Distributed artificial influence weather operation cooperation instruction issuing system based on Internet of Things

Abstract

The invention belongs to the technical field of software systems, and discloses a distributed artificial influence weather operation cooperation instruction issuing system based on the Internet of Things, which adopts a cloud side end three-layer architecture. The cloud platform layer fuses multi-source meteorological data through a perception model, the rain cloud tracking prediction model analyzes rain cloud changes, and the task model generates an operation instruction based on a large language model; the edge node layer processes and distributes a cache instruction, and supports local decision when no cloud instruction exists; and the terminal equipment layer completes data collection and job execution. According to the method, multi-source data space-time fusion, rain cloud trajectory dynamic tracking and intelligent generation of the operation scheme are realized, the optimal catalysis window can be determined, the equipment type, position and operation parameters can be accurately planned, and dynamic adjustment of redundant equipment is supported. The operation accuracy and efficiency are remarkably improved in effect, the operation opportunity judgment error is reduced, the equipment response delay is shortened to the second level, the rain increasing efficiency is improved, and reliable support is provided for disaster prevention, activity guarantee and the like.

Description

Technical Field

[0001] The present invention belongs to the technical field of meteorological systems, and in particular relates to a distributed artificial weather modification operation collaborative instruction issuing system based on the Internet of Things. Background Art

[0002] Meteorological intervention techniques aim to regulate local weather processes through scientific means to address natural disasters such as droughts, floods, and hail, or to ensure the smooth progress of major events. Their core goal is to promote or inhibit precipitation by altering cloud microphysical structure. For example, artificial rainmaking involves seeding cold clouds with ice-forming agents such as silver iodide and dry ice to produce large ice crystals, accelerating the conversion of cloud water into precipitation. Meanwhile, artificial rain suppression involves excessively seeding ice nuclei upwind of a target area, preventing clouds from forming large enough raindrops or causing them to fall prematurely. Common intervention methods include aerial seeding, rocket launches, and anti-aircraft artillery bombardment, with catalysts primarily consisting of silver iodide, dry ice, and powdered salt. This technology, developed in the 1940s, has been widely used worldwide. Its principles are based on cloud physics: cold cloud catalysis utilizes the ice crystal effect to promote precipitation, while warm cloud catalysis utilizes hygroscopic agents to enlarge cloud droplets, resulting in rainfall. Furthermore, artificial hail suppression suppresses the formation of large hailstones by increasing the number of hail embryos, which compete for water. These technologies have played an important role in agricultural drought relief, ecological restoration, and disaster prevention, but they still face technical bottlenecks.

[0003] Current weather modification operations are highly dependent on human decision-making, and present the following core issues: Data processing and decision-making are inefficient. Traditional methods rely on manual analysis of weather radar and satellite data, combined with empirical evidence to determine the timing and location of operations. These methods lack the ability to integrate multi-source data in real time. For example, operators must manually identify cloud features and calculate catalyst quantities, which is time-consuming and susceptible to subjective factors.

[0004] Equipment coordination is insufficient. The scheduling and parameter settings of rockets, aircraft, ground-based smoke furnaces, and other equipment require manual coordination and cannot be dynamically adjusted according to real-time cloud conditions. For example, when the speed or intensity of rain clouds changes, manual instructions may lag, resulting in poor operational results.

[0005] Evaluating operational effectiveness is difficult. Traditional statistical methods struggle to accurately quantify the impact of human intervention, and there is a lack of real-time feedback mechanisms. For example, the actual effectiveness of artificial rainmaking can be masked by natural precipitation fluctuations, requiring verification through complex comparative experiments or numerical simulations.

[0006] Furthermore, operational efficiency is hampered by issues such as insufficient technological maturity, aging equipment, and uneven personnel expertise. For example, some areas still use outdated anti-aircraft gun equipment, which has low catalytic efficiency and poses safety risks. Inadequate operator training can lead to operational errors or inappropriate parameter settings. Summary of the Invention

[0007] In order to solve the problems existing in the prior art, the present invention provides a distributed artificial weather modification operation collaborative instruction issuance system based on the Internet of Things. Through the automated data acquisition and integration of multi-source meteorological data, rain clouds are judged and predicted, so that the system can automatically generate tasks according to needs and issue instructions through the Internet of Things.

[0008] The technical solution adopted in the present invention is: In a first aspect, the present invention provides a distributed weather modification operation collaborative instruction issuance system based on the Internet of Things, which uniformly issues task instructions for weather modification operations within a determined target area, including: The cloud platform layer, serving as the data control center, includes a perception model that integrates and processes multi-source meteorological data for the target area through data channels, a rain cloud tracking and prediction model, and a task model based on a large language model. The terminal equipment layer, which serves as the underlying data collection and execution end, includes meteorological monitoring equipment, artificial influence operation equipment, and auxiliary equipment; and The edge node layer, as the middle layer of the distributed architecture, is deployed in the regional operation center or edge computing server, including the data processing station and the device management gateway. The data processing station processes and caches the instructions from the cloud platform layer and the data from the terminal device layer within the management scope, and the device management gateway distributes the instructions and converts the data protocol.

[0009] In combination with the first aspect, the present invention provides a first implementation scheme of the first aspect, wherein the edge node layer includes the smallest meteorological management unit within a determined target area as a node, and is equipped with a node server, which interacts with the server of the cloud platform layer through a wired communication network and interacts with the terminal device through a wireless IoT communication network.

[0010] In combination with the first aspect, the present invention provides a second implementation of the first aspect, wherein the edge node layer further includes a local decision module, which processes the data obtained from the terminal device layer through the local decision module equipped with a locally deployed and trained large language model to obtain local weather status data and upload it to the cloud platform layer; The preliminary instructions issued by the cloud platform layer are processed. If the preliminary instructions from the cloud platform layer are not obtained within the set rain cloud window time, the edge node layer processes the latest preliminary instructions based on the obtained meteorological status data to form task instructions and sends them to the corresponding terminal device layer.

[0011] In combination with the first aspect, the present invention provides a third implementation of the first aspect, wherein the meteorological monitoring equipment includes a meteorological radar, an anemometer, a rain gauge, a thermometer, a hygrometer, a barometer, and a sounding balloon; The artificial influence operation equipment includes ground launching equipment and aerial spreading equipment.

[0012] In combination with the first aspect, the present invention provides a fourth implementation of the first aspect, wherein the perception model obtains multi-source meteorological data for the target area through an online data channel, and performs data fusion with weight distribution after processing the multi-source data to form operation reference data for the target area.

[0013] In combination with the fourth implementation manner of the first aspect, the present invention provides a fifth implementation manner of the first aspect, wherein the online data acquired by the perception model includes local monitoring data acquired by a plurality of meteorological detection devices at the terminal device layer, and meteorological satellite images acquired through an access network, and the perception model recognizes and processes the acquired meteorological satellite images using an image processing algorithm to acquire meteorological satellite data; After the perception model cleans and spatially unifies the local monitoring data and meteorological satellite data, it assigns weights to the data with spatial and temporal labels through dynamic weighting, performs spatial and temporal association through a multimodal Transformer model, and generates a grid fusion feature tensor for the target area. Finally, the perception model outputs a grid dataset and a time series dataset. The rain cloud tracking prediction model processes the grid dataset and the time series dataset and outputs visualization data of rain cloud change data within the target area for no more than 72 hours, as well as the predicted trajectory changes of rain clouds.

[0014] In combination with the fifth embodiment of the first aspect, the present invention provides a sixth embodiment of the first aspect, wherein the task model processes and obtains the optimal catalytic window T of the rain cloud based on the rain cloud change data for the target area and obtained by the rain cloud tracking and prediction model; Then, based on the grid dataset, time series dataset, and visualization data of rain cloud predicted trajectory changes obtained by the perception model, the location where the artificial influence operation equipment needs to reach and the preliminary instructions for the corresponding operations are determined. The cloud platform layer sends the instructions to the edge node layer, which processes the instructions and forms task instructions corresponding to the terminal devices within the communication range.

[0015] In combination with the sixth implementation manner of the first aspect, the present invention provides a seventh implementation manner of the first aspect, wherein the task model is trained based on an existing large language model, and historical meteorological data is collected during training. The historical meteorological data includes a grid dataset and a time series dataset output by the perception model, and visualization data of rain cloud prediction trajectory changes output by the rain cloud tracking prediction model. The collected data is cleaned, normalized, and feature extracted, and then the actual operation status and effect evaluation data of the corresponding artificial weather modification operation are recorded. The preprocessed data is divided into training set, validation set and test set in proportion. The training set is used for model parameter learning, the validation set is used to adjust the model's hyperparameters, and the test set is used to evaluate the model's final performance. A deep learning model is used to process time series data and spatial features, with historical meteorological data as input and the corresponding optimal catalytic window, operating equipment type, location and operating instructions as output. The model parameters are continuously adjusted through the backpropagation algorithm to output task instructions.

[0016] In combination with the sixth implementation manner of the first aspect, the present invention provides an eighth implementation manner of the first aspect, wherein, when generating a preliminary instruction, the task model first obtains device information obtained by the cloud platform through the device management gateway of the edge node layer to perform a full-area scan of devices in the target area, and the task model generates a device capability matrix based on the device information; the preliminary instruction generated by the task model includes alternative adjustment parameters, and the alternative parameters are associated with performance data of alternative devices in the device capability matrix; When a terminal device fails during the execution of a task instruction, the local decision module of the edge node starts the adjustment logic, calls the pre-stored alternative adjustment parameters, selects alternative devices from the device capability matrix, and calculates the replacement coefficient to correct the parameters, keeping the total catalyst amount change value less than the set threshold, and sends collaborative instructions to the surrounding devices through the device management gateway to synchronously correct the device startup timing.

[0017] The beneficial effects of the present invention are: (1) The technology of this invention realizes the automation of the entire process from meteorological data collection and analysis to operation instruction generation by building a three-layer distributed architecture of cloud, edge and end and an intelligent task model, significantly improving the accuracy and efficiency of weather modification operations. The system can integrate multi-source data such as satellites, radars, and ground sensors in real time to accurately identify rain cloud types, movement trajectories, and optimal catalytic windows, avoiding the subjectivity and lag of manual analysis and reducing errors in judging the timing of operations; (2) The task model of the present invention, based on deep learning and meteorological principles, can autonomously generate detailed plans including equipment type, operation location, time window and catalyst parameters, replacing the traditional operation planning model that relies on manual experience. Through dynamic optimization algorithms, the system can adjust the equipment coordination strategy according to real-time cloud conditions. For example, when rain clouds arrive early or their intensity suddenly changes, it automatically triggers redundant equipment and corrects the sowing parameters to ensure that the matching degree between operation instructions and actual meteorological conditions is improved, significantly reducing the problem of operation failure caused by manual decision-making delays; (3) The present invention adopts a distributed architecture to support seamless collaboration of multiple types of equipment, such as rockets, drones, and ground smoke stoves. By combining local processing of edge nodes with global scheduling in the cloud, it can achieve real-time monitoring of equipment status and rapid issuance of instructions. In remote areas or complex terrain scenarios, the system uses multi-mode transmission technology to ensure equipment connection, solving the problem of blind spots in traditional manual operations without ground networks, shortening equipment response delays from tens of minutes in manual operations to seconds, and expanding the scope of operation coverage. (4) The present invention introduces a closed-loop mechanism of reinforcement learning and effect evaluation. By training the model with historical operation data, it continuously optimizes core parameters such as catalyst dosage calculation and sowing height selection, thereby improving the efficiency of artificial rainmaking and the success rate of rain elimination. At the same time, the system's built-in safety rules and equipment redundancy strategies can automatically avoid sensitive areas such as residential areas and airports, reduce the risk of human operational errors, and improve the safety factor of operations, providing reliable technical support for scenarios such as major event security, agricultural drought resistance, and ecological restoration. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is an architectural diagram of an instruction issuing system in an embodiment of the present invention; Figure 2 It is a process diagram of the task model of the present invention. DETAILED DESCRIPTION

[0019] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Generally, the components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations.

[0021] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for protection, but merely represents selected embodiments of the present application. All other embodiments obtained by persons of ordinary skill in the art based on the embodiments in the present application without making any creative efforts shall fall within the scope of protection of the present application.

[0022] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings.

[0023] In the description of this application, it should be noted that if the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. appear, the orientation or position relationship indicated is based on the orientation or position relationship shown in the accompanying drawings, or the orientation or position relationship in which the product of the application is usually placed when in use. It is only for the convenience of describing this application and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation. Therefore, it cannot be understood as a limitation on this application. In addition, if the terms "first", "second", etc. appear in the description of this application, they are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0024] Furthermore, the use of terms such as "horizontal" and "vertical" in the description of this application does not necessarily imply that a component must be absolutely horizontal or suspended, but rather that it can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal than "vertical" and does not mean that the structure must be completely horizontal, but rather that it can be slightly tilted.

[0025] It should also be noted that, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," and "connected" should be understood broadly. For example, they may refer to fixed connections, detachable connections, or integral connections; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal connections between two components. A person of ordinary skill in the art will understand the specific meanings of the above terms in this application based on the specific circumstances.

[0026] Example 1: This embodiment discloses a distributed artificial weather modification operation collaborative instruction issuance system based on the Internet of Things, which is intended to automatically generate and issue task instructions for regional artificial weather modification operations in a target area, and the determined time window is generally within 72 hours.

[0027] The artificial weather modification involved in this embodiment is the activity of intervening in local atmospheric physical processes through scientific and technological means to achieve specific weather control objectives, specifically the management of rain clouds in this embodiment. Its core operations include: Artificial rainmaking involves spreading catalysts such as silver iodide and dry ice into clouds through aircraft, rockets, and ground smoke stoves to promote the formation of ice crystals or the condensation of water droplets, thereby increasing precipitation efficiency.

[0028] Artificial rain elimination can catalyze precipitation in advance upstream of the target area, or spread excessive amounts of ice nuclei to inhibit the formation of raindrops to ensure weather conditions for major events.

[0029] The system automates the entire operation process through distributed control and precise collaboration. The core mechanisms are as follows: Real-time monitoring and data fusion: Integrate data from multiple sources such as meteorological satellites, radars, and drones to build a three-dimensional monitoring model for cloud and water resources. Edge nodes transmit information such as equipment status and catalyst remaining in real time through the Internet of Things module to achieve dynamic monitoring.

[0030] Next, collaborative operation instructions are generated. Based on numerical forecast models and AI algorithms, combined with cloud structure and catalyst diffusion simulations, a coordinated plan for multiple operation points is generated. The instructions are then issued and transmitted to the operation terminal via multiple links, including 5G and satellite communications, supporting resumable transmission and encrypted authentication. The terminal device automatically executes the instructions and provides status feedback, using Beidou time synchronization and local logic analysis, without the need for human intervention.

[0031] To evaluate the effectiveness of the system's commands, radar, satellite remote sensing, and ground-based observation data, combined with AI quantitative assessment models, analyze cloud microphysical changes before and after operations to verify rain enhancement efficiency. The system continuously optimizes its operational strategies through machine learning, for example, adjusting catalyst type and placement based on historical data.

[0032] Specifically, the system in this embodiment refers to Figure 1 The architecture is shown in Figure 1, including: The cloud platform layer, serving as the data control center, includes a perception model that integrates and processes multi-source meteorological data for the target area through data channels, a rain cloud tracking and prediction model, and a task model based on a large language model. The terminal equipment layer, which serves as the underlying data collection and execution end, includes meteorological monitoring equipment, artificial influence operation equipment, and auxiliary equipment; and The edge node layer, as the middle layer of the distributed architecture, is deployed in the regional operation center or edge computing server, including the data processing station and the device management gateway. The data processing station processes and caches the instructions from the cloud platform layer and the data from the terminal device layer within the management scope, and the device management gateway distributes the instructions and converts the data protocol.

[0033] The distributed architecture's middle layer, deployed in regional operations centers or edge computing servers, handles local data processing, real-time control logic execution, and collaborative interaction with the cloud platform. The edge node layer comprehensively plans the administrative divisions and topographical conditions within the target area, establishing nodes based on the smallest meteorological management units within the target area. Node servers are deployed, interacting with cloud platform servers via wired communication networks and with terminal devices via wireless IoT communication networks.

[0034] The purpose of setting up the edge node layer includes: Data preprocessing and aggregation: receiving raw meteorological data uploaded by terminal devices, performing denoising, calibration and local storage, such as caching high-frequency data within 10 minutes, filtering out invalid information and uploading key data, such as cloud movement trajectory and real-time water content, to the cloud platform to reduce network transmission pressure.

[0035] Real-time status monitoring and local control: real-time monitoring of the operating status of all terminal devices within the jurisdiction, such as the ammunition remaining of rocket launchers and the flight endurance of drones. In emergency situations such as network interruptions, emergency operations can be performed based on local preset rules to ensure system reliability.

[0036] Command distribution and collaborative scheduling: receiving operation plans issued by the cloud platform, such as launching three rockets in area A at 9 o'clock tomorrow, breaking them down into specific equipment control instructions, and coordinating multiple devices to execute them in a time sequence.

[0037] Edge nodes are regionally distributed units that collaborate with the cloud platform for local computing and global optimization. They upload pre-processed regional data and device status to the cloud platform and receive global operation plans and model parameter updates from the cloud platform. The cloud platform provides edge nodes with global meteorological data and operation model algorithm support, and coordinates the operation plans of each edge node across regions.

[0038] And refer to Figure 1 The architecture diagram and Figure 2 As shown in the flowchart, the edge node layer and the cloud platform layer, as well as the edge node layer and the terminal device layer, all interact with each other through the IoT network data. The construction of the IoT network includes the following technical means: Edge nodes in remote areas connect to the cloud platform via satellite communications, receiving global operation plans and distributing instructions to end devices. In areas with densely populated end devices, edge nodes communicate with end devices via LoRa or Mesh networks, reducing the load on satellite links. For example, multiple ground-based stoves connect to edge nodes via LoRa, which then upload data via satellite.

[0039] Many mobile terminal devices are usually moved manually, but they can also be operated unmanned, using a separate satellite communication method similar to Starlink for positioning and data exchange.

[0040] Furthermore, the perception model acquires multi-source meteorological data for the target area through online data channels, processes the multi-source data, and performs weighted data fusion to form operational reference data for the target area. The online data acquired by the perception model includes local monitoring data obtained by several meteorological detection devices at the terminal device layer, as well as meteorological satellite imagery acquired through the access network. The perception model uses image processing algorithms to identify and process the acquired meteorological satellite images and obtain meteorological satellite data.

[0041] After the perception model cleans and spatially unifies the local monitoring data and meteorological satellite data, it assigns weights to the data with spatial and temporal labels through dynamic weighting, performs spatial and temporal association through a multimodal Transformer model, and generates a grid fusion feature tensor for the target area. Finally, the perception model outputs a grid dataset and a time series dataset. The rain cloud tracking prediction model processes the grid dataset and the time series dataset and outputs visualization data of rain cloud change data within the target area for no more than 72 hours, as well as the predicted trajectory changes of rain clouds.

[0042] Data cleaning for the perception model primarily involves outlier detection and correction. For example, it removes temperature values ​​above 50°C or below -50°C, and filters out ground clutter below 0 dBZ for radar reflectivity. Machine learning employs a spatiotemporal interpolation algorithm, combining neighboring nodes and historical data to predict missing values. It then uses an isolation forest algorithm to detect outliers and iteratively corrects them using a Kalman filter.

[0043] Spatiotemporal unification refers to the alignment of terminal devices with the Coordinated Universal Time (Coordinated Universal Time) of network data via Beidou / GPS timing. Asynchronous data is synchronized to a unified time grid using linear interpolation. Local device data is then converted to the UTM projection with a conversion error of less than 0.5 meters and resampled to the target area's unified grid using bilinear interpolation.

[0044] Furthermore, the task model processes and obtains the optimal catalytic window T of the rain cloud based on the rain cloud change data for the target area and obtained by the rain cloud tracking and prediction model; then, based on the grid dataset, time series dataset and visualization data of the rain cloud predicted trajectory changes obtained by the perception model, the task model determines the location where the artificial influence operation equipment needs to reach and the preliminary instructions for the corresponding operations. The cloud platform layer sends the instructions to the edge node layer, which processes the instructions to form task instructions corresponding to the terminal devices within the communication range.

[0045] Among them, the task model is trained based on the existing large language model, and historical meteorological data is collected during training. The historical meteorological data includes the grid data set and time series data set output by the perception model, and the visualization data of the rain cloud prediction trajectory changes output by the rain cloud tracking and prediction model. The collected data is cleaned, normalized and feature extracted, and then the actual operation status and effect evaluation data of the corresponding artificial weather modification operation are recorded; the preprocessed data is divided into training set, validation set and test set in proportion. The training set is used for model parameter learning, the validation set is used to adjust the model's hyperparameters, and the test set is used to evaluate the model's final performance; a deep learning model is used to process time series data and spatial features, with historical meteorological data as input and the corresponding optimal catalytic window, operating equipment type, location and operation instructions as output. The parameters of the model are continuously adjusted through the back propagation algorithm to output task instructions.

[0046] Furthermore, when issuing instructions, if the original task cannot be executed due to weather changes or equipment failure, a protection mechanism is set up based on this system to automatically adjust the task instructions using edge nodes to provide redundant alternative solutions. The details are as follows: Based on the original file cloud-edge three-tier architecture and artificial weather modification operation scenarios, the goal is achieved through logical deepening to meet the needs of formulating 72-hour tasks based on device information on the cloud and local adjustment of edge nodes when old equipment suddenly fails.

[0047] Before generating a 72-hour task plan, the cloud performs a full scan of devices in the target area through the device management gateway at the edge node layer, including: The operational status of old equipment is monitored by collecting data such as mechanical wear and ammunition remaining through local sensors on edge nodes, and the status monitoring sensors installed on the equipment provide operational status data.

[0048] The core parameter library of the equipment, which includes data such as maximum range, single-shot catalyst dosage, and collaborative response delay of new equipment, is fused by the perception model to generate an equipment capability matrix. The matrix includes equipment type, location, and substitutability weight.

[0049] When the cloud-based task model generates a plan based on rain cloud prediction data, it must include the main task instructions and alternative adjustment parameters, where the alternative parameters are associated with the performance data of alternative devices in the device capability matrix.

[0050] (1) Fault detection and triggering conditions The data processing station at the edge node monitors the status of legacy equipment in real time through the following methods: For old equipment that requires manual operation, operation feedback signals are collected through LoRa wireless communication, such as launch button trigger feedback, ammunition loading completion signal, etc. If the preset signal is not received for 30 consecutive seconds, it is judged as unable to operate.

[0051] Combined with the lightweight model of the local decision-making module, it analyzes the historical failure data of the equipment. For example, if the failure rate of a certain type of rocket frame rises sharply below -5°C, it can provide an early warning of potential failures 10 minutes in advance.

[0052] Local adjustment is triggered when the following conditions are met: the faulty device is the core execution unit of the current task; the network quality monitoring module of the edge node determines that the cloud communication is interrupted or the response delay exceeds 1 minute; the remaining time to the optimal catalytic window T is less than 30 minutes.

[0053] The local decision module of the edge node starts the adjustment logic, calls the equipment coordination strategy library generated by historical plan training based on the original file task model, selects alternative equipment from the equipment capability matrix, such as replacing a faulty rocket rack with two ground smoke stoves, and calculates the replacement coefficient, such as the sowing amount of one smoke stove in one hour = the catalytic effect of 0.3 rockets, to ensure that the total catalyst amount meets the standard.

[0054] At the same time, perform parameter correction and adjust the operating parameters of the remaining equipment, including: Powder dosage: If the original single-shot dosage of the faulty rocket rack is 300g, and it is replaced with two smoke furnaces, the single-shot dosage of each smoke furnace will be increased to 200g, and the total dosage will be 300g × 1.3 times the compensation coefficient; Position offset: Based on the real-time trajectory of rain clouds, the operating position of the replacement equipment is offset by 1-2 km in the direction of the rain cloud movement; Collaborative scheduling: Send collaborative instructions to peripheral devices through the device management gateway and synchronously correct the device startup timing.

[0055] The edge node collects micro-meteorological data after the operation in real time through local meteorological monitoring equipment, and compares it with the preset effect threshold. For example, if the cloud top temperature drops by ≥2°C 10 minutes after catalysis, if it does not meet the standard, a secondary adjustment will be initiated to further increase the amount of medicine released by adjacent equipment until the core goal of the overall mission is met, that is, the rain-enhancing efficiency remains above 85% of the original plan.

[0056] A specific embodiment of the present invention takes a city marathon event support scenario as an example to demonstrate how the system can realize the generation and coordinated execution of rain removal operation instructions within 72 hours.

[0057] The cloud platform layer serves as the core control center and synchronously accesses multi-source meteorological data through data channels. On the one hand, it collects data from local monitoring equipment such as micro-meteorological stations deployed at the terminal device layer that obtain real-time temperature, humidity, wind speed and direction, and millimeter-wave radars that monitor reflectivity and vertical liquid water content in real time. On the other hand, it obtains the FY-4B satellite cloud images and numerical forecast grid data of the National Meteorological Administration through the network interface.

[0058] The perception model preprocesses both types of data, first using image recognition algorithms to extract key features such as cloud boundaries and TBB gradients from satellite cloud images. Local radar reflectivity data is then fused with satellite TBB data using dynamic weighting. Radar data is assigned a weight of 0.6 due to its high spatiotemporal resolution, while satellite data is assigned a weight of 0.4 due to its wide coverage. The fused data is processed by a multimodal Transformer model to generate a gridded fusion feature tensor with a resolution of 200 meters, containing 12 meteorological parameters such as temperature, humidity, and reflectivity. This ultimately forms a gridded dataset and a time series dataset for subsequent modeling.

[0059] The rain cloud tracking prediction model, based on the output data from the perception model, first processes the fused meteorological data using U-Net+ and a network to accurately segment the rain cloud outline. Using a random forest algorithm, the model identifies the cloud type as cumulonimbus based on features such as a VIL value of 12 kg / m² and a TBB value of 225 K. Using a Kalman filter algorithm to dynamically track its movement, the model determines that the rain cloud is moving southeastward at a speed of 15 m / s.

[0060] The model further combined localized WRF numerical simulation with an LSTM neural network to comprehensively analyze the thermodynamic and kinetic characteristics of clouds, predicting that rain clouds would pass over the stadium between 10:00 and 12:00 PM on the day of the event. Real-time monitoring and analysis of microphysical parameters such as cloud top height and ice crystal concentration determined that the optimal catalytic window was 1-2 hours before the arrival of rain clouds, from 8:00 to 9:00 AM. During this time, cloud top heights stabilize at 4-5 km, and ice crystal concentrations can be effectively increased to 5 crystals per liter through silver iodide seeding, providing ideal physical conditions for rain suppression operations.

[0061] Based on these predictions, the task model first extracted similar scenarios from historical data, including rain suppression operations for similar events in 2023. After preprocessing through cleaning and normalization, this data was fed into a fine-tuned large language model. When generating task plans, the model fully considered equipment capabilities and safety constraints, such as the prohibition of rocket launch equipment within 5 km of the stadium and the 60-minute single-flight endurance limit for drones.

[0062] Based on this, preliminary operation instructions were generated. Three rocket launchers were deployed 50km upstream of the stadium. Each launcher was equipped with 20 silver iodide rockets. The catalyst dosage of each rocket was 300g. They were launched in batches between 8:30 and 9:00 to consume the water vapor in the cloud in advance. At the same time, five drones with a maximum load of 10kg were dispatched to spread salt particles at an altitude of 3km from the ground in the middle of the cloud according to a preset route. The particle size of the salt particles was controlled at 60 microns to inhibit the merging process of raindrops.

[0063] Taking into account the possible 10% fluctuation in the movement speed of rain clouds, the mission model automatically deploys two drones as redundant equipment at alternative operation points based on historical operation experience and real-time data uncertainty analysis, and sets dynamic trigger conditions. When the main equipment fails or the movement speed of rain clouds suddenly increases, the redundant equipment will automatically start to perform supplementary seeding tasks to ensure the stability and effectiveness of rain elimination operations.

[0064] The edge node layer deploys node servers with the administrative district where the event is located as the smallest meteorological management unit, realizes high-speed data interaction with the cloud platform through the wired fiber optic network, and connects with terminal devices with the help of 5G communication network and LoRa wireless communication technology to build a communication network covering the entire target area.

[0065] Furthermore, based on a mechanism for adjusting for equipment failures, this implementation utilizes a cloud-based solution where two older rocket mounts in Area A each launch three rockets daily between 8:00 AM and 9:00 AM, complementing a single drone in Area B for seeding. On the second day of the mission, the edge node detected a lack of launch feedback from R1, declaring it faulty. At this point, 25 minutes remained before the optimal window T, and cloud-based communications were interrupted by heavy rain. The edge node activated its policy library and calculated that one drone, R2, in Area B, could replace R1 and R2. It instructed R2 to increase the charge per shot from 300g to 450g and the number of launches from three to four. It also instructed the drone to shift its seeding route 1.5km toward Area A, covering R1's original operating area. It also simultaneously activated a backup smoke stove in Area C, commencing seeding 10 minutes earlier to compensate for the time difference.

[0066] Effect verification: Through local radar monitoring, the cloud top temperature drop in the area after adjustment deviated from the original plan by no more than 1.2℃, and the rain elimination effect met the standard.

[0067] The node server's built-in data processing station locally parses and caches the operation instructions issued from the cloud, and performs secondary verification and optimization based on local real-time meteorological data to ensure the timeliness and accuracy of the instructions. The device management gateway is responsible for protocol conversion between different types of terminal devices, enabling operational equipment such as rocket launchers and drones to accurately receive and execute instructions. It also transmits device status and operation data in real time, forming a complete closed loop from data collection and instruction generation to operation execution, providing reliable meteorological support.

[0068] The present invention is not limited to the above optional embodiments. Anyone can derive various other forms of products based on the teachings of the present invention. The above specific embodiments should not be construed as limiting the scope of protection of the present invention. The scope of protection of the present invention shall be based on the scope defined in the claims, and the description can be used to interpret the claims.

Claims

1. A distributed weather modification operation collaborative instruction issuing system based on the Internet of Things (IoT) that uniformly issues instructions for weather modification operations within a specific target area. Its features include: include: The cloud platform layer, serving as the data control center, includes a perception model that integrates and processes multi-source meteorological data for the target area through data channels, a rain cloud tracking and prediction model, and a task model based on a large language model. The terminal equipment layer, serving as the underlying data collection and execution end, includes meteorological monitoring equipment, artificial influence operation equipment, and auxiliary equipment; as well as The edge node layer, as the middle layer of the distributed architecture, is deployed in a regional operation center or edge computing server and includes a data processing station and a device management gateway. The data processing station processes and caches instructions from the cloud platform layer and data from the terminal device layer within the management scope, and the device management gateway distributes instructions and performs protocol conversion on the data. The edge node layer also includes a local decision module, which processes the data obtained from the terminal device layer through the local decision module equipped with a locally deployed and trained large language model to obtain local weather status data and upload it to the cloud platform layer. The preliminary instructions issued by the cloud platform layer are processed. If the preliminary instructions from the cloud platform layer are not obtained within the set rain cloud window time, the edge node layer processes the latest preliminary instructions based on the obtained meteorological status data to form task instructions and sends them to the corresponding terminal device layer.

2. The distributed weather modification operation collaborative instruction issuing system based on the Internet of Things according to claim 1 is characterized by: The edge node layer includes the smallest meteorological management unit within the determined target area as a node, and is equipped with a node server. It interacts with the server of the cloud platform layer through a wired communication network and interacts with the terminal device through a wireless IoT communication network.

3. The distributed weather modification operation collaborative instruction issuing system based on the Internet of Things according to claim 1 is characterized by: The meteorological monitoring equipment includes weather radar, anemometer, rain gauge, thermometer, hygrometer, barometer, and sounding balloon; The artificial influence operation equipment includes ground launching equipment and aerial spreading equipment.

4. The distributed weather modification operation collaborative instruction issuing system based on the Internet of Things according to claim 1 is characterized by: The perception model obtains multi-source meteorological data for the target area through an online data channel, and performs data fusion with weight distribution after processing the multi-source data to form operational reference data for the target area.

5. The distributed weather modification operation collaborative instruction issuing system based on the Internet of Things according to claim 4 is characterized by: The online data acquired by the perception model includes local monitoring data acquired by several meteorological detection devices at the terminal device layer, and meteorological satellite images acquired through the access network. The perception model identifies and processes the acquired meteorological satellite images through image processing algorithms and acquires meteorological satellite data. After the perception model cleans and spatially unifies the local monitoring data and meteorological satellite data, it assigns weights to the data with spatial and temporal labels through dynamic weighting, performs spatial and temporal association through a multimodal Transformer model, and generates a grid fusion feature tensor for the target area. Finally, the perception model outputs a grid dataset and a time series dataset, which are then processed by the rain cloud tracking and prediction model to output rain cloud change data for no more than 72 hours in the target area and visualization data of the predicted trajectory changes of rain clouds.

6. The distributed weather modification operation collaborative instruction issuing system based on the Internet of Things according to claim 5 is characterized by: The task model processes and obtains the optimal catalytic window T of rain clouds based on the rain cloud change data for the target area and obtained by the rain cloud tracking and prediction model; Then, based on the grid dataset, time series dataset, and visualization data of rain cloud predicted trajectory changes obtained by the perception model, the location where the artificial influence operation equipment needs to reach and the preliminary instructions for the corresponding operations are determined. The cloud platform layer sends the preliminary instructions to the edge node layer, which processes the preliminary instructions and forms task instructions corresponding to the terminal devices within the communication range.

7. The distributed weather modification operation collaborative instruction issuing system based on the Internet of Things according to claim 6 is characterized by: The task model is trained based on an existing large language model. During training, historical meteorological data is collected. The historical meteorological data includes grid datasets and time series datasets output by the perception model, as well as visualization data of rain cloud trajectory changes output by the rain cloud tracking prediction model. The collected data is cleaned, normalized, and features extracted. The actual operation status and effect evaluation data of the corresponding artificial weather modification operation are then recorded. The preprocessed data is divided into training set, validation set and test set. The training set is used for model parameter learning, the validation set is used to adjust the model's hyperparameters, and the test set is used to evaluate the model performance. A deep learning model is used to process time series data and spatial features, with historical meteorological data as input and the corresponding optimal catalytic window, operating equipment type, location and operating instructions as output. The model parameters are continuously adjusted through the backpropagation algorithm to output task instructions.

8. The distributed weather modification operation collaborative instruction issuing system based on the Internet of Things according to claim 6 is characterized by: When generating preliminary instructions, the task model first obtains device information from a global scan of devices in the target area by the cloud platform through the device management gateway of the edge node layer. The task model generates a device capability matrix based on the device information. The preliminary instructions generated by the task model include alternative adjustment parameters, which are associated with performance data of alternative devices in the device capability matrix. When a terminal device fails during the execution of a task instruction, the local decision module of the edge node starts the adjustment logic, calls the pre-stored alternative adjustment parameters, selects alternative devices from the device capability matrix, and calculates the replacement coefficient to correct the parameters, keeping the total catalyst amount change value less than the set threshold, and sends collaborative instructions to the surrounding devices through the device management gateway to synchronously correct the device startup timing.

Images