloud data, J is the second measurement data obtained by the laser device, L is the second measurement data obtained by the radar device, and k1 is the weight coefficient. In some embodiments, k1 may be a preset value, and its specific value may be set according to actual needs. For example, k1 may be 0.5, etc.

[0106] In some embodiments, the processor may also determine the target cloud data in other ways. For example, the processor generates second cloud data based on the second measurement data; calculates a first difference based on the second cloud data; and generates the target cloud data based on the first difference. For more information on how the processor determines the target cloud data, see Figure 3 the relevant description.

[0107] In some embodiments of the present specification, by estimating the first measurement accuracy, the accuracy of the first measurement data can be judged. According to the different accuracies of the first measurement data, the target cloud data is subsequently determined in various ways. This can improve the accuracy and automation of cloud measurement, and can reasonably allocate the use of measurement devices and computing power resources, which is beneficial to improving the operation efficiency of the cloud measurement system and avoiding the generation of redundant data.

[0108] Figure 3 is an exemplary flowchart of the process for generating the target cloud data according to some embodiments of the present specification. As Figure 3 shown, the process for generating the target cloud data includes the following steps. In some embodiments, the process for generating the target cloud data may be executed by a processor.

[0109] Step 310, generate second cloud data based on the second measurement data.

[0110] For more information on the second measurement data, see the relevant description in step 222.

[0111] In some embodiments, the content included in the second cloud data is similar to that of the first cloud data, but the accuracy of the second cloud data is higher than that of the first cloud data. For more information on the second cloud data, see the relevant description of the first cloud data in step 210.

[0112] In some embodiments, the processor may determine the second cloud data based on the second measurement data. In some embodiments, the processor determines the second cloud data based on the second measurement data, which is similar to determining the target cloud data based on the first measurement data. For more information on determining the target cloud data based on the second measurement data, see the relevant description of determining the target cloud data based on the first measurement data in step 221.

[0113] Step 320, calculate a first difference based on the second cloud data.

[0114] The first difference refers to the difference between the second cloud data corresponding to the laser device and the second cloud data corresponding to the radar device. In some embodiments, the first difference may be a specific value.

[0115] In some embodiments, the processor may calculate a first difference based on the average height and average thickness in the second cloud layer data. For more information about the average height and average thickness, reference may be made to the relevant description in step 210.

[0116] Exemplarily, the processor may calculate the height difference between the average height corresponding to the laser device and the average height corresponding to the radar device, and the thickness difference between the average thickness corresponding to the laser device and the average thickness corresponding to the radar device. The processor may calculate the mean square error of the height difference and the thickness difference, and confirm this mean square error as the first difference.

[0117] Step 330: Generate target cloud layer data based on the first difference.

[0118] In some embodiments, the processor may calculate the target cloud layer data through a weighted fusion algorithm based on the second measurement data respectively obtained by k1, the laser device, and the radar device. In some embodiments, the processor may determine k1 based on the first difference. For more information about the weighted fusion algorithm and k1, reference may be made to the relevant description in step 223.

[0119] In some embodiments, the smaller the first difference, the closer k1 can be to 0.5. When the first difference is smaller and the visibility is smaller, k1 can be smaller, with a minimum of 0. When the first difference is larger and the visibility is larger, k1 can be larger, with a maximum of 1.

[0120] In some embodiments, the processor may construct a second data table based on historical data. The second data table includes historical first differences, historical visibilities, historical k1s, and their corresponding relationships. The processor may determine k1 by querying the second data table based on the current first difference and visibility.

[0121] In some embodiments, the processor may also determine the target cloud layer data in other ways. For example, when the first difference is greater than a preset difference threshold, update the measurement parameters of the laser device and the radar device; obtain third measurement data based on the laser device and the radar device with updated parameters; generate target cloud layer data based on the third measurement data. For more information about generating target cloud layer data based on the third measurement data, reference may be made to Figure 4 the relevant description.

[0122] In some embodiments, the processor is configured to, when the first difference is less than the preset difference threshold, determine the target cloud layer data based on the second cloud layer data. For example, when the first difference is less than the preset difference threshold, it may be determined that the second measurement data obtained by using the laser device and the radar device for the same preset point of the cloud layer is the same or has a small difference, and the corresponding second cloud layer data is also equal to or close to the actual data of the cloud layer. At this time, the second cloud layer data may be used as the target cloud layer data.

[0123] The preset difference threshold is a preset value for comparison with the first difference, and the preset difference threshold can be set according to actual needs.

[0124] In some embodiments, the preset difference threshold is determined based on the visibility value and the precipitation intensity distribution value. For more information about the visibility value and the precipitation intensity distribution value, reference can be made to the relevant description in step 210.

[0125] In some embodiments, the preset difference threshold is negatively correlated with the visibility value and positively correlated with the precipitation intensity distribution value. When the visibility value is high and the precipitation intensity distribution value is low, the expected first difference is small, and the preset difference threshold can be set low. When the visibility value is low or the precipitation intensity distribution value is high, the interference of the environment on the measurement increases, and the expected first difference increases. At this time, the preset difference threshold can be increased accordingly.

[0126] The visibility value and the precipitation intensity distribution value can affect the accuracy and efficiency of detecting clouds by laser or microwave signals. When determining the preset difference threshold, by considering the visibility value and the precipitation intensity distribution value, the preset difference threshold can be dynamically adjusted according to the actual situation, which is beneficial to improving the accuracy of determining the preset difference threshold.

[0127] When the first difference is less than the preset difference threshold, it can be determined that the second measurement data obtained by the laser device and the radar device are the same or the error between them is small. The processor can directly determine the second cloud data as the target cloud data, which can improve the efficiency of determining the target cloud data and reduce the waste of computing power resources.

[0128] By calculating the first difference, the magnitude of the difference between the second measurement data obtained by the laser device and the radar device can be judged. When determining the target cloud data, by considering the first difference, the target cloud data can be comprehensively calculated in combination with the actual measurement conditions of the laser device and the radar device, which is beneficial to improving the accuracy of determining the target cloud data.

[0129] Figure 4 It is a schematic diagram of determining target cloud data shown in some embodiments of this specification.

[0130] In some embodiments, the processor is configured to update the measurement parameters of the laser device and the radar device when the first difference is greater than the preset difference threshold; obtain third measurement data based on the laser device and the radar device after updating the parameters; and generate target cloud data based on the third measurement data.

[0131] The measurement parameters of the laser device refer to various indicators and values involved in the measurement process of the laser device. For example, the number of measurements, the reflection power of the laser, the wavelength, the pulse frequency, etc.

[0132] The measurement parameters of the radar device refer to various indicators and values involved in the measurement process of the radar device. For example, the number of measurements, the transmit power of the radar, the frequency, the antenna gain, etc.

[0133] For more information about the first difference and the preset difference threshold, see Figure 3 and its related description.

[0134] In some embodiments, the processor may determine the updated measurement parameters of the laser device and the radar device through a parameter determination model based on the difference value between the second cloud layer data obtained by laser cloud measurement and radar cloud measurement. The parameter determination model is a machine learning model.

[0135] The parameter determination model refers to a model used to determine the updated measurement parameters of the laser device and the radar device. In some embodiments, the parameter determination model is a machine learning model. For example, Neural Networks (NN).

[0136] In some embodiments, the input of the parameter determination model includes the difference value between the second cloud layer data obtained by laser cloud measurement and radar cloud measurement. The difference value refers to the difference between the second cloud layer data corresponding to the laser device and the second cloud layer data obtained by the radar device. The output includes the updated measurement parameters of the laser device and the radar device. For more information about the second cloud layer data, see the relevant description in step 310.

[0137] In some embodiments, the processor may train the parameter determination model based on the second sample data set.

[0138] The second sample data set includes second training samples and their corresponding second labels.

[0139] In some embodiments, the processor may determine the second training samples through historical data. In some embodiments, the second training samples and their corresponding labels may be obtained based on experimental or historical test data.

[0140] Merely by way of example, the processor may record the difference value between the sample cloud layer data collected by the sample laser cloud measurement and the sample radar cloud measurement at the first historical time as the second training sample. Then the processor may update the measurement parameters of the laser cloud measurement and the radar cloud measurement. For example, multiple different update schemes including the sample laser cloud measurement and the sample radar cloud measurement may be randomly generated. For those in which the difference of the cloud layer data re-obtained after updating the parameters based on the update scheme is less than the preset value, the updated measurement parameters corresponding to the update scheme are used as the labels corresponding to the second training sample. For multiple second samples, labels are obtained in this way, and then multiple training samples and labels are obtained. Among them, the preset value may be set according to manual experience or historical data.

[0141] In some embodiments, the processor may use, as a historical preferred update scheme, an update scheme for which the difference in the cloud layer data re-obtained after updating the parameters based on the update scheme is less than a preset value.

[0142] The training steps for determining a model based on the second training samples with the second label are similar to the steps for training and predicting the accuracy of the model. For the specific steps, reference may be made to the relevant content above.

[0143] In some embodiments of this specification, determining the updated measurement parameters of the laser device and the radar device through the parameter determination model can improve the acquisition accuracy of cloud measurement data.

[0144] The third measurement data refers to the measurement data of the cloud layer information obtained by the laser device and the radar device after updating the parameters. In some embodiments, the third measurement data is similar to the first measurement data. For the specific content, reference may be made to the relevant description of the first measurement data above.

[0145] In some embodiments, the processor may generate target cloud layer data based on the third measurement data.

[0146] In some embodiments, the processor determines the target cloud layer data based on the third measurement data, which is similar to the processor determining the target cloud layer data based on the first measurement data. For more content on determining the target cloud layer data based on the third measurement data, reference may be made to the relevant description of determining the target cloud layer data based on the first measurement data in step 221.

[0147] In some embodiments, the processor may generate target cloud layer data through a prediction model based on the second measurement data and the third measurement data. Among them, the prediction model is a machine learning model. For more content on the second measurement data, reference may be made to the relevant description in step 222.

[0148] The prediction model refers to a model for generating target cloud layer data. In some embodiments, the prediction model is a machine learning model. For example, Neural Networks (NN).

[0149] In some embodiments, the input of the prediction model includes the second measurement data and the third measurement data, and the output includes the target cloud layer data.

[0150] In some embodiments, the processor may train the prediction model based on the third sample data set.

[0151] The third sample data set includes the third training samples and their corresponding third labels.

[0152] In some embodiments, the third training samples and their corresponding labels may be obtained based on experiments or historical test data.

[0153] By way of example only, the processor may use the measurement data of the laser device and the radar device before the update of the measurement parameters corresponding to the historical preferred update scheme and the measurement data of the laser device and the radar device after the update of the measurement parameters as the third training sample.

[0154] In some embodiments, the third label includes real cloud data, which can be a real cloud parameter sequence, a real cloud change trend index, a real cloud development range, etc.

[0155] In some embodiments, the third label can be obtained in various ways. For example, fine information is obtained through devices such as multi-band radars and satellite remote sensing, and the third label is obtained through methods such as multi-sensor data fusion. For multiple third training samples, labels are obtained in this way, and then multiple training samples and the third label are obtained.

[0156] The training steps for training the prediction model based on the third training sample with the third label are similar to the steps for the training accuracy estimation model. For the specific steps, reference can be made to the relevant content above.

[0157] In some embodiments of this specification, the target cloud data is generated through a prediction model based on the multiple acquisition data of the radar device and the laser device, making the prediction of the cloud data more accurate.

[0158] In some embodiments of this specification, when the first difference exceeds the preset difference threshold, the system updates the measurement parameters of the laser device and the radar device, obtains the third measurement parameter according to the measurement of the device after the parameter update, and dynamically adjusts the measurement parameter according to the real-time data, enhancing the accuracy and reliability of the measurement. The system can effectively reduce the error caused by the measurement deviation, thereby improving the overall quality of the finally generated target cloud data.

[0159] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only an example and does not constitute a limitation to this specification. Although not explicitly stated here, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are proposed in this specification, so such modifications, improvements, and corrections still belong to the spirit and scope of the exemplary embodiments of this specification.

[0160] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other deformations may also fall within the scope of this specification. Therefore, by way of example rather than limitation, alternative configurations of the embodiments of this specification can be considered to be in line with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.

Claims

1. A cloud measurement system, comprising a laser device, a radar device, and a processor. The laser device refers to a laser ceilometer, and the radar device refers to a cloud radar. The processor is configured to: Determine a first measurement device according to the visibility value and precipitation intensity distribution value determined from weather data, and obtain first measurement data based on the first measurement device to determine first cloud layer data and estimate the first measurement accuracy. The first measurement device refers to the measurement device for obtaining the first measurement data, and the first measurement device refers to the laser device or the radar device; When the first measurement accuracy meets the first preset condition, generate target cloud layer data based on the first measurement data through a preset algorithm. The first preset condition means that the first measurement accuracy is greater than a first threshold. The preset algorithm includes Kalman filtering, Doppler frequency shift extraction algorithm, spatial interpolation algorithm, or trend surface analysis method; When the first measurement accuracy meets the second preset condition, determine a second measurement device to obtain second measurement data and generate target cloud layer data based on the preset algorithm. The second preset condition means that the first measurement accuracy is less than the first threshold and greater than a second threshold. The second measurement device refers to the measurement device for obtaining the second measurement data, and the second measurement device refers to the same measurement device as the first measurement device. By using the same measurement device for the second measurement, the random error is reduced; When the first measurement accuracy meets the third preset condition, obtain the second measurement data through the laser device and the radar device at the same time, generate second cloud layer data based on the second measurement data, and calculate a first difference based on the average height and average thickness in the second cloud layer data. The first difference refers to the difference between the second cloud layer data corresponding to the laser device and the second cloud layer data corresponding to the radar device; when the first difference is greater than a preset difference threshold, based on the difference value between the second cloud layer data obtained from the laser device and the radar device, determine the updated measurement parameters of the laser device and the radar device through a parameter determination model. The parameter determination model is a machine learning model; obtain third measurement data based on the laser device and the radar device with updated parameters; generate the target cloud layer data based on the second measurement data and the third measurement data through a prediction model. The third preset condition means that the first measurement accuracy is less than the second threshold. The second measurement data refers to the data for determining the target cloud layer data. The prediction model is a machine learning model; Among them, The first preset condition, the second preset condition, and the third preset condition are mutually exclusive. The first threshold and the second threshold are determined according to the morphological complexity of the cloud layer to be measured. The morphological complexity is determined by the fractal dimension, compactness, and standard deviation of contour curvature obtained by recognizing the cloud layer image.

2. The system according to claim 1, wherein The processor is configured to: Estimate the first measurement accuracy based on the first cloud layer data and the cloud layer image.

3. The system according to claim 1, characterized in that The processor is configured to: When the first difference is less than the preset difference threshold, based on the second cloud layer data corresponding to the laser device and the second cloud layer data corresponding to the radar device, determine the target cloud layer data.

4. A cloud measurement method, characterized in that, The method includes: According to the visibility value and precipitation intensity distribution value determined from weather data, determine the first measurement device and obtain first measurement data based on the first measurement device to determine the first cloud layer data and estimate the first measurement accuracy. The first measurement device refers to the measurement device for obtaining the first measurement data, and the first measurement device refers to a laser device or a radar device; When the first measurement accuracy meets the first preset condition, based on the first measurement data, generate the target cloud layer data through a preset algorithm. The first preset condition means that the first measurement accuracy is greater than the first threshold, and the preset algorithm includes Kalman filtering, Doppler frequency shift extraction algorithm, spatial interpolation algorithm, or trend surface analysis method; When the first measurement accuracy meets the second preset condition, determine the second measurement device to obtain second measurement data and generate the target cloud layer data based on the preset algorithm. The second preset condition means that the first measurement accuracy is less than the first threshold and greater than the second threshold. The second measurement device refers to the measurement device for obtaining the second measurement data, and the second measurement device refers to the same measurement device as the first measurement device. By using the same measurement device for the second measurement, the random error is reduced; When the first measurement accuracy meets the third preset condition, simultaneously obtain the second measurement data through the laser device and the radar device, generate the second cloud layer data based on the second measurement data, calculate the first difference based on the average height and average thickness in the second cloud layer data. The first difference refers to the difference between the second cloud layer data corresponding to the laser device and the second cloud layer data corresponding to the radar device; when the first difference is greater than the preset difference threshold, based on the difference value between the second cloud layer data obtained by the laser device and the radar device, determine the updated measurement parameters of the laser device and the radar device through a parameter determination model. The parameter determination model is a machine learning model; obtain the third measurement data based on the laser device and the radar device with updated parameters; based on the second measurement data and the third measurement data, generate the target cloud layer data through a prediction model. The third preset condition means that the first measurement accuracy is less than the second threshold. The second measurement data refers to the data for determining the target cloud layer data. The prediction model is a machine learning model. The laser device refers to a laser ceilometer, and the radar device refers to a cloud radar; Among them, the first preset condition, the second preset condition, and the third preset condition are mutually exclusive. The first threshold and the second threshold are determined according to the morphological complexity of the cloud layer to be measured. The morphological complexity is determined by the fractal dimension, compactness, and standard deviation of the contour curvature obtained by identifying the cloud layer image.

5. The method according to claim 4, wherein The estimating the first measurement accuracy includes: Estimate the first measurement accuracy based on the first cloud data and the cloud image.

6. The method according to claim 4, wherein The generating the target cloud data based on the first difference includes: When the first difference is less than the preset difference threshold, determine the target cloud data based on the second cloud data corresponding to the laser device and the second cloud data corresponding to the radar device.

Images