Regional atmospheric waveguide real-time detection method and system based on multi-source data fusion

Abstract

The invention discloses a regional atmospheric waveguide real-time detection method and system based on multi-source data fusion, and the method comprises the steps: obtaining atmospheric waveguide data collected by a multi-source sensor, and determining the spatial grid distribution of the atmospheric waveguide data and the uncertainty value of each grid point; according to the uncertainty value, constructing a self-adaptive detection scheduling model based on ant colony optimization, and generating a scheduling strategy of the detection frequency and spatial distribution of the multi-source sensor; according to the scheduling strategy, controlling the multi-source sensor to perform data acquisition to obtain optimized multi-source atmospheric waveguide data; performing fusion processing on the optimized multi-source atmospheric waveguide data to generate regional atmospheric waveguide feature data; and determining a real-time detection result corresponding to the atmospheric waveguide data according to the regional atmospheric waveguide characteristic data. According to the invention, through multi-source data fusion, ant colony optimization scheduling and real-time feature generation, the efficiency, precision and adaptability of regional atmospheric waveguide detection are significantly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of atmospheric waveguide detection, and in particular to a regional atmospheric waveguide real-time detection method and system based on multi-source data fusion. BACKGROUND

[0002] With the rapid development of wireless communication and radar technology, atmospheric waveguide, as an important meteorological phenomenon affecting electromagnetic wave propagation, has important application value in the fields of marine communication, navigation and meteorological monitoring. The formation of atmospheric waveguide is closely related to atmospheric parameters such as temperature, humidity and air pressure, and its dynamic change has a significant impact on the refraction and attenuation of signal propagation. In order to realize efficient regional atmospheric waveguide detection, the existing technology usually uses multi-source sensors (such as radar, radiosonde) to collect atmospheric data, and generates waveguide features through data fusion.

[0003] The existing atmospheric waveguide detection technology has made certain progress in multi-source data acquisition and fusion, but still has some limitations. The traditional detection method usually adopts a fixed scheduling strategy, which is difficult to adapt to the spatial heterogeneity and rapid change of atmospheric waveguide, resulting in uneven coverage or resource waste. In addition, the existing fusion method lacks dynamic optimization of data uncertainty when processing multi-source data, affecting the accuracy and real-time performance of waveguide feature generation. The present application realizes efficient data acquisition and accurate feature generation through adaptive scheduling and matching degree weighted fusion strategy of ant colony optimization, which makes up for the shortcomings of the existing technology in dynamic environment adaptability and real-time processing capability. SUMMARY

[0004] This section aims to summarize some aspects of the embodiments of the present application and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of the specification to avoid obscuring the purpose of this section, abstract and title, and such simplifications or omissions cannot be used to limit the scope of the present application.

[0005] In view of the above problems of the existing regional atmospheric waveguide real-time detection method and system based on multi-source data fusion, the present application is proposed.

[0006] Therefore, the purpose of the present application is to provide a regional atmospheric waveguide real-time detection method and system based on multi-source data fusion, which is suitable for solving the problem that the existing technology is difficult to adapt to the spatial heterogeneity and rapid change of atmospheric waveguide, resulting in uneven coverage or resource waste.

[0007] To solve the above technical problems, the present application provides the following technical solutions: In a first aspect, the embodiments of the present application provide a regional atmospheric waveguide real-time detection method based on multi-source data fusion, comprising: Acquire atmospheric waveguide data collected by a plurality of source sensors, determine spatial grid distribution of the atmospheric waveguide data and uncertainty value of each grid point; According to the uncertainty value, an adaptive detection scheduling model based on ant colony optimization is constructed to generate a scheduling strategy of detection frequency and spatial distribution of the plurality of source sensors; According to the scheduling strategy, the plurality of source sensors are controlled to collect data to obtain optimized multi-source atmospheric waveguide data; The optimized multi-source atmospheric waveguide data are fused to generate regional atmospheric waveguide feature data; According to the regional atmospheric waveguide feature data, real-time detection results corresponding to the atmospheric waveguide data are determined.

[0008] As a preferred scheme of the regional atmospheric waveguide real-time detection method based on multi-source data fusion, according to the uncertainty value, an adaptive detection scheduling model based on ant colony optimization is constructed to generate a scheduling strategy of detection frequency and spatial distribution of the plurality of source sensors, including: In the case that the uncertainty value is higher than a preset threshold, the pheromone distribution of the ant colony optimization model is initialized based on the uncertainty value of the grid point; According to the pheromone distribution, the probability of the plurality of source sensors being allocated to each grid point is calculated to generate a scheduling strategy of detection frequency and spatial distribution of high-priority grid points; The pheromone distribution is iteratively updated to optimize the scheduling strategy to preferentially cover high-uncertainty grid points.

[0009] As a preferred scheme of the regional atmospheric waveguide real-time detection method based on multi-source data fusion, according to the uncertainty value, an adaptive detection scheduling model based on ant colony optimization is constructed to generate a scheduling strategy of detection frequency and spatial distribution of the plurality of source sensors, including: In the case that the uncertainty value is lower than or equal to a preset threshold, the pheromone distribution of the ant colony optimization model is initialized based on the energy consumption constraint of the plurality of source sensors; According to the pheromone distribution and the energy consumption constraint, the probability of the plurality of source sensors being allocated to each grid point is calculated to generate a scheduling strategy of detection frequency and spatial distribution of low-priority grid points; The pheromone distribution is iteratively updated to optimize the scheduling strategy to balance detection coverage and energy consumption efficiency.

[0010] As a preferred scheme of the regional atmospheric waveguide real-time detection method based on multi-source data fusion, wherein: according to the uncertainty value, an adaptive detection scheduling model based on ant colony optimization is constructed to generate a scheduling strategy of the detection frequency and spatial distribution of the multi-source sensor, and the method further comprises: If the multi-source sensor comprises a mobile sensor, the energy consumption type of the mobile sensor is identified. In the case of dynamic constraint of the energy consumption type, the current position and residual energy consumption of the mobile sensor are analyzed to generate a path planning strategy of the mobile sensor. According to the path planning strategy and the pheromone distribution, the probability of the mobile sensor being assigned to a low-priority grid point is calculated to generate a scheduling strategy of the detection frequency and spatial distribution corresponding to the mobile sensor.

[0011] As a preferred scheme of the regional atmospheric waveguide real-time detection method based on multi-source data fusion, wherein: the optimized multi-source atmospheric waveguide data is fused to generate regional atmospheric waveguide feature data, comprising: Identify the sensor identifier of the optimized multi-source atmospheric waveguide data. According to the sensor identifier, the multi-source atmospheric waveguide data is grouped to obtain all data belonging to the same regional data group. Determine the fusion data set corresponding to the regional data group, and generate regional atmospheric waveguide feature data based on the fusion data set.

[0012] As a preferred scheme of the regional atmospheric waveguide real-time detection method based on multi-source data fusion, wherein: determining the fusion data set corresponding to the regional data group, and generating regional atmospheric waveguide feature data based on the fusion data set, comprising: According to the matching degree of the optimized multi-source atmospheric waveguide data and the regional data group, the data belonging to the same regional data group is weighted to obtain the weighted fusion degree of the regional data group. According to the order from large to small, the value is taken from the weighted fusion degree, and if the largest weighted fusion degree is greater than the first preset fusion threshold, the regional data group corresponding to the weighted fusion degree is determined as the target data set. Based on the target data set, the regional atmospheric waveguide feature data corresponding to the multi-source atmospheric waveguide data is generated.

[0013] As a preferred scheme of the regional atmospheric waveguide real-time detection method based on multi-source data fusion, wherein: based on the target data set, the regional atmospheric waveguide feature data corresponding to the multi-source atmospheric waveguide data is generated, comprising: If the maximum weighted fusion degree is less than or equal to a first preset fusion threshold, and the matching degree of the regional data set is greater than a second preset fusion threshold, a supplementary waveguide feature corresponding to the multi-source atmospheric waveguide data is generated through a machine learning model; wherein the first preset fusion threshold is greater than the second preset fusion threshold. According to the multi-source data of the regional data set and the supplementary waveguide feature, a regional atmospheric waveguide feature data corresponding to the multi-source atmospheric waveguide data is generated.

[0014] In a second aspect, to further solve the problem that the prior art is difficult to adapt to the spatial heterogeneity and rapid change of atmospheric waveguide, resulting in uneven coverage of detection or waste of resources, the embodiment of the present application provides a regional atmospheric waveguide real-time detection system based on multi-source data fusion, comprising: A data acquisition module is configured to acquire atmospheric waveguide data collected by a multi-source sensor, determine the spatial grid distribution of the atmospheric waveguide data and the uncertainty value of each grid point. A scheduling optimization module is configured to construct an adaptive detection scheduling model based on ant colony optimization according to the uncertainty value, and generate a scheduling strategy of the detection frequency and spatial distribution of the multi-source sensor. A data acquisition module is configured to acquire atmospheric waveguide data collected by a multi-source sensor, determine the spatial grid distribution of the atmospheric waveguide data and the uncertainty value of each grid point. A data fusion module is configured to fuse the optimized multi-source atmospheric waveguide data to generate regional atmospheric waveguide feature data. A feature output module is configured to determine a real-time detection result corresponding to the atmospheric waveguide data according to the regional atmospheric waveguide feature data.

[0015] In a third aspect, the embodiment of the present application provides a computer device, comprising a memory and a processor, and the memory stores a computer program, wherein when the computer program is executed by the processor, any step of the real-time detection method of regional atmospheric waveguide based on multi-source data fusion according to the first aspect of the present application is realized.

[0016] In a fourth aspect, the embodiment of the present application provides a computer readable storage medium, which stores a computer program, wherein when the computer program is executed by the processor, any step of the real-time detection method of regional atmospheric waveguide based on multi-source data fusion according to the first aspect of the present application is realized.

[0017] The beneficial effects of the present application: the present application significantly improves the efficiency, accuracy and adaptability of regional atmospheric waveguide detection through multi-source data fusion, ant colony optimization scheduling and real-time feature generation; through the adaptive detection scheduling model based on ant colony optimization, the detection frequency and spatial distribution of multi-source sensors are dynamically adjusted according to the grid point uncertainty value, high uncertainty areas are preferentially covered, and energy consumption and coverage are balanced in low uncertainty areas, sensor resource allocation is optimized, redundant detection is reduced, and overall detection efficiency is improved; Through sensor identification grouping, weighted fusion degree and target data set screening, accurate integration of multi-source data is realized, high-precision regional waveguide feature data is generated, compared with traditional fusion methods, through matching degree weighting and threshold screening, data mixing and noise interference are reduced, and the accuracy of feature extraction is improved; using a machine learning model to generate supplementary waveguide features, making up for scenes with insufficient data quality, combining real-time scheduling and efficient fusion, ensuring that the detection results are generated quickly, adapting to the dynamic changes of atmospheric waveguide, significantly improving the real-time detection capability, and being suitable for rapidly changing meteorological environment. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creating labor. Among them: Figure 1 The overall flowchart of the regional atmospheric waveguide real-time detection method based on multi-source data fusion proposed by the present application; Figure 2 The uncertainty value judgment flowchart of the regional atmospheric waveguide real-time detection method based on multi-source data fusion proposed by the present application; Figure 3 The weighted fusion degree judgment flowchart of the regional atmospheric waveguide real-time detection method based on multi-source data fusion proposed by the present application. DETAILED DESCRIPTION

[0019] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification.

[0020] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, therefore the present application is not limited by the specific embodiments disclosed below.

[0021] Second, the "one embodiment" or "an embodiment" referred to herein means a particular feature, structure, or characteristic including an implementation that can be included in at least one implementation of the application. The appearances of the "in one embodiment" or "in an embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily all referring to a single, alternative embodiment, or a single, alternative implementation.

[0022] Third, the application is described in detail in conjunction with the schematic diagram. In the detailed description of the embodiments of the application, the cross-sectional view of the device structure is partially enlarged without the general proportion for the convenience of description, and the schematic diagram is only an example, which should not limit the scope of protection of the application herein. In addition, the three-dimensional spatial dimensions of length, width and depth should be included in actual production.

[0023] Embodiment one Reference Figures 1-3 For one embodiment of the application, a regional atmospheric waveguide real-time detection method based on multi-source data fusion is provided.

[0024] Existing atmospheric waveguide detection technology has made some progress in multi-source data acquisition and fusion, but still has some limitations: traditional detection methods usually use fixed scheduling strategies, which are difficult to adapt to the spatial heterogeneity and rapid changes of atmospheric waveguide, leading to uneven coverage or resource waste. In addition, the existing fusion method lacks dynamic optimization of data uncertainty when processing multi-source data, affecting the accuracy and real-time performance of waveguide feature generation.

[0025] The present application can effectively solve the above-mentioned problems, and the following will be described in detail how to realize the regional atmospheric waveguide real-time detection method based on multi-source data fusion in combination with multiple embodiments.

[0026] Figure 1 The overall flowchart of the regional atmospheric waveguide real-time detection method based on multi-source data fusion is shown, which includes: S1: Obtain atmospheric waveguide data collected by multi-source sensors, determine the spatial grid distribution of atmospheric waveguide data and the uncertainty value of each grid point.

[0027] The multi-source sensors include, but are not limited to, the following types: radar: such as weather radar, used to collect atmospheric refractive index, temperature and humidity data in the region; unmanned aerial vehicle: equipped with temperature and humidity sensors and barometers, collecting high-resolution local atmospheric data, especially suitable for low-altitude or marine environments; radiosonde: by releasing a radiosonde balloon, obtaining atmospheric parameters (such as temperature, humidity, wind speed) in the vertical direction; ground station: fixed sensors, collecting long-term atmospheric data near the ground; the frequency and range of data collection are determined according to the requirements of the detection area, for example, in a 10km x 10km marine area, the radar collects horizontal distribution data every 5 minutes, the unmanned aerial vehicle collects local high-precision data every minute, and the radiosonde releases every hour to obtain vertical profile data.

[0028] Spatial grid distribution processing: using spatial grid division technology, the detection area is divided into a three-dimensional grid (for example, horizontal resolution 1km x 1km, vertical resolution 100m), and the collected data is mapped to the grid points by inverse distance weighted interpolation technology to generate a multi-dimensional data vector.

[0029] The uncertainty value is calculated by statistical analysis technology, and the uncertainty value of each grid point is calculated by statistical analysis technology (such as weighted average or linear combination) based on the variance of the data and the measurement error of the sensor.

[0030] S2: According to the uncertainty value, an adaptive detection scheduling model based on ant colony optimization is constructed to generate a scheduling strategy for the detection frequency and spatial distribution of the multi-source sensors, including: As shown in Figure 2 , in the case where the uncertainty value (calculated by the above step S1) is higher than the preset threshold, the pheromone distribution of the ant colony optimization model is initialized based on the uncertainty value of the grid point; The initialization of the pheromone distribution of the ant colony optimization model includes: The uncertainty value of the grid point is used to initialize the pheromone distribution of the ant colony optimization model, and the pheromone concentration reflects the detection priority of the grid point, and the formula is: ; Wherein, is the initial pheromone concentration of the grid point , is a proportional coefficient, is a default pheromone value, and the grid point higher than the threshold value obtains a higher pheromone concentration and preferentially absorbs sensor resources; According to the pheromone distribution, the probability of the multi-source sensor being allocated to each grid point is calculated to generate a scheduling strategy for the detection frequency and spatial distribution of the high-priority grid point; wherein the probability of the multi-source sensor being allocated to each grid point is calculated, a scheduling strategy of the detection frequency and spatial distribution of the high-priority grid points is generated, and the scheduling strategy comprises: ; wherein, is the probability of the multi-source sensor being allocated to each grid point at time t, is the probability of the multi-source sensor being allocated to each grid point at time t, is the pheromone concentration, and the heuristic factor formula is: , is the pheromone weighting factor, which adjusts the influence of the pheromone concentration, is the heuristic factor weighting parameter, which adjusts the influence of the heuristic factor, j represents a single index in the index set of the grid point, and specifically represents the y-axis direction grid index of the three-dimensional space grid point , J is the optional grid point set, represents all grid points to which the sensor can be allocated, t is the time step, and represents the iteration time of the ant colony optimization algorithm; The pheromone distribution is iteratively updated to optimize the scheduling strategy to preferentially cover the high-uncertainty grid points.

[0031] wherein the pheromone distribution is iteratively updated to optimize the scheduling strategy, and the scheduling strategy comprises: The pheromone distribution is updated through detection feedback, and the formula is: ; wherein, is the pheromone concentration of the grid point at time t+1, is the updated detection priority, is the pheromone evaporation coefficient, which controls the decay rate of the pheromone, is the pheromone concentration of the grid point at time t, is the pheromone concentration of the grid point at time t, is the pheromone increment coefficient, which adjusts the gain of the pheromone from the detection feedback.

[0032] In the case where the uncertainty value is lower than or equal to a preset threshold, the pheromone distribution of the ant colony optimization model is initialized based on the energy consumption constraint of the multi-source sensor; According to the pheromone distribution and the energy consumption constraint, the probability of the multi-source sensor being allocated to each grid point is calculated, and a scheduling strategy of the detection frequency and spatial distribution of the low-priority grid points is generated; The pheromone distribution is iteratively updated to optimize the scheduling strategy to balance the detection coverage and energy consumption efficiency.

[0033] In the embodiments of the present application, the proportion coefficient c, the default pheromone value , the pheromone weighting factor , and the heuristic factor weighting parameter The pheromone increment coefficient Q can be found by minimizing the scheduling error using historical atmospheric waveguide detection data, so as to find the parameter value most suitable for the current detection scene, or can be obtained through iterative training of an ant colony optimization algorithm or an expert experience method, and the embodiment is not limited in this regard.

[0034] If the multi-source sensor includes a mobile sensor, the energy consumption type of the mobile sensor is identified; The energy consumption type of the mobile sensor is identified, including: The energy consumption type of the mobile sensor is identified, including dynamic constraints and static constraints. Dynamic constraints refer to real-time changes in energy consumption with sensor operating states (such as flight distance, speed, and detection tasks), such as the flight and hovering energy consumption of a UAV. The energy consumption type is determined by the working mode of the sensor (such as flight mode, detection frequency) and the real-time state (such as battery remaining capacity). Only when the energy consumption type is dynamic constraint, subsequent path planning and scheduling are performed.

[0035] In the case of dynamic constraint of energy consumption type, the current position and remaining energy consumption of the mobile sensor are analyzed to generate a path planning strategy for the mobile sensor. The current position and remaining energy consumption of the mobile sensor are analyzed to generate a path planning strategy for the mobile sensor, including: Current position analysis: Obtain the real-time coordinates of the mobile sensor, calculate the distance to the low-priority grid points, and preferentially select grid points with shorter distances to reduce energy consumption. Remaining energy consumption analysis: Evaluate the ratio of the remaining energy consumption of the mobile sensor to the maximum energy consumption to ensure that the path planning is within the energy consumption support range. Path planning strategy: According to the current position and remaining energy consumption, generate an optimal path that covers low-priority grid points, preferentially select grid points with short distances and low energy consumption requirements, and the path planning adopts the principle of minimum energy consumption path, such as selecting the nearest grid point or planning a cruise path to cover multiple low-priority grid points. According to the path planning strategy and pheromone distribution, the probability of the mobile sensor being assigned to the low-priority grid points is calculated, and a scheduling strategy corresponding to the detection frequency and spatial distribution is generated.

[0036] It should be noted that the scheduling strategy is based on uncertain values to generate a multi-source sensor detection frequency and spatial distribution allocation scheme through an ant colony optimization algorithm, which is optimized for the above two cases to balance detection efficiency and energy consumption, and the specific generation process is as follows: First, determine the scheduling priority: when the uncertainty value of the grid point is higher than the preset threshold, mark it as high priority, these grid points usually correspond to the area of severe change of waveguide (such as abnormal refractive index area), which needs high-performance sensors (such as unmanned aerial vehicle, radar) to detect preferentially to ensure data accuracy; when the uncertainty value of the grid point is less than or equal to the preset threshold, mark it as low priority, combine the sensor energy consumption constraint (based on the normalized ratio of current energy consumption to maximum energy consumption, range [0,1]), preferentially allocate low energy consumption sensors (such as ground station) to reduce energy consumption while maintaining coverage; Second, generate the detection frequency and spatial distribution strategy (use ant colony optimization algorithm, determine the sensor allocation priority according to the uncertainty value and energy consumption constraint, generate the scheduling strategy): for high uncertainty grid points: set a higher detection frequency (preferably 1-2 times per minute), ensure fast access to dynamic change data, preferentially allocate high-performance sensors (such as unmanned aerial vehicle, radar) to high uncertainty grid points, cover the area with significant waveguide change, allocate based on the uncertainty value of the grid point, the higher the value, the higher the priority; for low uncertainty grid points: set a lower detection frequency (preferably 1 time per 5-10 minutes), reduce the number of detections to reduce energy consumption, preferentially allocate low energy consumption sensors (such as ground station) to low uncertainty grid points, ensure regional coverage while optimizing energy efficiency, consider the current energy consumption state of the sensor, preferentially select sensors with lower energy consumption; Finally, optimize the scheduling strategy: update the allocation priority by iteration, dynamically adjust the detection frequency and spatial distribution, after each detection, update the priority according to the detection feedback (such as data consistency or energy consumption change), high uncertainty grid points continue to maintain high frequency detection, low uncertainty grid points further optimize energy consumption allocation, ensure the balance of overall detection coverage and resource efficiency.

[0037] For example, assume that in a certain ocean waveguide detection area (20km×20km×5km, grid 200×200×50), a group of multi-source sensors (including unmanned aerial vehicle, radar, ground station) are performing detection tasks, the system detects that the uncertainty value U=0.75> preset threshold 0.6 of grid point , indicating that there is an abnormal refractive index, at the same time, the uncertainty value U=0.4≤ preset threshold 0.6 of grid point , the data is stable, the current position of the unmanned aerial vehicle is (5km, 5km, 0.5km), the remaining energy consumption is 40% of the maximum energy consumption, the system generates a scheduling strategy based on the ant colony optimization algorithm; high uncertainty grid point: allocate unmanned aerial vehicle to detect , frequency is 1 time per minute, preferentially capture abnormal data; low uncertainty grid point: allocate ground station to detect , frequency is 1 time per 10 minutes, reduce energy consumption; mobile sensor scheduling: unmanned aerial vehicle is identified as a dynamic constraint, plan the path to preferentially detect low priority grid points within a distance of 2km , the frequency is once every 8 minutes; after the scheduled execution, the system dynamically adjusts the allocation, optimizes the detection coverage and energy consumption in a timely manner, prevents detection blind spots or energy waste caused by improper resource allocation, and ensures the accuracy and real-time performance of the waveguide data.

[0038] It should be noted that the traditional detection scheduling method often uses fixed frequency and uniform distribution strategy, which cannot effectively handle the non-uniform distribution of grid point uncertainty and the dynamic energy consumption constraint of mobile sensor in atmospheric waveguide detection, resulting in low detection efficiency or high energy consumption. The embodiment constructs an adaptive scheduling model based on ant colony optimization, and introduces a dynamic processing mechanism for uncertainty value and energy consumption constraint, wherein: the priority scheduling mechanism of high uncertainty grid points enhances the rapid response capability to the area of severe change of waveguide, ensuring the data accuracy; the energy consumption optimization mechanism of low uncertainty grid points effectively reduces the resource consumption while maintaining the coverage; the mobile sensor path planning strategy adjusts the detection path and frequency flexibly by real-time analysis of position and energy consumption, solving the resource allocation problem under dynamic constraint. The model realizes the dynamic balance of detection efficiency, coverage and energy efficiency, breaks through the limitations of traditional fixed scheduling, overcomes the one-sidedness of single index scheduling, makes the scheduling strategy more objective and comprehensive and has foresight, provides reliable technical support for real-time detection of atmospheric waveguide, and effectively improves the detection precision and resource utilization.

[0039] S3: According to the scheduling strategy, control the multi-source sensor to collect data, and obtain the optimized multi-source atmospheric waveguide data.

[0040] Preferably, according to the scheduling strategy, controlling the multi-source sensor to collect data, comprising: based on the high uncertainty grid points, low uncertainty grid points and mobile sensor detection frequency and spatial distribution strategy generated in step S2, respectively configuring the detection task of each sensor, coordinating the working state of the sensor (such as turning on, turning off, moving path adjustment), ensuring that the high uncertainty grid points preferentially collect high frequency data, the low uncertainty grid points reduce the detection frequency to save energy, and the mobile sensor efficiently covers the specified grid points according to the path planning. Through real-time monitoring of the sensor running state and data collection feedback, dynamically adjusting the collection parameters (such as sampling interval, detection range), generating the optimized multi-source atmospheric waveguide data, containing temperature, humidity, air pressure and other key parameters, providing high quality input for data fusion in step S4.

[0041] S4: Fusion processing of the optimized multi-source atmospheric waveguide data to generate regional atmospheric waveguide feature data.

[0042] Preferably, as shown in Figure 3 , the fusion processing of the optimized multi-source atmospheric waveguide data to generate regional atmospheric waveguide feature data, comprising: The sensor identifier of the optimized multi-source atmospheric duct data is identified. Preferably, the sensor identifier of the optimized multi-source atmospheric duct data is identified, including: parsing the optimized multi-source atmospheric duct data, and extracting the sensor identifier of each piece of data. The sensor identifier is a unique identifier (such as the device number of the radar, the serial number of the unmanned aerial vehicle, and the station number of the ground station), which is used to distinguish the data collected by different sensors (such as radars, unmanned aerial vehicles, and ground stations), record the source, collection time, and spatial position.

[0043] According to the sensor identifier, the multi-source atmospheric duct data is grouped to obtain all data belonging to the same regional data group; Preferably, according to the sensor identifier, the multi-source atmospheric duct data is grouped, including: according to the sensor identifier, the optimized multi-source atmospheric duct data is classified, and the data belonging to the same region (such as the same grid point or adjacent grid point cluster) is grouped to obtain all data of the same regional data group. Each group of data contains multiple parameters (such as temperature, humidity, and air pressure), and is associated with the spatial position and sensor type, ensuring the correspondence of the data in space and source, and providing structured input for subsequent fusion processing.

[0044] The fusion data set corresponding to the regional data group is determined, and the regional atmospheric duct feature data is generated based on the fusion data set.

[0045] Further, the fusion data set corresponding to the regional data group is determined, and the regional atmospheric duct feature data is generated based on the fusion data set, including: According to the matching degree of the optimized multi-source atmospheric duct data and the regional data group, the data belonging to the same regional data group is weighted to obtain the weighted fusion degree of the regional data group; Wherein, the data belonging to the same regional data group is weighted, including: the matching degree is determined based on data consistency (such as the similarity of parameters such as temperature, humidity, and air pressure at the same grid point or adjacent grid points) and sensor reliability (such as high precision of radar and stability of ground station). Weighting processing assigns a weight (normalized to [0, 1]) to each piece of data, and the weight reflects the credibility of the data and the contribution degree to the regional characteristics. The weighted fusion degree is calculated by weighted average or weighted voting, and the fusion quality of the data group is quantified.

[0046] The value is taken from the weighted fusion degree in descending order, and if the largest weighted fusion degree is greater than the first preset fusion threshold, the regional data group corresponding to the weighted fusion degree is determined as the target data set; If the weighted fusion degree with the largest value is greater than a first preset fusion threshold, the regional data set corresponding to the weighted fusion degree is determined as the target data set, including: the corresponding regional data set is determined as the target data set. The target data set represents a regional data set with the highest fusion quality and the strongest data consistency, and is used to generate regional atmospheric waveguide feature data.

[0047] Based on the target data set, regional atmospheric waveguide feature data corresponding to the multi-source atmospheric waveguide data is generated.

[0048] Based on the target data set, regional atmospheric waveguide feature data corresponding to the multi-source atmospheric waveguide data is generated. If the weighted fusion degree with the largest value is less than or equal to the first preset fusion threshold, and the matching degree of the regional data set is greater than a second preset fusion threshold, a supplementary waveguide feature corresponding to the multi-source atmospheric waveguide data is generated through a machine learning model; wherein the first preset fusion threshold is greater than the second preset fusion threshold. Wherein, generating a supplementary waveguide feature corresponding to the multi-source atmospheric waveguide data through a machine learning model comprises: If the weighted fusion degree with the largest value is less than or equal to the first preset fusion threshold, it indicates that the fusion quality of the regional data set does not meet the requirements for directly generating feature data, but if the matching degree of the regional data set is greater than the second preset fusion threshold, it is considered that the data set still has reliability. The system inputs the regional data set into a pre-trained machine learning model (such as a neural network or a random forest), and generates a supplementary waveguide feature (such as a predicted refractive index or propagation attenuation value) based on the pattern of historical waveguide data and current multi-source data using the model. These supplementary features are combined with the multi-source data of the regional data set to generate the final regional atmospheric waveguide feature data, ensuring the integrity and accuracy of the feature data.

[0049] Based on the multi-source data of the regional data set and the supplementary waveguide feature, regional atmospheric waveguide feature data corresponding to the multi-source atmospheric waveguide data is generated.

[0050] In the embodiments of the present application, with respect to the various thresholds in data fusion, the first preset fusion threshold is determined by statistically analyzing a large number of samples of multi-source atmospheric waveguide data collected under normal detection conditions, calculating the standard deviation and confidence interval of data consistency, and conducting a comprehensive assessment based on the technical specifications and detection accuracy requirements provided by the sensor manufacturer. This is intended to screen out regional data groups with high fusion quality. The second preset fusion threshold is derived based on long-term observations of detection data in a dynamically changing environment, combined with an analysis of the response characteristics and data stability of sensor types (e.g., high-precision radar, flexibility of drones), primarily taking into account differences in data reliability under different detection scenarios. Furthermore, in actual applications, the first and second preset fusion thresholds can also be obtained by determining initial thresholds based on small-scale test data, then continuously optimizing these thresholds through continuous monitoring and feedback, or by introducing machine learning methods to obtain a more optimal threshold combination through training with historical data, etc. This embodiment does not specifically limit this.

[0051] For example, assuming that in a certain ocean waveguide detection area (20 km × 20 km × 5 km, grid size 200 × 200 × 50), the optimized multi-source atmospheric waveguide data collected in step S3 above includes detection results from radar, UAV, and ground station. The system performs the fusion processing as follows: Sensor ID grouping: parse the data stream, extract the sensor ID, and group the data into grid points The regional data set includes the temperature of the radar, the humidity of the drone and the air pressure of the ground station; the weighted fusion degree calculation is: The weights of the data groups are assigned (radar 0.5, drone 0.3, ground station 0.2), and the weighted fusion degree is calculated as 0.85 by weighted average; the target data set is determined: the weighted fusion degree 0.85> the first preset fusion threshold 0.8, and the data group is determined as the target data set, and the waveguide feature generation is supplemented: for the grid points , the weighted fusion degree is 0.75<0.8, but the matching degree is 0.6>the second preset fusion threshold 0.5. The system inputs the data group into the neural network model to generate a supplementary refractive index value of 1.45, which is combined with the original data to generate feature data.

[0052] It should be noted that the traditional atmospheric waveguide data fusion method usually adopts simple average or fixed weight fusion, which is difficult to adapt to the heterogeneity and dynamic change of multi-source data, resulting in insufficient precision of fusion results or serious noise interference. The embodiment designs a multi-level fusion mechanism based on sensor identification grouping, matching degree weighting, threshold screening and machine learning supplementary features, establishes a refined fusion system including data grouping, fusion degree calculation, target data set screening and supplementary feature generation, realizes the stereoscopic processing of multi-source atmospheric waveguide data by comprehensively considering data consistency and sensor reliability; this multi-level fusion mechanism overcomes the single processing mode of traditional fusion method, realizes the progressive and process optimization of data quality, can take appropriate fusion or supplementary strategy according to the matching degree and fusion quality of data; significantly improves the precision and reliability of regional atmospheric waveguide feature data, reduces the feature generation error caused by insufficient data or noise interference, avoids the negative influence of low-quality data on the fusion result, and realizes the optimal allocation of multi-source data fusion resources.

[0053] S5: determining real-time detection results corresponding to the atmospheric waveguide data according to the regional atmospheric waveguide feature data.

[0054] Specifically, the system analyzes the regional atmospheric waveguide feature data, extracts key parameters (such as refractive index, waveguide height, propagation attenuation), combines the predefined waveguide characteristic model or classification rule, generates real-time detection results, and the detection results include waveguide type (such as evaporation waveguide, surface waveguide), intensity level (such as strong, weak) and its spatial distribution, which are used to support ocean communication, radar detection or meteorological monitoring and other applications. The system realizes real-time processing and output, ensures that the detection results can quickly reflect the dynamic change of regional atmospheric waveguide.

[0055] In summary, the present application significantly improves the efficiency, accuracy and adaptability of regional atmospheric waveguide detection through multi-source data fusion, ant colony optimization scheduling and real-time feature generation; through the adaptive detection scheduling model based on ant colony optimization, the detection frequency and spatial distribution of multi-source sensors are dynamically adjusted according to the grid point uncertainty value, high uncertainty areas are preferentially covered, and energy consumption and coverage rate are balanced in low uncertainty areas, which optimizes the allocation of sensor resources, reduces redundant detection and improves overall detection efficiency; Through sensor identification grouping, weighted fusion and target data set screening, accurate integration of multi-source data is realized, high-precision regional waveguide feature data is generated, compared with traditional fusion method, through matching degree weighting and threshold screening, data mixing and noise interference are reduced, the accuracy of feature extraction is improved; the machine learning model is used to generate supplementary waveguide features to make up for the scene of insufficient data quality, combined with real-time scheduling and efficient fusion, the detection results are quickly generated, which adapts to the dynamic change of atmospheric waveguide, significantly improves the real-time detection ability, and is suitable for rapidly changing meteorological environment.

[0056] Embodiment 2, as an embodiment of the present application, a regional atmospheric waveguide real-time detection system based on multi-source data fusion, comprising: A data acquisition module is configured to acquire atmospheric waveguide data collected by the multi-source sensors, determine the spatial grid distribution of the atmospheric waveguide data and the uncertainty value of each grid point; A scheduling optimization module is configured to construct an adaptive detection scheduling model based on ant colony optimization according to the uncertainty value, and generate a scheduling strategy of the detection frequency and spatial distribution of the multi-source sensors; A data acquisition module is configured to acquire atmospheric waveguide data collected by the multi-source sensors, determine the spatial grid distribution of the atmospheric waveguide data and the uncertainty value of each grid point; A data fusion module is configured to fuse the optimized multi-source atmospheric waveguide data to generate regional atmospheric waveguide feature data; A feature output module is configured to determine the real-time detection result corresponding to the atmospheric waveguide data according to the regional atmospheric waveguide feature data.

[0057] Embodiment 3, as an embodiment of the present application, which is different from the previous embodiment is: If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the parts that essentially contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk, and various media that can store program codes.

[0058] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered a list of executable instructions for implementing logical functions, and can be specifically embodied in any computer-readable medium for use by an instruction execution system, device or apparatus, such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute the instructions, or in conjunction with these instructions. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transport programs for use by an instruction execution system, device or apparatus, or in conjunction with these instruction execution systems, devices or apparatuses.

[0059] More specific examples (a non-exhaustive list) of the computer- readable medium include the following: an electrical connection (electronic) having one or more wires, a portable computer diskette (magnetic), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer readable medium can be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, for instance via an optical scanner, then compiled, interpreted or otherwise processed in a suitable manner, if necessary, to generate an electronically readable version of the program, which can be stored in the computer memory.

[0060] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the embodiments described above, various steps or methods can be implemented, for example in software or firmware, stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, and in another embodiment, any of the following technologies, known in the art, or combinations thereof, can be used: discrete logic circuitry having logic gates for implementing logic functions upon an application of data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), and the like.

[0061] It should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, but not limit the present application, and although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalent replaced, without departing from the spirit and scope of the technical solutions of the present application, and all of them should be covered in the scope of the claims of the present application.

Claims

1. A real-time regional atmospheric duct detection method based on multi-source data fusion, characterized in that: include: Acquire atmospheric duct data collected by multi-source sensors, and determine the spatial grid distribution of the atmospheric duct data and the uncertainty value of each grid point; According to the uncertainty value, constructing an adaptive detection scheduling model based on ant colony optimization to generate a scheduling strategy for the detection frequency and spatial distribution of the multi-source sensor; According to the scheduling strategy, the multi-source sensor is controlled to collect data to obtain optimized multi-source atmospheric waveguide data; performing fusion processing on the optimized multi-source atmospheric duct data to generate regional atmospheric duct characteristic data; According to the regional atmospheric duct characteristic data, a real-time detection result corresponding to the atmospheric duct data is determined.

2. The method for real-time regional atmospheric duct detection based on multi-source data fusion according to claim 1, characterized in that: According to the uncertainty value, an adaptive detection scheduling model based on ant colony optimization is constructed to generate a scheduling strategy for the detection frequency and spatial distribution of the multi-source sensor, including: When the uncertainty value is higher than a preset threshold, initializing the pheromone distribution of the ant colony optimization model based on the uncertainty value of the grid point; Calculating the probability of the multi-source sensor being assigned to each grid point based on the pheromone distribution, and generating a scheduling strategy for the detection frequency and spatial distribution of high-priority grid points; The scheduling strategy is optimized by iteratively updating the pheromone distribution to preferentially cover high uncertainty grid points.

3. The method for real-time regional atmospheric duct detection based on multi-source data fusion according to claim 1, characterized in that: According to the uncertainty value, an adaptive detection scheduling model based on ant colony optimization is constructed to generate a scheduling strategy for the detection frequency and spatial distribution of the multi-source sensor, further comprising: When the uncertainty value is lower than or equal to a preset threshold, initializing the pheromone distribution of the ant colony optimization model based on the energy consumption constraint of the multi-source sensor; Calculating the probability of allocating the multi-source sensor to each grid point based on the pheromone distribution and the energy consumption constraint, and generating a scheduling strategy for the detection frequency and spatial distribution of low-priority grid points; The scheduling strategy is optimized by iteratively updating the pheromone distribution to balance detection coverage and energy efficiency.

4. The method for real-time regional atmospheric duct detection based on multi-source data fusion according to claim 3 is characterized in that: According to the uncertainty value, an adaptive detection scheduling model based on ant colony optimization is constructed to generate a scheduling strategy for the detection frequency and spatial distribution of the multi-source sensor, further comprising: If the multi-source sensor includes a mobile sensor, identifying an energy consumption type of the mobile sensor; When the energy consumption type is a dynamic constraint, analyzing the current position and remaining energy consumption of the mobile sensor to generate a path planning strategy for the mobile sensor; According to the path planning strategy and the pheromone distribution, the probability of the mobile sensor being assigned to a low-priority grid point is calculated, and a corresponding scheduling strategy for detection frequency and spatial distribution is generated.

5. The method for real-time regional atmospheric duct detection based on multi-source data fusion according to claim 1, characterized in that: The optimized multi-source atmospheric duct data is fused to generate regional atmospheric duct characteristic data, including: Identifying a sensor identifier of the optimized multi-source atmospheric duct data; Grouping the multi-source atmospheric waveguide data according to the sensor identifier to obtain all data belonging to the same regional data group; A fused data set corresponding to the regional data group is determined, and regional atmospheric duct characteristic data is generated based on the fused data set.

6. The method for real-time regional atmospheric duct detection based on multi-source data fusion according to claim 5, characterized in that: Determining a fused data set corresponding to the regional data group, and generating regional atmospheric duct characteristic data based on the fused data set, comprising: weighting the data belonging to the same regional data group according to the matching degree between the optimized multi-source atmospheric waveguide data and the regional data group to obtain a weighted fusion degree of the regional data group; Taking values ​​from the weighted fusion degrees in descending order, if the weighted fusion degree with the largest value is greater than a first preset fusion threshold, determining the regional data group corresponding to the weighted fusion degree as the target data set; Based on the target data set, regional atmospheric waveguide characteristic data corresponding to the multi-source atmospheric waveguide data is generated.

7. The method for real-time regional atmospheric duct detection based on multi-source data fusion according to claim 6, characterized in that: Generating regional atmospheric duct characteristic data corresponding to the multi-source atmospheric duct data based on the target data set includes: If the weighted fusion degree with the largest value is less than or equal to a first preset fusion threshold, and the matching degree of the regional data group is greater than a second preset fusion threshold, generating a supplementary waveguide feature corresponding to the multi-source atmospheric waveguide data through a machine learning model; wherein the first preset fusion threshold is greater than the second preset fusion threshold; Regional atmospheric waveguide characteristic data corresponding to the multi-source atmospheric waveguide data are generated based on the multi-source data of the regional data group and the supplementary waveguide characteristics.

8. A regional atmospheric duct real-time detection system based on multi-source data fusion, based on the real-time detection method according to any one of claims 1 to 7, characterized in that: include: Data acquisition module: used to obtain atmospheric waveguide data collected by multi-source sensors, determine the spatial grid distribution of the atmospheric waveguide data and the uncertainty value of each grid point; Scheduling optimization module: used to construct an adaptive detection scheduling model based on ant colony optimization according to the uncertainty value, and generate a scheduling strategy for the detection frequency and spatial distribution of the multi-source sensor; Data acquisition module: used to control the multi-source sensor to collect data according to the scheduling strategy to obtain optimized multi-source atmospheric waveguide data; Data fusion module: used for fusing the optimized multi-source atmospheric waveguide data to generate regional atmospheric waveguide characteristic data; Feature output module: determines the real-time detection results corresponding to the atmospheric waveguide data according to the regional atmospheric waveguide feature data.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for real-time detection of regional atmospheric ducts based on multi-source data fusion according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for real-time detection of regional atmospheric ducts based on multi-source data fusion according to any one of claims 1 to 7 are implemented.

Images