Dynamic weighted fusion processing method for mobile meteorological multi-source detection data

Abstract

This invention relates to the field of meteorological data processing technology, specifically to a dynamic weighted fusion processing method for mobile meteorological multi-source detection data. The method includes: constructing a two-factor correction model to accurately correct the radial reflectivity factor using attitude influence coefficients and pulse width correction coefficients, effectively offsetting systematic errors caused by the platform's dynamic characteristics; simultaneously, segmenting the data based on cloud water density distribution and correcting the ground-based microwave radiometer data by combining the contribution values ​​of each segment to brightness temperature observations, solving the observation bias problem caused by uneven cloud structure, and providing high-quality preprocessed data for subsequent fusion; and dynamically calculating confidence weights by quantifying the adaptability errors and dynamic reliability factors of various data sources, enabling the data fusion weights to be adjusted in real time according to the quality of the detection data, thereby improving the accuracy and reliability of multi-source data fusion.

Description

Technical Field

[0001] This invention relates to the field of meteorological data processing technology, specifically to a dynamic weighted fusion processing method for mobile meteorological multi-source detection data. Background Technology

[0002] With the rapid development of meteorological detection technology, mobile meteorological detection systems (such as mobile radar and ground-based microwave radiometers) have been widely used in fields such as severe weather monitoring and short-term nowcasting due to their advantages of high flexibility and wide coverage.

[0003] The dynamic characteristics of mobile detection platforms bring additional data quality challenges. During movement, fluctuations in platform attitude (such as pitch and roll angles) can cause systematic deviations in radar detection signals, and dynamic changes in pulse width can also affect the accuracy of echo signals, thereby reducing the accuracy of inversion of key parameters such as cloud water density. Observational data from ground-based microwave radiometers are easily affected by the inhomogeneity of the vertical structure of clouds, and a single observation cannot fully reflect the true thermal radiation state of the target cloud. Directly fusing uncorrected raw data will lead to distorted fusion results.

[0004] Existing multi-source meteorological data fusion methods mostly adopt fixed-weight fusion strategies, which fail to fully consider the dynamic error characteristics of mobile sounding data and the adaptability differences of different data sources. It is difficult to dynamically adjust the fusion weights according to the data quality. Furthermore, there is a lack of targeted preprocessing mechanisms for heterogeneous data, which fails to effectively correct observation biases caused by platform attitude, pulse width fluctuations, and cloud structure inhomogeneity. As a result, the accuracy and reliability of the fusion results are insufficient, and they cannot meet the demand for high-quality data for refined meteorological services.

[0005] Therefore, there is an urgent need for a processing method that can adapt to the characteristics of multi-source meteorological data, dynamically correct observation biases, and achieve adaptive weighted fusion, so as to improve the fusion quality and application value of multi-source meteorological data. Summary of the Invention

[0006] The purpose of this invention is to provide a dynamic weighted fusion processing method for mobile meteorological multi-source detection data: aiming to solve the technical problem that existing multi-source meteorological data fusion methods have insufficient accuracy and reliability of fusion results, and cannot meet the demand for high-quality data for refined meteorological services.

[0007] A dynamic weighted fusion processing method for multi-source mobile meteorological data, the method includes: Collect mobile meteorological data sources, including ground-based microwave radiometer observation data and mobile radar detection data; The water content density distribution of the target cloud layer is determined based on mobile radar detection data. Based on the water content density distribution of the target cloud layer, the target cloud layer is segmented to obtain several labeled cloud layer segments. The contribution of the labeled cloud segment to the brightness temperature observation value of the target cloud is calculated. Based on the contribution value of the brightness temperature observation value, the radiation brightness temperature in the ground-based microwave radiometer observation data is corrected to obtain the ground-based microwave radiometer observation data of the target cloud. A unified geographic grid is constructed to map the corrected ground-based microwave radiometer observation data and mobile radar detection data to the same grid. The confidence weights of various data sources are dynamically calculated to achieve adaptive fusion of multi-source data.

[0008] Furthermore, determining the water density distribution of the target cloud layer based on mobile radar detection data specifically includes the following processes: Mobile radar detection data includes echo signals from target clouds. These echo signals are collected after the mobile radar system transmits detection signals towards the target clouds. Cloud height is determined based on these echo signals. Place Radial reflectivity factor at time Radial Doppler velocity and pulse width Mobile radar detection data includes data from mobile radar systems. The attitude angles and position coordinates at each moment, where the attitude angles include the pitch angle. and roll angle ; Construct a two-factor correction model to calculate the corrected equivalent reflectance factor. ; ; in, This is the attitude influence coefficient. This is the pulse width correction factor. The standard pulse width of radar. Atmospheric attenuation coefficient, To detect altitude; Based on the corrected equivalent reflectivity factor Determine the water density distribution of the target cloud layer : ; in, The constant coefficients, These are the inversion coefficients. The Doppler velocity is used as a reference to correct the effect of particle motion on water density.

[0009] Furthermore, the process of obtaining the attitude influence coefficient includes the following steps: First, the overall acquisition period of attitude data is set, and the period is evenly divided into several consecutive sub-periods. The attitude angle data sequence in each sub-period is collected through the synchronous cooperation of the Beidou high-precision positioning and timing unit and the inertial measurement unit. This sequence contains the pitch angle and roll angle corresponding data collected multiple times in each sub-period. At the same time, the measured standard data in the same period is obtained through the ground reference station and compared with the detection data collected by the mobile platform at the same time to calculate the true error value of the detection data in each sub-period. For each sub-period, first calculate the average value of all pitch angle data and the average value of all roll angle data within that sub-period. Then, calculate the squared difference between each pitch angle data and the average pitch angle, and the squared difference between each roll angle data and the average roll angle. Add the squared pitch angle difference and the squared roll angle difference at the same acquisition time to obtain the sum of squared attitude deviations at that time. Sum the sum of squared attitude deviations at all acquisition times within a sub-period, and then divide by the total number of acquisition points within that sub-period to obtain the mean squared attitude deviation. Finally, take the square root of this mean to obtain the attitude variability of that sub-period, which is used to characterize the stability of the attitude angles within that sub-period. A rectangular coordinate system is established with the attitude fluctuation of each sub-period as the horizontal axis and the actual error value of the detection data of the corresponding sub-period as the vertical axis. The attitude fluctuation and the corresponding actual error value of all sub-periods are marked as data points in the coordinate system, and the attitude-error correlation curve is plotted. Two consecutive data points on the attitude-error correlation curve are selected sequentially. The difference between the true error value of the latter data point and the true error value of the former data point is calculated. Then, the difference between the attitude variability of the latter data point and the attitude variability of the former data point is calculated. The error difference is divided by the variability difference to obtain the slope of the set of consecutive data points. The number of valid slopes with positive values ​​is counted. This number represents the number of correlations in which the detection error increases synchronously when the attitude variability increases. At the same time, the average value of the absolute values ​​of all slopes is calculated to characterize the overall change in the impact of attitude variability on the error. The product of the average of all absolute slope values ​​and the ratio of the number of effective slopes to the total number of slopes is denoted as the attitude influence coefficient.

[0010] Furthermore, the process of obtaining the pulse width correction coefficient includes the following steps: Based on the design parameters of the mobile radar system and meteorological detection standards, a fixed pulse width reference value is determined. This reference value represents the optimal pulse width for the radar under standard atmospheric conditions, fixed detection range, and a stationary platform. Actual pulse width variation curves under different detection scenarios are obtained. The portion of the curve exceeding the pulse width reference value is marked as a high-deviation curve, and the corresponding time period is marked as a high-deviation time period. The portion of the curve below the pulse width reference value is marked as a low-deviation curve, and the corresponding time period is marked as a low-deviation time period. By comparing ground-based reference station measured data with radar detection data, the difference between the measured and theoretical values ​​of the cloud equivalent reflectivity for each time period is obtained. This refers to reflectivity deviation. Periods with reflectivity deviations greater than a preset deviation threshold are marked as high-deviation periods, and periods with reflectivity deviations less than or equal to the preset deviation threshold are marked as low-deviation periods. Periods overlapping with high-deviation pulse width periods are marked as high-deviation overlapping periods, and periods overlapping with low-deviation pulse width periods are marked as low-deviation overlapping periods. The durations of high-deviation and low-deviation overlapping periods are summed to obtain the total overlap deviation duration. The ratio of the total overlap deviation duration to the total duration of the entire detection cycle is calculated to obtain the overlap deviation duration percentage. This overlap deviation duration percentage is recorded as the pulse width correction coefficient.

[0011] Furthermore, based on the water density distribution of the target cloud layer, the target cloud layer is segmented to obtain several labeled cloud layer segments. The specific process includes the following steps: For any two cloud layers at any altitude above the target cloud layer, obtain the water content density respectively. and Set a threshold for water content density difference. Calculate the water content density and absolute value of the difference ,Will and To make a comparison, if The cloud layers at two different altitudes are recorded as the same labeled cloud layer segment. The cloud layers at the two altitudes are marked as different labeled cloud layer segments.

[0012] Furthermore, the calculation of the contribution of the labeled cloud segment to the brightness temperature observation of the target cloud layer specifically includes the following process: Cloud segmentation based on ground-based microwave radiometer data acquisition Raw brightness temperature observations Simultaneously record the thickness of the marked cloud segments output by the BeiDou high-precision positioning and timing unit. The thickness of the marked cloud layer segment is the height difference between the upper and lower layers of the marked cloud layer segment. The thickness percentage of the labeled cloud layer segments is used as the basic weight. : ;in, The thickness of the target cloud layer; Calculate and label cloud layer segments Contribution of brightness temperature observations to the target cloud layer : .

[0013] Furthermore, the radiation brightness temperature in the ground-based microwave radiometer observation data is corrected based on the contribution value of the brightness temperature observation value. The specific process for obtaining the ground-based microwave radiometer observation data after target cloud correction includes the following steps: A radiation brightness temperature correction model is constructed, and the original observations are corrected by incorporating the contribution values ​​of each segment: ; in, The corrected radiative brightness temperature corresponding to the target cloud layer. To measure the radiation brightness temperature of the target cloud layer using a ground-based microwave radiometer. To indicate the number of cloud segments, To label the atmospheric transmittance corresponding to cloud layer segments, The total atmospheric transmittance of the target cloud layer. This refers to the temperature of the cosmic background radiation.

[0014] Furthermore, constructing a unified geographic grid and mapping the corrected ground-based microwave radiometer observation data and mobile radar detection data to the same grid specifically includes the following processes: A unified geographic grid is constructed, and data mapping is achieved using adaptive weighted interpolation. ;in, Data to be mapped Mapping weights corresponding to grid cells Data to be mapped Spatial resolution, This represents the total amount of data on the grid cell, where the data to be mapped includes corrected ground-based microwave radiometer observation data and mobile radar detection data.

[0015] Furthermore, dynamically calculating the confidence weights of various data sources to achieve adaptive fusion of multi-source data specifically includes the following processes: The adaptive error of various data sources is quantified, the dynamic reliability factor of various data sources is calculated, and the adaptive fusion weight is calculated based on the adaptive error and the dynamic reliability factor to achieve adaptive fusion of multi-source data.

[0016] Compared to existing solutions, the beneficial effects achieved by this invention are: Improving data preprocessing accuracy and reducing the impact of observation bias: This method constructs a two-factor correction model to comprehensively consider the impact of mobile radar platform attitude fluctuations and pulse width changes on the detection data. It uses the attitude influence coefficient and pulse width correction coefficient to accurately correct the radial reflectivity factor, effectively offsetting the systematic errors caused by the platform's dynamic characteristics. At the same time, it performs segmented processing based on the cloud water density distribution and corrects the ground-based microwave radiometer data by combining the contribution values ​​of each segment to the brightness temperature observation value. This solves the observation bias problem caused by the uneven cloud structure and provides high-quality preprocessed data for subsequent fusion.

[0017] Achieving dynamic adaptive fusion and optimizing the reliability of fusion results: This method abandons the traditional fixed-weight fusion mode. By quantifying the adaptive error and dynamic confidence factor of various data sources, it dynamically calculates the confidence weight, so that the data fusion weight can be adjusted in real time according to the quality of the detected data. When the data accuracy is high, the corresponding weight increases, and when the error is large, the weight automatically decreases, ensuring that the fusion result is more consistent with the real meteorological conditions and improving the accuracy and reliability of multi-source data fusion. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0019] Figure 1 This is a flowchart of the first dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating the second dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to an embodiment of the present invention. Figure 3 This is a flowchart of the third dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to an embodiment of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more exemplary embodiments. Numerous specific details are provided in the following description to give a full understanding of exemplary embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced with one or more of the specific details omitted, or other methods, components, steps, etc., can be employed. In other instances, well-known structures, methods, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of this disclosure.

[0022] This embodiment provides a dynamic weighted fusion processing method for mobile meteorological multi-source detection data. Figure 1 This is a flowchart illustrating the first dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to an embodiment of the present invention. Figure 1 As shown, the method includes the following steps: Step S101: Collect mobile meteorological data sources, including ground-based microwave radiometer observation data and mobile radar detection data; Ground-based microwave radiometer observation data can include near-surface temperature, humidity, air pressure, etc.

[0023] Step S102: Determine the water content density distribution of the target cloud layer based on the mobile radar detection data, and perform cloud layer segmentation processing on the target cloud layer based on the water content density distribution of the target cloud layer to obtain several labeled cloud layer segments. Step S103: Calculate the contribution value of the marked cloud segment to the brightness temperature observation value of the target cloud, and correct the radiation brightness temperature in the ground-based microwave radiometer observation data based on the contribution value of the brightness temperature observation value to obtain the corrected ground-based microwave radiometer observation data of the target cloud. Step S104: Construct a unified geographic grid, map the corrected ground-based microwave radiometer observation data and mobile radar detection data to the same grid, dynamically calculate the confidence weights of various data sources, and achieve adaptive fusion of multi-source data.

[0024] In summary, this method constructs a two-factor correction model that comprehensively considers the impact of mobile radar platform attitude fluctuations and pulse width variations on the detection data. It uses the attitude influence coefficient and pulse width correction coefficient to accurately correct the radial reflectivity factor, effectively offsetting the systematic errors caused by the platform's dynamic characteristics. Simultaneously, it performs segmented processing based on cloud water density distribution, and combines the contribution values ​​of each segment to the brightness temperature observations to correct the ground-based microwave radiometer data. This solves the observation bias problem caused by uneven cloud structure, providing high-quality preprocessed data for subsequent fusion. By quantifying the adaptability error and dynamic confidence factor of various data sources and dynamically calculating the confidence weight, the data fusion weight can be adjusted in real time according to the quality of the detection data, improving the accuracy and reliability of multi-source data fusion.

[0025] In some embodiments, in step S102, determining the water density distribution of the target cloud layer based on mobile radar detection data specifically includes the following process: Mobile radar detection data includes echo signals from target clouds. These echo signals are collected after the mobile radar system transmits detection signals towards the target clouds. Cloud height is determined based on these echo signals. Place Radial reflectivity factor at time Radial Doppler velocity and pulse width Mobile radar detection data includes data from mobile radar systems. The attitude angles and position coordinates at each moment, where the attitude angles include the pitch angle. and roll angle ; Construct a two-factor correction model to calculate the corrected equivalent reflectance factor. ; ; in, This is the attitude influence coefficient. This is the pulse width correction factor. The standard pulse width of radar. Atmospheric attenuation coefficient, For altitude detection; it is worth noting that the standard radar pulse width It is 0.8 Atmospheric attenuation coefficient Selection criteria: The detection range of mobile radar (vehicle-mounted or ship-mounted) is typically 0.5-10km. At a detection distance of 1km, the attenuation is approximately 15% (signal strength retained 85%); at a detection distance of 5km, the attenuation is approximately 53% (signal strength retained 47%); and at a detection distance of 10km, the attenuation is approximately 78% (signal strength retained 22%). Based on the attenuation pattern, the atmospheric attenuation coefficient is derived. Take 0.15 .

[0026] Based on the corrected equivalent reflectivity factor Determine the water density distribution of the target cloud layer : ; in, The constant coefficients, These are the inversion coefficients. For reference Doppler velocity, a value of 5 is used. This is used to correct the effect of particle motion on water density. In this embodiment, the constant coefficient... Set as The inversion coefficient for fixed radar is usually set to 0.2, but the signal stability of mobile radar is slightly lower than that of fixed radar, so the coefficient needs to be appropriately reduced to weaken the error amplification effect; a value of 0.18 can reduce the sensitivity of the inversion results to reflectivity measurement errors.

[0027] Furthermore, Figure 2 This is a flowchart illustrating the second dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to an embodiment of the present invention. Figure 2 As shown, the process of obtaining the attitude influence coefficient includes the following steps: Step S201: Segmented acquisition of attitude data; First, the overall acquisition period of attitude data is set, and the period is evenly divided into several consecutive sub-periods. The attitude angle data sequence in each sub-period is collected through synchronous cooperation between the Beidou high-precision positioning and timing unit and the inertial measurement unit. At the same time, the measured standard data in the same period is obtained through the ground reference station and compared with the detection data collected by the mobile platform at the same time to calculate the true error value of the detection data in each sub-period. The attitude angle data sequence includes pitch and roll angle data collected multiple times within each sub-period. Step S202, calculate the attitude fluctuation of the sub-period; For each sub-period, first calculate the average value of all pitch angle data and the average value of all roll angle data within that sub-period. Then, calculate the squared difference between each pitch angle data and the average pitch angle, and the squared difference between each roll angle data and the average roll angle. Add the squared pitch angle difference and the squared roll angle difference at the same acquisition time to obtain the sum of squared attitude deviations at that time. Sum the sum of squared attitude deviations at all acquisition times within a sub-period, and then divide by the total number of acquisition points within that sub-period to obtain the mean squared attitude deviation. Finally, take the square root of this mean to obtain the attitude fluctuation of that sub-period. Among them, attitude fluctuation is used to characterize the stability of attitude angles during this period.

[0028] Step S203: Plot the attitude-error correlation curve; A rectangular coordinate system is established with the attitude fluctuation of each sub-period as the horizontal axis and the actual error value of the detection data of the corresponding sub-period as the vertical axis. The attitude fluctuation and the corresponding actual error value of all sub-periods are marked as data points in the coordinate system, and the attitude-error correlation curve is plotted. Step S204, Calculation of attitude influence coefficient and influence factor; Two consecutive data points on the attitude-error correlation curve are selected sequentially. The difference between the true error value of the latter data point and the true error value of the former data point is calculated. Then, the difference between the attitude variability of the latter data point and the attitude variability of the former data point is calculated. The error difference is divided by the variability difference to obtain the slope of the set of consecutive data points. The number of valid slopes with positive values ​​is counted. This number represents the number of correlations in which the detection error increases synchronously when the attitude variability increases. At the same time, the average value of the absolute values ​​of all slopes is calculated to characterize the overall change in the impact of attitude variability on the error. Step S205: Multiply the average of all absolute slope values ​​and the ratio of the number of effective slopes to the total number of slopes, and record the product as the attitude influence coefficient.

[0029] Furthermore, Figure 3 This is a flowchart illustrating the third dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to an embodiment of the present invention. Figure 3 As shown, obtaining the pulse width correction coefficient specifically includes the following steps: Step S301: Based on the design parameters of the mobile radar system and the meteorological detection standards, determine a fixed pulse width reference value. This reference value is the optimal pulse width of the radar under standard atmospheric conditions, fixed detection distance, and stationary platform conditions. Step S302: Obtain the actual pulse width variation curves under different detection scenarios. Mark the part of the curve that is higher than the pulse width reference value as the pulse width high deviation curve and the corresponding time period as the pulse width high deviation time period; mark the part of the curve that is lower than the pulse width reference value as the pulse width low deviation curve and the corresponding time period as the pulse width low deviation time period. Step S303: By comparing the measured data from the ground reference station with the radar detection data, the difference between the measured value and the theoretical value of the cloud equivalent reflectivity for each time period is obtained, i.e., reflectivity deviation; the time periods with reflectivity deviation greater than the preset deviation threshold are marked as high reflectivity deviation time periods, and the time periods with reflectivity deviation less than or equal to the preset deviation threshold are marked as low reflectivity deviation time periods. Step S304: Mark the overlapping periods between the pulse width high deviation period and the reflectivity high deviation period as high deviation overlapping periods, and mark the overlapping periods between the pulse width low deviation period and the reflectivity low deviation period as low deviation overlapping periods; sum the durations of the high deviation overlapping periods and the low deviation overlapping periods to obtain the total overlap deviation duration. Step S305: Calculate the ratio of the total coincidence deviation duration to the total duration of the entire detection cycle to obtain the coincidence deviation duration ratio; record the coincidence deviation duration ratio as the pulse width correction coefficient.

[0030] In some embodiments, segmenting the target cloud layer based on its water density distribution to obtain several labeled cloud segments specifically includes the following process: For any two cloud layers at any altitude above the target cloud layer, obtain the water content density respectively. and Set a threshold for water content density difference. Calculate the water content density and absolute value of the difference ,Will and To make a comparison, if The cloud layers at two different altitudes are recorded as the same labeled cloud layer segment. The cloud layers at the two altitudes were designated as different labeled cloud layer segments. It's worth noting that, to ensure the observation data of the two labeled cloud layer segments are more likely to be similar, The value can be five percent.

[0031] In some embodiments, calculating the contribution of labeled cloud segment to the brightness temperature observation of the target cloud layer specifically includes the following process: Cloud segmentation based on ground-based microwave radiometer data acquisition Raw brightness temperature observations Simultaneously record the thickness of the marked cloud segments output by the BeiDou high-precision positioning and timing unit. The thickness of the marked cloud layer segment is the height difference between the upper and lower layers of the marked cloud layer segment. The thickness percentage of the labeled cloud layer segments is used as the basic weight. : ;in, The thickness of the target cloud layer; Calculate and label cloud layer segments Contribution of brightness temperature observations to the target cloud layer : .

[0032] Furthermore, the radiation brightness temperature in the ground-based microwave radiometer observation data is corrected based on the contribution value of the brightness temperature observation value. The specific process for obtaining the ground-based microwave radiometer observation data after target cloud correction includes the following steps: A radiation brightness temperature correction model is constructed, and the original observations are corrected by incorporating the contribution values ​​of each segment: ; in, The corrected radiative brightness temperature corresponding to the target cloud layer. To measure the radiation brightness temperature of the target cloud layer using a ground-based microwave radiometer. To indicate the number of cloud segments, To label the atmospheric transmittance corresponding to cloud layer segments, The total atmospheric transmittance of the target cloud layer. This refers to the cosmic background radiation temperature. It's worth noting that atmospheric transmittance is obtained as follows: the detection channels of a microwave radiometer (such as the 22GHz and 31GHz water vapor channels, and the 60GHz and 118GHz oxygen channels) are sensitive to atmospheric molecule absorption. By comparing the observed brightness temperature values ​​at different altitudes with the cosmic background radiation temperature, and combining this with an atmospheric radiative transfer model, the atmospheric transmittance along the observation path (including the altitude of the cloud layer) can be directly calculated.

[0033] In some embodiments, constructing a unified geographic grid and mapping the corrected ground-based microwave radiometer observation data and mobile radar detection data to the same grid specifically includes the following process: A unified geographic grid is constructed, and an adaptive weighted interpolation method is used to achieve data mapping: ;in, Data to be mapped Mapping weights corresponding to grid cells Data to be mapped Spatial resolution, This represents the total amount of data on the grid cell, where the data to be mapped includes corrected ground-based microwave radiometer observation data and mobile radar detection data.

[0034] In some embodiments, dynamically calculating the confidence weights of various data sources to achieve adaptive fusion of multi-source data specifically includes the following processes: Quantifying the adaptive error of various data sources: For the corrected ground-based microwave radiometer observation data: Operating parameters of the ground-based microwave radiometer (Value range [0,1], 1 indicates normal operation) and data consistency test value (Values ​​range [0,1], where 1 indicates perfect consistency), calculate the fitness error. : ; Among them, the operating status parameters of the ground-based microwave radiometer The following parameters were obtained: the number of normally operating ground-based microwave radiometers was counted relative to the total number of ground-based microwave radiometers. The ratio of the number of normally operating radiometers to the total number was recorded as the operating status parameter of the ground-based microwave radiometer. Data consistency test value The method for obtaining the data is as follows: If multiple ground-based microwave radiometers of the same type exist, calculate the mean deviation of the current device's data from other devices' data, and record this mean deviation as the data consistency test value. .

[0035] For mobile radar detection data: Calculate the adaptive error Target cloud visibility based on mobile radar detection data collected from target grid points Wind speed and cloud coverage (Obtained through satellite cloud image inversion); ; in, The threshold is 20km (good visibility threshold). The extreme wind speed threshold is 20 m / s; considering that visibility and cloud cover have a more significant impact on radar perception, the weight is adjusted accordingly. Set to 0.4, weight Set to 0.2, weight Set it to 0.4.

[0036] Calculate the dynamic reliability factor for various data sources; Calculate the dynamic reliability factor of the corrected ground-based microwave radiometer observation data. : ;in, To ensure the accuracy of the equipment calibration, data is obtained by periodically calibrating the ground-based microwave radiometer. This is the data transmission packet loss rate, measured by communication link monitoring instruments.

[0037] Calculate the dynamic reliability factor of mobile radar detection data : ; in, To ensure data freshness, the interval between the timestamp of the mobile radar detection data acquisition and the timestamp of the current data fusion is recorded. ; .

[0038] Adaptive fusion weights are calculated based on adaptive error and dynamic reliability factor to achieve adaptive fusion of multi-source data: Corrected adaptive fusion weights for ground-based microwave radiometer observations: ; Adaptive fusion weights for mobile radar detection data: ; Based on adaptive fusion weights, the fusion value of the target grid point data is calculated: ; in, The fused value of the target grid point data. This is the corrected ground-based microwave radiometer observation data. This is data detected by mobile radar.

[0039] It is worth noting that the above formulas are all dimensionless calculations, and the preset parameters in the formulas should be set by those skilled in the art according to the actual situation.

[0040] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0041] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0042] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0043] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0044] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0045] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A dynamic weighted fusion processing method for mobile meteorological multi-source detection data, characterized in that, The methods include: Collect mobile meteorological data sources, including ground-based microwave radiometer observation data and mobile radar detection data; The water content density distribution of the target cloud layer is determined based on mobile radar detection data. Based on the water content density distribution of the target cloud layer, the target cloud layer is segmented to obtain several labeled cloud layer segments. The contribution of the labeled cloud segment to the brightness temperature observation value of the target cloud is calculated. Based on the contribution value of the brightness temperature observation value, the radiation brightness temperature in the ground-based microwave radiometer observation data is corrected to obtain the ground-based microwave radiometer observation data of the target cloud. A unified geographic grid is constructed to map the corrected ground-based microwave radiometer observation data and mobile radar detection data to the same grid. The confidence weights of various data sources are dynamically calculated to achieve adaptive fusion of multi-source data.

2. The dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to claim 1, characterized in that, Determining the water density distribution of a target cloud layer based on mobile radar detection data specifically includes the following process: Mobile radar detection data includes echo signals from target clouds. These echo signals are collected after the mobile radar system transmits detection signals towards the target clouds. Cloud height is determined based on these echo signals. Place Radial reflectivity factor at time Radial Doppler velocity and pulse width Mobile radar detection data includes data from mobile radar systems. The attitude angles and position coordinates at each moment, where the attitude angles include the pitch angle. and roll angle ; Construct a two-factor correction model to calculate the corrected equivalent reflectance factor. ; ； in, This is the attitude influence coefficient. This is the pulse width correction factor. The standard pulse width of radar. Atmospheric attenuation coefficient, To detect altitude; Based on the corrected equivalent reflectance factor Determine the water density distribution of the target cloud layer : ； in, The constant coefficients, These are the inversion coefficients. The Doppler velocity is used as a reference to correct the effect of particle motion on water density.

3. The dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to claim 2, characterized in that, The process of obtaining the attitude influence coefficient includes the following steps: Attitude data is collected in segments; Calculation of attitude variability in sub-periods; Plotting the attitude-error correlation curve; Calculation of attitude influence coefficient and influence factor; The product of the average of all absolute slope values ​​and the ratio of the number of effective slopes to the total number of slopes is denoted as the attitude influence coefficient.

4. The dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to claim 2, characterized in that, Specific details on obtaining pulse width correction coefficient The process includes the following: Determine a fixed pulse width reference value; The actual pulse width variation curves under different detection scenarios are obtained. The portion of the curve exceeding the pulse width reference value is marked as a high-deviation curve, and the corresponding time period is marked as a high-deviation time period. The portion of the curve below the pulse width reference value is marked as a low-deviation curve, and the corresponding time period is marked as a low-deviation time period. By comparing the measured data from the ground reference station with the radar detection data, the difference between the measured and theoretical values ​​of the cloud equivalent reflectivity for each time period is obtained, i.e., reflectivity deviation. Time periods with reflectivity deviations greater than a preset deviation threshold are marked as high-deviation time periods. The time periods when the reflectivity deviation is less than or equal to the preset deviation threshold are marked as low reflectivity deviation periods; the time periods that overlap with the pulse width high deviation periods are marked as high deviation overlapping periods; the time periods that overlap with the pulse width low deviation periods are marked as low deviation overlapping periods; the durations of the high deviation overlapping periods and the low deviation overlapping periods are summed to obtain the total overlap deviation duration; the ratio of the total overlap deviation duration to the total duration of the entire detection cycle is calculated to obtain the overlap deviation duration percentage; the overlap deviation duration percentage is recorded as the pulse width correction coefficient.

5. The dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to claim 1, characterized in that, Based on the water content density distribution of the target cloud layer, the target cloud layer is segmented to obtain several labeled cloud layer segments. The specific process includes the following steps: For any two cloud layers at any altitude above the target cloud layer, obtain the water content density respectively. and Set a threshold for water content density difference. Calculate the water content density and absolute value of the difference ,Will and To make a comparison, if The cloud layers at two different altitudes are recorded as the same labeled cloud layer segment. The cloud layers at the two altitudes are marked as different labeled cloud layer segments.

6. The dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to claim 1, characterized in that, The calculation of the contribution of labeled cloud segments to the brightness temperature observation of the target cloud layer includes the following process: Cloud segmentation based on ground-based microwave radiometer data acquisition Raw brightness temperature observations Simultaneously record the thickness of the marked cloud segments output by the BeiDou high-precision positioning and timing unit. The thickness of the marked cloud layer segment is the height difference between the upper and lower layers of the marked cloud layer segment. The thickness percentage of the labeled cloud layer segments is used as the basic weight. ; Calculate and label cloud layer segments Contribution of brightness temperature observations to the target cloud layer : .

7. The dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to claim 6, characterized in that, The process of correcting the radiation brightness temperature in the ground-based microwave radiometer observation data based on the contribution value of the brightness temperature observation value to obtain the ground-based microwave radiometer observation data after target cloud layer correction includes the following steps: A radiation brightness temperature correction model is constructed, and the original observations are corrected by incorporating the contribution values ​​of each segment: ； in, The corrected radiative brightness temperature corresponding to the target cloud layer. To measure the radiation brightness temperature of the target cloud layer using a ground-based microwave radiometer. To indicate the number of cloud segments, To label the atmospheric transmittance corresponding to cloud layer segments, The total atmospheric transmittance of the target cloud layer. This refers to the temperature of the cosmic background radiation.

8. The dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to claim 1, characterized in that, Constructing a unified geographic grid and mapping the corrected ground-based microwave radiometer observation data and mobile radar detection data to the same grid specifically includes the following process: constructing a unified geographic grid and using an adaptive weighted interpolation method to achieve data mapping.

9. The dynamic weighted fusion processing method for mobile meteorological multi-source detection data according to claim 1, characterized in that, Dynamically calculating the confidence weights of various data sources to achieve adaptive fusion of multi-source data includes the following processes: The adaptive error of various data sources is quantified, the dynamic reliability factor of various data sources is calculated, and the adaptive fusion weight is calculated based on the adaptive error and the dynamic reliability factor to achieve adaptive fusion of multi-source data.

Images