Low-altitude meteorological early warning method and system based on Internet of Things

Abstract

The invention discloses a low-altitude meteorological early warning method and system based on the Internet of Things, and relates to the technical field of low-altitude flight traffic control, and the low-altitude meteorological early warning method comprises the steps: arranging a plurality of Internet of Things sensor devices in a low-altitude flight region, and collecting meteorological data in real time, including temperature, humidity, wind speed, air pressure and the like; uploading the collected meteorological data to a cloud server by using a data transmission network; performing real-time analysis and prediction on a low-altitude meteorological condition in combination with a meteorological prediction model and historical big data; and when abnormal or potential dangerous meteorological conditions are detected, early warning information is sent to related users through multiple communication modes. According to the invention, the accuracy and timeliness of low-altitude meteorological early warning are improved, and the safety of air traffic in low-altitude flight is effectively guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of low-altitude flight traffic control, and in particular to a low-altitude meteorological early warning method and system based on the Internet of Things. Background Art

[0002] With the widespread use of low-altitude aircraft such as drones, low-altitude flight areas are gradually opened. In order to ensure the safety of low-altitude flight, especially to avoid flight accidents caused by weather factors, the Internet of Things technology is used to perceive real-time environmental changes in low-altitude areas, monitor flight status, and transmit and process data to help low-altitude aircraft effectively respond to meteorological changes. At present, the low-altitude meteorological early warning system based on the Internet of Things is difficult to respond in time and issue graded early warnings according to the interference caused by meteorological changes in different periods, and it is difficult to plan the optimal flight path and avoid the safety hazards caused by meteorological influences.

[0003] To sum up, how to respond in time and give graded warnings of potential safety hazards brought about by meteorological changes in order to plan the optimal flight path to avoid flight risks is an issue that needs to be urgently solved and optimized in the low-altitude meteorological warning system based on the Internet of Things. Summary of the invention

[0004] The present invention provides a low-altitude meteorological early warning method and system based on the Internet of Things, which solves the technical problem of how to timely respond to and grade early warning of safety hazards caused by meteorological changes, so as to plan the optimal flight path to avoid flight risks.

[0005] In order to solve the above technical problems, the present invention provides a low-altitude meteorological early warning method and system based on the Internet of Things. The specific technical solutions are as follows:

[0006] In the first aspect, a low-altitude meteorological early warning method based on the Internet of Things includes:

[0007] S100, collecting a variety of historical meteorological data to obtain a meteorological data set; obtaining a variety of environmental parameters corresponding to each meteorological data item according to the meteorological data set, and assembling the environmental parameter set; obtaining the low-altitude flight space coordinates of the current time node to obtain flight coordinate data;

[0008] S200, constructing and training the meteorological data set and the environmental parameter set to obtain a meteorological prediction model;

[0009] S300, based on the meteorological prediction model, real-time acquisition of low-altitude flight meteorological data environment parameters at the current time node, inputting the meteorological data environment parameters at multiple time nodes into the meteorological prediction model to output a meteorological data prediction result representing the next time node;

[0010] S400, judging the meteorological change trend according to the meteorological data prediction result of the meteorological prediction model; setting multiple change trend warning thresholds, responding to the preset warning thresholds according to different meteorological change trends, and outputting warning information;

[0011] A variety of low-altitude flight strategies are planned according to meteorological changes; the low-altitude flight strategy obtains flight path coordinate nodes according to warning information; by adding the original coordinate node to the open list and continuously expanding the neighboring nodes, the path cost and estimated value are updated in real time until the target node is found or the open list is empty, that is, the safe coordinate node is obtained.

[0012] As a further optimization scheme of the present invention, low-altitude flight strategies corresponding to various meteorological changes are planned according to various meteorological changes; based on the low-altitude flight strategies, the low-altitude flight strategies are timely adjusted according to the early warning information of various meteorological change trends.

[0013] As a further optimization scheme of the present invention, the weather forecast model includes:

[0014] The meteorological data set and the environmental parameter set are used to construct a data set to generate structured data, and the structured data is encoded into sequence data to train the meteorological prediction model;

[0015] The sequence data is input into the weather forecast model; the weather forecast model comprises an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output layer, the intermediate representation data of multiple hidden layers are transmitted to the output layer, and the output layer outputs the weather data recognition results corresponding to the multiple environmental parameters;

[0016] At least two environmental parameter data items at the current time node are input into the meteorological prediction model, and the output layer output of the meteorological prediction model represents the current meteorological data prediction result.

[0017] As a further optimization scheme of the present invention, judging the meteorological change trend according to the meteorological data prediction results of the meteorological prediction model includes:

[0018] According to the meteorological data prediction results, meteorological data and environmental parameters of multiple time nodes are obtained; based on the environmental parameters of meteorological data at different time nodes, Obtain the change rates of various environmental parameters at different time nodes;

[0019] In the formula, H represents the change rate between adjacent time nodes, and Respectively represent the time node t i and t i+1 At least one environment parameter variable value A, represents the time difference between the i-th and i+1-th adjacent time nodes, and t i+1 Indicates t i The time node after the time node;

[0020] Obtain the change rates between at least three different adjacent time nodes to obtain a change rate control group; compare at least two of the change rate control groups to obtain a comparison result; and judge the meteorological change trend based on the existence of a plurality of different comparison results in different change rate control groups.

[0021] As a further optimization scheme of the present invention, multiple change trend warning thresholds are set, and according to different meteorological change trends, the preset warning thresholds are responded to and warning information is output, including:

[0022] Based on various different comparison results of different change rate control groups, various change trend warning mechanisms are set; the various change trend warning mechanisms set change trend warning threshold ranges [G i , G j ], when the change rate of various environmental parameters at different time nodes of meteorological data changes significantly within the warning threshold range, the first meteorological warning information is output;

[0023] The first meteorological warning information includes upward trend warning information and downward trend warning information; when the comparison result of the change rate control group of different environmental parameters is in an obvious upward trend, the change rate is gradually increasing to output the upward trend warning information; when the comparison result of the change rate control group of different environmental parameters is in an obvious downward trend, the change rate is gradually decreasing to output the downward trend warning information.

[0024] As a further optimization solution of the present invention, the warning information further includes:

[0025] Based on the meteorological data set and the environmental parameter set, the environmental parameter characteristics of different meteorological data are obtained; the environmental parameters include air pressure, temperature, humidity, wind speed and direction, precipitation, visibility, cloud height and thunderstorm activity; characteristic properties of environmental parameters existing in different meteorological changes are obtained through multiple sensors;

[0026] The first meteorological warning information monitors the rapid changes in environmental parameters to determine whether an extreme meteorological event is likely to occur; first, a warning threshold range is set, which is usually based on the changes in meteorological data and obvious fluctuations in environmental parameters over a period of time;

[0027] Based on the first meteorological warning information, the impact results of the increase or decrease of the environmental parameters that have changed significantly within the warning threshold range are summarized to obtain the final meteorological change trend information, so as to output the second meteorological warning information; the impact results of the increase or decrease of the environmental parameters represent different meteorological changes caused by various environmental parameters with different change trends, and the impacts on low-altitude flights are different accordingly;

[0028] The second meteorological warning information includes warning information for improved meteorological changes and warning information for worse meteorological changes; when the change trend of each environmental parameter at the same time node meets the environmental parameter characteristics of improved meteorological changes, the warning information for improved meteorological changes is output; when the change trend of each environmental parameter at the same time node meets the environmental parameter characteristics of worse meteorological changes, the warning information for worse meteorological changes is output;

[0029] Compare the predicted meteorological change trend results with the actual meteorological changes to obtain the comparison results; and according to the comparison results, The meteorological change trend results predicted by calibration are: MSE represents the mean square error calibration algorithm, N represents the type of meteorological data, represents the predicted value; Represents the true value.

[0030] As a further optimization scheme of the present invention, according to various weather changes, low-altitude flight strategies corresponding to various weather changes are planned, including:

[0031] Based on the flight coordinate data, obtain the low-altitude flight coordinate position of the current node; according to the meteorological data, set a variety of low-altitude flight strategies;

[0032] The low-altitude flight strategy includes: in normal weather conditions, the low-altitude flight is smoothly flown along a preset path; when abnormal weather conditions occur, the environmental parameters based on the abnormal weather data are To obtain the estimated values ​​of the coordinate nodes involved in the abnormal weather flight path;

[0033] In the formula, f(x) represents the estimated value of the planned path coordinate node, g(x) represents the distance cost from the original coordinate node to the current coordinate node, h(x) represents the distance cost from the current coordinate node to the target coordinate node, and x represents a coordinate node in the path planning.

[0034] As a further optimization scheme of the present invention, the original coordinate node is added to the open list, and g(x)=0 is set; h(s) is the heuristic estimation value of the original coordinate node, where s represents the original coordinate node;

[0035] Select the node with the smallest distance from the open list, which is the current coordinate node S; expand each neighbor node S' of S to obtain g(S') and h(S'), where g(S') represents the update distance cost from the original coordinate node to the current coordinate node, and h(S') is the estimated value of the real-time update distance coordinate node;

[0036] If the neighbor node S' has not been added to the open list, add it to the open list; if the neighbor node S' is already in the list, the obtained path distance cost is smaller, and the path cost value is updated in real time; move the current coordinate node S to the closed list to avoid repeated expansion;

[0037] Repeat the operation until the target coordinate node is found or the open list is empty.

[0038] As a further optimization scheme of the present invention, based on the low-altitude flight strategy, according to the early warning information of various meteorological change trends, the low-altitude flight strategy is timely adjusted, including:

[0039] Based on the low-altitude flight strategy, To obtain the low-altitude flight yaw scheduling coordinates; the low-altitude flight yaw scheduling coordinates are the target coordinates for regulating the low-altitude flight strategy;

[0040] Where C(path) is the coordinate point of the low-altitude flight path after yaw scheduling, x i ,y i 、z i They represent the control parameters of the current low-altitude flight, which include the coordinate position, flight direction and flight angle of the current low-altitude flight, T i represents the flight time, cost represents the flight risk or energy consumption cost based on weather changes, and n represents the number of low-altitude flight control parameters;

[0041] Based on the low-altitude flight yaw scheduling coordinates, a plurality of the low-altitude flight yaw scheduling coordinates are obtained to obtain a low-altitude flight route after regulation, The flight strategy is gradually optimized to obtain the optimal flight path strategy; the interference of weather changes is avoided according to the optimal flight path strategy; where V(t) represents the optimal strategy cost at time t, Indicates from time The flight cost to t.

[0042] In a second aspect, the system is provided with an electronic device including a memory, a processor, and a low-altitude meteorological early warning method program based on the Internet of Things stored in the memory and executable on the processor, wherein the low-altitude meteorological early warning method program based on the Internet of Things implements the steps of a low-altitude meteorological early warning method based on the Internet of Things when executed by the processor, and the system includes:

[0043] Data collection module: It is used to collect a variety of historical meteorological data to obtain a meteorological data set; according to the meteorological data set, to obtain a variety of environmental parameters corresponding to each meteorological data item, and to compile an environmental parameter set; to obtain the low-altitude flight space coordinates of the current time node to obtain flight coordinate data;

[0044] Model training module: used to construct and train the meteorological data set and the environmental parameter set to obtain a meteorological prediction model; based on the meteorological prediction model, various environmental parameters in the low-altitude flight meteorological data at the current time node are obtained in real time, and the meteorological data environmental parameters at multiple time nodes are input into the meteorological prediction model to output the meteorological data prediction result representing the next time node;

[0045] Early warning response module: It is used to judge the meteorological change trend according to the meteorological data prediction results of the meteorological forecast model; set multiple change trend early warning thresholds, respond to the preset early warning thresholds according to different meteorological change trends, and output early warning information.

[0046] The present invention has at least the following beneficial effects: the present invention collects a variety of historical meteorological data and constructs a meteorological data set; by collecting a variety of historical meteorological data, the diversity and comprehensiveness of data sources are ensured. These data include various meteorological factors such as air pressure, temperature, humidity, wind speed and direction, precipitation, visibility, cloud height, and thunderstorm activity, providing rich information for subsequent modeling; the construction of the meteorological data set provides high-quality data input for subsequent steps, can provide accurate historical data support for the prediction model, and reduce errors; by integrating a variety of meteorological data items and environmental parameters, it is possible to comprehensively evaluate the impact of meteorological conditions on low-altitude flights and improve the accuracy of prediction results.

[0047] By building a weather forecast model; based on training of historical meteorological data sets and environmental parameter sets, a special weather forecast model can be built according to the actual flight environment and needs, and the weather changes in the low-altitude flight environment can be accurately predicted; by associating meteorological data and environmental parameters, the model can not only predict common weather changes, but also take into account the special impact of the flight environment on the weather (such as terrain, flight trajectory and flight time), thereby improving the prediction accuracy; by training the model with advanced algorithms (such as machine learning or deep learning), the complex patterns in historical data can be automatically captured, thereby enhancing the model's prediction ability.

[0048] Then, the low-altitude flight meteorological data is obtained in real time and predicted. Based on the trained meteorological prediction model, the low-altitude flight meteorological data of the current time node can be obtained in real time, and future meteorological conditions can be dynamically predicted, thereby providing immediate decision support. By inputting data from multiple time nodes into the model, the trend of meteorological changes can be captured, rather than predictions at a single time point, providing more comprehensive trend information. With the input of real-time meteorological data, the model can continuously adjust the prediction during the flight to ensure the timeliness and accuracy of the meteorological data and avoid the errors that may be caused by static predictions.

[0049] According to the meteorological forecast results, the meteorological change trend is judged and the warning information is output. By judging the meteorological change trend through the meteorological forecast results, potential severe weather or meteorological changes (such as strong winds, heavy rains, and rain and snow weather) can be warned in advance to reduce flight safety risks. According to different meteorological change trends and set warning thresholds, multi-level warning responses can be achieved to provide targeted response measures. For example, a higher level of response is taken for sudden meteorological changes, while a lower level of response is taken for minor changes. The analysis of meteorological change trends and the output of warning information provide scientific decision-making support for flight personnel or dispatch centers, helping them adjust flight paths, control flight altitudes, or take other necessary safety measures. By accurately grasping the meteorological change trend, safety warnings can be made in real time to prevent the potential impact of weather on low-altitude flights, significantly improving flight safety.

[0050] Through the coordination and cooperation between various steps, effective support is provided for weather forecasting and decision-making during low-altitude flight. From data collection, model training to real-time forecasting and warning output, the entire process builds a closed-loop weather forecasting system to provide more accurate and real-time weather information for flight missions. By identifying weather change trends in advance and issuing warnings, flight safety is enhanced. Real-time weather data and dynamic forecasts can provide flight personnel with real-time and scientific decision-making basis to reduce risks during flight. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a schematic flow chart of a low-altitude meteorological early warning method based on the Internet of Things provided by an embodiment of the present invention;

[0052] Figure 2 It is a schematic diagram of a low-altitude meteorological early warning system based on the Internet of Things provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The present application is further described in detail below in conjunction with the accompanying drawings. It is necessary to point out here that the following specific implementation methods are only used to further illustrate the present application and cannot be understood as limiting the scope of protection of the present application. Technical personnel in this field can make some non-essential improvements and adjustments to the present application based on the above application content.

[0054] This embodiment provides a low-altitude meteorological early warning method and system based on the Internet of Things, and the specific implementation methods are as follows:

[0055] like Figure 1 As shown, a low-altitude meteorological early warning method based on the Internet of Things includes the following steps:

[0056] S100, collecting a variety of historical meteorological data to obtain a meteorological data set; obtaining a variety of environmental parameters corresponding to each meteorological data item according to the meteorological data set, and assembling the environmental parameter set; obtaining the low-altitude flight space coordinates of the current time node to obtain flight coordinate data;

[0057] S200, constructing and training the meteorological data set and the environmental parameter set to obtain a meteorological prediction model;

[0058] S300, based on the meteorological prediction model, real-time acquisition of low-altitude flight meteorological data environment parameters at the current time node, inputting the meteorological data environment parameters at multiple time nodes into the meteorological prediction model to output a meteorological data prediction result representing the next time node;

[0059] S400, judging the meteorological change trend according to the meteorological data prediction results of the meteorological prediction model; setting multiple change trend warning thresholds, responding to the preset warning thresholds according to different meteorological change trends, and outputting warning information.

[0060] In the implementation of the present invention, historical meteorological data from various sources are collected based on step S100, including air pressure, temperature, humidity, wind speed and direction, precipitation, visibility, cloud height, and basic meteorological elements of thunderstorm activity. The data may come from a variety of channels such as meteorological stations, satellite remote sensing, drone monitoring, or various related environmental parameter sensors (lidar sensors to monitor wind speed and direction, temperature sensors, humidity sensors, and air pressure sensors); for example, in clear weather, the typical characteristics of various environmental parameters can be summarized as follows:

[0061] Air Pressure: Usually high and stable.

[0062] Temperature: Higher during the day and lower at night, with a large temperature difference between day and night.

[0063] Humidity: Low.

[0064] Wind speed: Low, steady.

[0065] Precipitation: Almost zero.

[0066] Visibility: Very good, usually maximum.

[0067] Cloud Height: No clouds or only thin clouds at high altitude.

[0068] Thunderstorm activity: None or extremely rare.

[0069] On rainy days, meteorological parameters usually present the following characteristics:

[0070] Air Pressure: Low, usually in low pressure areas or frontal systems.

[0071] Temperature: During rainfall, temperatures vary less but may drop, especially behind a cold front.

[0072] Humidity: Humidity has risen significantly, approaching 100%.

[0073] Wind speed: The wind speed is relatively high and may increase dramatically during heavy rain.

[0074] Wind Direction: Wind direction varies greatly and is usually affected by fronts or low pressure areas.

[0075] Precipitation: Precipitation is high and may range from light rain to heavy rain.

[0076] Visibility: Visibility is reduced. The heavier the precipitation, the worse the visibility.

[0077] Cloud height: Clouds are low and dense, with low-level clouds and cumulonimbus clouds being common.

[0078] Thunderstorm activity: Thunderstorm activity is common in summer, especially during heavy precipitation or rainstorms; it is occasional or absent in winter.

[0079] On snowy days, meteorological parameters usually appear as follows:

[0080] Air Pressure: Low, usually under the influence of a low pressure area or frontal system.

[0081] Temperature: The temperature is low, usually below 0°C. The lower the temperature, the more intensive the snowfall.

[0082] Humidity: Humidity is high, but slightly lower than on rainy days because the air contains less water vapor at low temperatures.

[0083] Wind Speed: Wind speeds may be high, especially during a snowstorm or cold snap.

[0084] Wind direction: Wind direction often changes, affected by low pressure systems and fronts, especially during storms, when wind direction changes dramatically.

[0085] Precipitation: Precipitation is usually light, mainly snow, and the amount of snowfall is affected by temperature, humidity and wind speed.

[0086] Visibility: Visibility is poor, especially in heavy snow, blizzards or blizzards, when visibility may be significantly reduced.

[0087] Cloud Height: Clouds are low and dense, with a generally low cloud base.

[0088] Thunderstorm activity: Thunderstorm activity is rare, but it occasionally occurs during special weather conditions (such as winter thunderstorms).

[0089] On haze days, meteorological parameters usually show:

[0090] Air pressure: The air pressure is high. Usually under high air pressure control, the air does not flow, resulting in the inability of pollutants to spread.

[0091] Temperature: Temperature changes are small and are usually low, especially in winter when inversion effects may lead to lower temperatures in the lower atmosphere.

[0092] Humidity: When humidity is high, moisture combines with pollutants to form haze particles, which further affects visibility.

[0093] Wind speed: Low wind speed and poor air mobility cause pollutants to accumulate near the ground and are difficult to disperse.

[0094] Wind direction: The wind direction is stable, the wind force is weak, the air flow is not smooth, and the haze is easy to maintain.

[0095] Precipitation: Little or no precipitation. Haze precipitation may occur when humidity is high, but it cannot effectively remove pollutants.

[0096] Visibility: Visibility is poor. Haze makes visibility less than 1 km, and in severe cases it drops to several hundred meters.

[0097] Cloud height: The clouds are low and dense, usually stratus, with the cloud base close to the ground. Haze and low clouds together affect visibility.

[0098] Thunderstorm activity: Thunderstorm activity is extremely rare. Haze days are usually accompanied by stable atmospheric conditions, lacking the conditions for thunderstorm formation.

[0099] According to the collected meteorological data (clear weather, rainy days, snowy days, foggy days and daytime periods), the associated environmental parameters are obtained and the relevant environmental parameters involved in the relevant meteorological data are preprocessed and normalized to obtain data of uniform dimension. The unified dimensional data is then used to form an environmental parameter set through deep learning model rules or multi-task learning algorithms or induction as the background input for meteorological forecasting.

[0100] According to the mission requirements of the current time node, collect the relevant spatial coordinate data of low-altitude flight, such as the coordinates of the drone route or the three-dimensional spatial range of the flight area. This part of the data is used to combine with meteorological data to locate the low-altitude flight meteorological environment.

[0101] S200 inputs the collected meteorological data and environmental parameter sets into the data model training process. This process includes data cleaning, normalization and feature extraction to ensure the validity and consistency of the model input data; using machine learning or deep learning algorithms (such as regression analysis, time series prediction or convolutional neural networks) to build a meteorological forecast model. By learning from historical meteorological data and environmental parameters, the model can extract the laws of weather changes.

[0102] S300 uses the real-time meteorological data and environmental parameters (such as current temperature, humidity, wind speed, etc.) of the current time node, combined with the coordinate data of low-altitude flight, as model input; the meteorological data of multiple time nodes are input into the prediction model, and the meteorological forecast results of the next time node are gradually calculated, such as the wind speed or rainfall in the next 10 minutes, etc. This recursive prediction method can provide real-time updated meteorological information for flight path planning.

[0103] S400 compares the changes in meteorological parameters at different time points based on the meteorological data forecast results output by the model, and identifies meteorological trends, such as sudden temperature drops, increased wind speeds, or air pressure fluctuations. It sets warning thresholds for different meteorological change trends. For example, when the wind speed exceeds a certain critical value or visibility is lower than a certain standard, a corresponding warning is triggered. When the predicted meteorological data reaches the set warning threshold, the system automatically issues a warning message (such as an alarm sound, graphic prompt, or text prompt) to alert relevant personnel and provide response suggestions.

[0104] The above steps work together to improve the accuracy of weather forecasts, especially the prediction of local weather changes in low-altitude flight environments, through multi-dimensional data input and machine learning models; predict and output future weather trends based on real-time data to provide real-time weather support for low-altitude aircraft such as drones; through the analysis and early warning functions of weather change trends, the impact of bad weather on low-altitude flight missions can be effectively avoided, thereby improving flight safety; using a variety of historical meteorological data and environmental parameters, the system can adapt to different terrain and climatic conditions and is widely used in a variety of low-altitude flight missions, such as express drones, agricultural spraying or post-disaster rescue; the early warning mechanism helps pilots or drone operators make quick decisions, such as adjusting routes or terminating missions, which helps save resources and avoid risks.

[0105] In another preferred implementation of the present invention, low-altitude flight strategies corresponding to various meteorological changes are planned according to various meteorological changes; based on the low-altitude flight strategy, the low-altitude flight strategy is adjusted in time according to the early warning information of various meteorological change trends.

[0106] In the implementation of the present invention, first, based on historical meteorological data, real-time meteorological data and environmental parameters, various meteorological changes that may affect low-altitude flight are identified. Common meteorological changes include increased wind speed, rain and snow, haze, and changes in air pressure. Corresponding low-altitude flight strategies are formulated for different meteorological changes. For example:

[0107] When the wind speed is high, the flight strategy may be to lower the flight altitude, choose a sheltered area, or adjust the flight path;

[0108] In thunderstorm weather, the flight strategy may be to suspend the flight and continue the mission after the weather stabilizes;

[0109] In low visibility or heavy fog, you may need to enhance the aircraft's automatic navigation capabilities, slow down the flight speed and maintain altitude.

[0110] According to the types of weather changes and the requirements of flight missions, a variety of optional flight strategies are planned, and the strategy library is kept updated and optimized to ensure flexibility in responding to sudden weather changes.

[0111] Based on the low-altitude flight strategy, the flight strategy is adjusted according to the warning information of the weather change trend, and the warning information of the weather change trend is received in real time according to the weather forecast model. This information usually includes possible wind speed changes, precipitation, temperature changes, etc., to predict the weather changes in the next few minutes to hours in advance; the received warning information is classified to determine which weather changes may have a direct impact on the flight mission. Set priorities according to the severity of the impact. For example, thunderstorms and extreme wind speed changes may be set as high priority warnings, while temperature changes may be low priority.

[0112] Based on real-time weather trends and warning information, the system automatically or manually adjusts the flight strategy. If the weather trend shows that the wind speed is about to exceed the safe range, the system will immediately adjust the aircraft's altitude or path, or suspend flight until the weather conditions return to safety. The pilot or operator will also receive adjustment suggestions and reminders to ensure the safety of the flight mission.

[0113] The collaboration between each step is to receive real-time weather data from the meteorological data source, combine the geographical location of the flight mission and the current meteorological conditions, obtain the trend forecast of meteorological changes, and analyze the received meteorological data to identify meteorological changes that may affect low-altitude flights. According to different meteorological changes, the system will automatically plan the corresponding flight strategy to ensure that the flight mission can be successfully completed under different meteorological conditions; during the flight, if meteorological changes occur, the system will adjust the flight strategy in time according to the warning information. For example, if the meteorological change warning shows that the wind speed is increasing, the system will automatically adjust the flight altitude or route according to the planned wind speed strategy; and continuously feedback the execution effect of meteorological changes and flight strategies. If an unexpected situation occurs during the actual flight (such as the warning fails to accurately predict the weather changes), the system will adjust the strategy based on the actual data feedback, and optimize the strategy library to improve the ability to deal with similar situations.

[0114] In a preferred embodiment of the present invention, the weather forecast model in step S200 includes:

[0115] S201, constructing a data set from the meteorological data set and the environmental parameter set to generate structured data, encoding the structured data into sequence data, so as to train and obtain the meteorological prediction model;

[0116] S202, inputting the sequence data into the meteorological prediction model; the meteorological prediction model includes an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output layer, transmitting the intermediate representation data of multiple hidden layers to the output layer, and the output layer outputs the meteorological data recognition results corresponding to the multiple environmental parameters;

[0117] S203, inputting at least two environmental parameter data items of the current time node into the meteorological prediction model, and the output layer output of the meteorological prediction model represents the current meteorological data prediction result.

[0118] In the implementation of the present invention, the collected meteorological data and related environmental parameters are integrated in step S201 to generate a data set. The meteorological data may include but is not limited to air pressure, temperature, humidity, wind speed and direction, precipitation, visibility, cloud height, and thunderstorm activity, while the environmental parameter set may include other factors such as geographic location and time. These data are usually organized in a table or other structured format.

[0119] Convert the organized meteorological data and environmental parameter sets into a structured dataset. Usually this structured dataset contains multiple feature columns, each column represents a data type or environmental parameter, and the rows represent different time nodes or different observation samples.

[0120] The constructed structured data is encoded so that it can be input into the deep learning model. Since meteorological data and environmental parameters usually have time series characteristics, the serialization process may be to deal with time dependency, that is, the correlation between previous and next data on the time axis. This step is converted into an input format suitable for deep learning models through certain encoding methods (for example, numerical standardization, time series windowing, etc.).

[0121] By constructing data sets and structured data, meteorological data and environmental parameters from different sources can be organically combined, so that subsequent model training can process rich and useful features; encoding into sequence data helps capture dependencies in time series and improves the model's prediction accuracy for future meteorological changes.

[0122] In step S202, the sequence data encoded in step S201 is input into a weather forecast model. The model may be a neural network (such as an LSTM or GRU network) containing multiple hidden layers to learn the temporal characteristics of the input data.

[0123] Weather forecast model structure:

[0124] Input layer: receives input data from step S201;

[0125] First, second, and third hidden layers: These hidden layers enable the model to learn complex patterns and time dependencies in the data by weighted combinations of the input data and performing nonlinear transformations;

[0126] Output layer: transfers the intermediate representation data of multiple hidden layers to the output layer and outputs the weather forecast results;

[0127] The intermediate representation is transmitted to the output layer: The output of each hidden layer will be used as the intermediate representation and passed to the final output layer layer by layer. The model will comprehensively consider the learning information of multiple hidden layers and make the final prediction decision.

[0128] Multiple hidden layers can make the model have stronger learning ability and can capture complex relationships in the data, especially the nonlinear relationship between meteorological data and environmental parameters; through the deep network architecture, the model can better infer future meteorological changes when faced with different meteorological and environmental parameters.

[0129] Based on the weather forecast model constructed in step S202, at least two environmental parameter data items of the current time node are input as real-time data into the weather forecast model in step S203. This step is performed after the training is completed and is used for actual weather forecasting.

[0130] Weather forecast result output: The model calculates the corresponding weather forecast result based on the environmental parameter data input at the current time node through the trained hidden layer. This result may be an estimate of weather data such as temperature and precipitation at a certain point in the future; by inputting the current environmental parameters, the model can achieve real-time weather forecasts to help decision makers respond quickly to weather changes; the trained weather forecast model can generate more accurate forecast results based on real-time input data, especially in complex weather environments, where the advantages of deep learning models can be better reflected.

[0131] In a preferred embodiment of the present invention, the step S400 determines the weather change trend according to the weather data prediction result of the weather prediction model, including:

[0132] S401, according to the meteorological data prediction results, to obtain meteorological data and environmental parameters of multiple time nodes; based on the environmental parameters of the meteorological data at different time nodes, Obtain the change rates of various environmental parameters at different time nodes;

[0133] In the formula, H represents the change rate between adjacent time nodes, and Respectively represent the time node t i and t i+1 At least one environment parameter variable value A, represents the time difference between the i-th and i+1-th adjacent time nodes, and t i+1 Indicates t i The time node after the time node;

[0134] S402, obtaining the change rates between at least three different adjacent time nodes to obtain a change rate control group; comparing at least two of the change rate control groups to obtain a comparison result; and judging the meteorological change trend according to the existence of a plurality of different comparison results in different change rate control groups.

[0135] In the implementation of the present invention, in step S401, meteorological data and environmental parameters of multiple time nodes are obtained:

[0136] In step S202, the weather forecast model has generated weather data forecast results for various future time nodes. At this point, the system will extract weather data (such as temperature, humidity, wind speed, etc.) and environmental parameters (such as air pressure, geographic location, etc.) at multiple time nodes from the forecast results.

[0137] Based on the environmental parameters of meteorological data at different time nodes, the time nodes are associated with the corresponding meteorological data and environmental parameters to construct meteorological forecasts and environmental data for each time node. These data can be continuous (e.g., time series of hours, days, weeks) or irregular (e.g., real-time data collected at specific time intervals).

[0138] For meteorological data and environmental parameters between different time nodes, calculate the rate of change between adjacent time nodes. The rate of change indicates the speed of change of meteorological data or environmental parameters, usually the relative change of a value. For example, the temperature change rate can indicate the rate of increase or decrease of temperature between two time points. The rate of change helps to quantify the changing trends of meteorological and environmental data; obtaining the rate of change can effectively quantify the dynamic changes of meteorological parameters and environmental data, and reveal the changing laws of meteorology;

[0139] Through the rate of change, the rate of change of meteorological parameters can be directly observed, which helps to better understand the urgency of meteorological changes, especially when predicting sudden meteorological events; obtaining meteorological data and change rates at multiple time nodes can lay the foundation for subsequent trend analysis and comparison.

[0140] In step S402, for the data of multiple time nodes, the change rate between them is calculated according to each pair of adjacent time nodes. For example, assuming there are time nodes t1, t2, t3, etc., the change rate between t1 and t2 and the change rate between t2 and t3 can be calculated; it is ensured that there are at least three change rate calculation results between different time nodes to ensure that the change trend can be accurately judged.

[0141] The change rates between different time nodes are combined into a "change rate control group", that is, the change rates of each pair of adjacent time nodes are compared. Through the control group, the trend and pattern of meteorological parameter changes between different time periods can be clearly observed. The change rate control group of multiple time nodes will provide rich information for judging the trend of meteorological changes.

[0142] The control group of change rates between multiple time nodes helps to compare meteorological changes from the perspective of multiple time periods, so as to have a more comprehensive understanding of the trend of meteorological changes; through the control group method, the cyclical and non-cyclical fluctuations in meteorological changes can be effectively captured, providing a basis for subsequent predictions or alarm mechanisms.

[0143] Compare at least two control groups of change rates and analyze the differences between them. For example, compare the change rates of meteorological parameters from t1 to t2 and from t2 to t3 to see if the changes are consistent, if there are large fluctuations, or if there is some kind of periodic or regular change.

[0144] When comparing multiple rate of change control groups, you can see different trends in the rate of change of meteorological data. For example, if the rate of change at different time nodes is positive and gradually increases, it may indicate that meteorological parameters such as temperature are continuing to rise. Conversely, if the rate of change fluctuates or decreases in multiple control groups, it may indicate that meteorological changes are stabilizing or decreasing.

[0145] Comparing the change rate control groups at different time nodes can help reveal the long-term or short-term trends of meteorological changes and help predict future meteorological trends. When the changes in one of the change rate control groups are abnormal (for example, the change rate in a certain time period is significantly different from that in other time periods), it is expressed as a meteorological anomaly, such as climate change, extreme weather, etc., and early warning can be given.

[0146] By integrating the comparison results of multiple change rate control groups, the accuracy of meteorological change trends can be enhanced, and scientific decisions can be made for the next forecast; finally, the trend of meteorological changes can be judged based on the results of the comparison. For example, if the trend of the rate of change at multiple time nodes shows a consistent increase or decrease, the model may judge that the meteorological change is a continuous upward or downward trend. If there are significant fluctuations in the rate of change, it may indicate that the meteorological change tends to be unstable or there are drastic changes; judging the trend of meteorological changes should not only pay attention to the rate of change, but also consider factors such as the direction of change (such as temperature increase or decrease), amplitude and periodicity. By integrating the comparison results of multiple change rate control groups, the system can give a more accurate forecast of meteorological trends and help decision makers formulate relevant response measures.

[0147] By judging the trend of meteorological changes, we can predict changes in meteorological conditions in advance and take countermeasures in advance, which is especially important in extreme weather warning and disaster prevention. Accurate judgment of meteorological change trends can help relevant departments and personnel to rationally allocate resources and provide a scientific basis for industries such as agriculture, transportation, and aviation. Combining the judgments of multiple change rate control groups can improve the accuracy of meteorological change predictions, avoid deviations from a single time node or a single change rate, and provide more stable and reliable meteorological change judgments.

[0148] In a preferred embodiment of the present invention, a plurality of change trend warning thresholds are set in step S400, and according to different meteorological change trends, the preset warning thresholds are responded to and warning information is output, including:

[0149] Step S403, based on the various comparison results of the different change rate control groups, multiple change trend warning mechanisms are set; the multiple change trend warning mechanisms set the change trend warning threshold range [G i , G j ], when the change rate of various environmental parameters at different time nodes of meteorological data changes significantly within the warning threshold range, the first meteorological warning information is output;

[0150] Step S404, the first meteorological warning information includes upward trend warning information and downward trend warning information; when the comparison result of the control group of the change rate of different environmental parameters is in an obvious upward trend, the change rate is gradually increasing to output the upward trend warning information; when the comparison result of the control group of the change rate of different environmental parameters is in an obvious downward trend, the change rate is gradually decreasing to output the downward trend warning information.

[0151] In the implementation of the present invention, in step S403, multiple change trend warning mechanisms are generated based on the results of the control group according to the comparison of the change rates of environmental parameters at different time nodes. Different change rates correspond to different warning levels or different warning processing methods; by setting the change trend warning threshold range [Gi, Gj], that is, setting a change rate interval, when the change rate of each environmental parameter in the meteorological data changes significantly within the interval, the warning is triggered.

[0152] The warning threshold [Gi, Gj] is a dynamically adjustable range, which is used to make flexible adjustments according to actual conditions. For example, when certain meteorological parameters change greatly within a specific time period, the threshold may be tighter and the warning may be more frequent; by setting specific change trend thresholds, customized warnings can be made for different environmental parameter change trends, not just for a single change, making the warning more accurate; the threshold range [Gi, Gj] can be dynamically adjusted according to real-time data and historical trends, avoiding the false alarm or missed alarm problems caused by fixed thresholds, and improving the flexibility of the system; it can comprehensively judge meteorological changes based on the change rates of different environmental parameters, ensuring multi-angle and multi-level environmental change monitoring.

[0153] In step S404, when the comparison result of the control group of the environmental parameter change rate shows that the change trend is significant (such as an upward trend or a downward trend), the corresponding first meteorological warning information is output through the set warning mechanism; when the change rate of one or more environmental parameters shows a continuous upward trend, it indicates that the weather or meteorological environment may become more severe or unstable, and the system outputs an upward trend warning information; when the change rate shows that the environmental parameters continue to decrease, it means that the meteorological environment may become more stable, and thus a downward trend warning information is output.

[0154] When weather changes are obvious (whether the trend is rising or falling), the system can quickly issue warnings in the early stages, providing an opportunity for early prevention; by analyzing the rate of change, it can accurately determine future weather trends (such as impending extreme weather or a sudden drop in temperature) to help relevant departments prepare; and it can output different types of warning information (upward or downward trends) based on different trends of change, allowing relevant departments and the public to take different response measures based on specific weather change trends.

[0155] In the process from step S403 to step S404, step S403 mainly sets the change trend warning mechanism based on historical and current meteorological data, and determines the change rate threshold [Gi, Gj], which lays the foundation for the subsequent warning mechanism; after setting the threshold range and warning mechanism in step S403, step S404 determines whether to trigger the warning according to the actual data, and outputs the corresponding warning information. In other words, S404 depends on the warning mechanism setting of S403; step S403 sets different threshold ranges for different change rates, and step S404 generates specific warning information based on these set warning mechanisms, so that the system can respond accurately to meteorological changes. Through this feedback mechanism, the system can continuously optimize the warning decision based on the actual data; the joint work of steps S403 and S404 makes the warning mechanism both flexible and able to adapt to environmental changes in real time, providing meteorological warnings with high accuracy and reliability.

[0156] In a preferred implementation of the present invention, the step S403 further includes:

[0157] Step S4031, based on the meteorological data set and the environmental parameter set, to obtain environmental parameter characteristics of different meteorological data; the environmental parameters include air pressure, temperature, humidity, wind speed and direction, precipitation, visibility, cloud height and thunderstorm activity; the characteristic properties of environmental parameters existing in different meteorological changes are obtained through multiple sensors;

[0158] Step S4032, the first meteorological warning information determines whether an extreme meteorological event may occur by monitoring the rapid changes in environmental parameters; first, a warning threshold range is set, which is usually based on the changes in meteorological data and the obvious fluctuations in environmental parameters over a period of time;

[0159] Step S4033, based on the first meteorological warning information, the impact results of the increase or decrease of the environmental parameters that have obviously changed within the warning threshold range are summarized to obtain the final meteorological change trend information, so as to output the second meteorological warning information; the impact results of the increase or decrease of the environmental parameters represent different meteorological changes caused by various environmental parameters with different change trends, and the impacts on low-altitude flight are different accordingly;

[0160] Step S4034, the second weather warning information includes weather change improvement warning information and weather change deterioration warning information; when the change trend of each environmental parameter at the same time node meets the environmental parameter characteristics of weather change improvement, the weather change improvement warning information is output; when the change trend of each environmental parameter at the same time node meets the environmental parameter characteristics of weather change deterioration, the weather change deterioration warning information is output;

[0161] Step S4035, compare the predicted weather change trend result with the actual weather change to obtain a comparison result; and according to the comparison result, The meteorological change trend results predicted by calibration are: MSE represents the mean square error calibration algorithm, N represents the type of meteorological data, represents the predicted value; Represents the true value.

[0162] In the implementation of the present invention, in step S4031, based on the meteorological data set and the environmental parameter set, the environmental parameters required for different meteorological changes are obtained in step S100; and the characteristics of the environmental parameters will reflect the changes in different meteorological conditions and have different effects on low-altitude flight traffic control; and can provide accurate basic data for subsequent meteorological forecasts and warnings. By extracting and analyzing the characteristics of environmental parameters, a clear meteorological change background can be provided for the subsequent warning system.

[0163] Step S4032 determines whether there is a risk of extreme weather events by monitoring the rapid changes in environmental parameters. Based on step S403, a warning threshold range is set, which is usually based on past meteorological data and fluctuations in environmental parameters; thus, possible extreme weather events such as strong winds, heavy rains, thunderstorms, etc. can be discovered in time, and the warning program can be activated in time, which helps to take preventive measures in advance.

[0164] In step S4033, after the first meteorological warning information is issued in step S403, the impact of significant changes (increase or decrease) in environmental parameters is summarized, the meteorological change trend is analyzed, and the second meteorological warning information is output. This step is to judge the specific trend of meteorological changes (such as worsening or improvement) based on the data changes after the first meteorological warning information; it can further refine the warning information, enhance the accuracy and response speed of meteorological warnings, provide more accurate information for decision makers, and help formulate emergency measures.

[0165] Step S4034 determines whether the weather is getting better or worse based on the changing trend of the environmental parameters. If the changing trend of the environmental parameters meets the characteristics of the weather getting better or worse, the weather warning information of the weather getting better or worse is output respectively; ensuring that the weather warning information is more detailed and accurate, different responses can be made to specific weather changes. This not only helps prevent extreme weather, but also reduces unnecessary overreactions.

[0166] Step S4035 compares the predicted weather change trend with the actual weather change, obtains the comparison result, and calibrates the prediction model according to the comparison result; by comparing the prediction with the actual situation, the accuracy of the prediction model can be corrected, making future weather warnings more accurate. This feedback mechanism continuously optimizes the warning system and improves its adaptability and reliability.

[0167] Through multiple warnings and gradual calibration, the system can provide more accurate weather warning information and reduce false alarms and missed alarms; early warnings and multiple forecasts enable decision makers to take timely response measures to reduce the impact of extreme weather events on flight, safety and other fields; by refining warning levels and comparing actual data, weather changes under different environmental conditions can be responded to flexibly and in real time, allowing the entire system to adapt to different weather change scenarios.

[0168] In a preferred embodiment of the present invention, the above-mentioned low-altitude flight strategies corresponding to various weather changes are planned according to various weather changes, and further include:

[0169] Based on the flight coordinate data, obtain the low-altitude flight coordinate position of the current node; according to the meteorological data, set a variety of low-altitude flight strategies;

[0170] The low-altitude flight strategy includes: in normal weather conditions, the low-altitude flight is smoothly flown along a preset path; when abnormal weather conditions occur, the environmental parameters based on the abnormal weather data are To obtain the estimated values ​​of the coordinate nodes involved in the abnormal weather flight path;

[0171] In the formula, f(x) represents the estimated value of the planned path coordinate node, g(x) represents the distance cost from the original coordinate node to the current coordinate node, h(x) represents the distance cost from the current coordinate node to the target coordinate node, and x represents a coordinate node in the path planning.

[0172] In the implementation of the present invention, the current spatial position of the aircraft is first extracted from the flight coordinate data. These coordinate data come from the aircraft's navigation system (such as GPS) or low-altitude flight monitoring system; the position of the aircraft is accurately located in real time, laying the foundation for the next step of flight path planning, ensuring that subsequent strategy adjustments are based on the actual aircraft status.

[0173] Based on the current meteorological data, the system distinguishes between normal and abnormal meteorological conditions and sets corresponding low-altitude flight strategies: in normal weather, the aircraft will fly smoothly along the preset path; in abnormal weather, the flight path will be dynamically adjusted using the environmental parameters of the abnormal meteorological data (such as wind speed, visibility, etc.); a flexible flight control mechanism is provided to enable the aircraft to operate smoothly under different meteorological conditions, reducing flight risks caused by meteorological changes.

[0174] Under abnormal weather conditions, path planning uses the distance cost from the original coordinate node to the current coordinate node; it represents the consumption of the flight path, which is usually related to weather conditions (wind speed, wind direction) and flight time.

[0175] The distance cost from the current coordinate node to the target coordinate node; represents the predicted cost of the aircraft from the current position to the final target position, estimated based on the current weather environment.

[0176] The path cost of the aircraft at different nodes will be calculated in real time to select the path with the lowest cost as the flight plan; this allows the system to dynamically plan the path under abnormal weather conditions; by comprehensively evaluating environmental influencing factors, the optimal path is selected to reduce flight risks; the flight distance, time and fuel consumption are balanced to achieve economical and efficient flight; the aircraft flies strictly according to the preset path, and there is no need to frequently adjust the path nodes; flight operations are simplified, complexity is reduced, and flight efficiency is improved; based on current weather changes, path planning is adjusted to increase the flexibility of flight strategies; reliable path adjustment solutions are provided in severe weather to improve the safety and resilience of the aircraft; by obtaining flight coordinates and meteorological data in real time, the system can respond quickly to environmental changes; a variety of flight strategies are set to adapt to normal and abnormal weather conditions and enhance the aircraft's ability to cope with complex environments; the cost of path planning is comprehensively considered) to minimize the threat of weather to low-altitude flights; the dynamic path planning model ensures flight efficiency and stability by balancing path cost and flight safety.

[0177] In a preferred implementation of the present invention, the original coordinate node is added to the open list, and g(x)=0 is set; h(s) is the heuristic estimate of the original coordinate node, where s represents the original coordinate node;

[0178] Select the node with the smallest distance from the open list, which is the current coordinate node S; expand each neighbor node S' of S to obtain g(S') and h(S'), where g(S') represents the update distance cost from the original coordinate node to the current coordinate node, and h(S') is the estimated value of the real-time update distance coordinate node;

[0179] If the neighbor node S' has not been added to the open list, add it to the open list; if the neighbor node S' is already in the list, the obtained path distance cost is smaller, and the path cost value is updated in real time; move the current coordinate node S to the closed list to avoid repeated expansion;

[0180] Repeat the operation until the target coordinate node is found or the open list is empty.

[0181] In the implementation of the present invention, the original coordinate node is added to the open list, which is a set of candidate nodes, initially including the starting point node. The starting point value of each node represents the actual path cost from the starting point to the node, while the estimated cost from the node to the target node is usually estimated by a heuristic function (such as Euclidean distance or Manhattan distance).

[0182] Initially, the node at the starting point is set to 0, because the actual cost from the starting point to the starting point is g(x)=0, and h(s) is the heuristic estimate of the original starting point; through this step, the system provides a basis for subsequent path expansion. Setting g(x) and h(s) for each node, and obtaining g(S') and h(S') through coordinate node expansion can help the algorithm determine which path target node is better and prepare for selecting the optimal path; select the node with the smallest distance from the open list, which is the current coordinate node. In the open list, select the node with the smallest current f(x)=g(x)+h(x) value as the current node, that is, the node most likely to go to the target. This node will be used for further path expansion.

[0183] Selecting the node with the smallest distance allows the algorithm to always expand in the target direction, thereby improving the efficiency and accuracy of path planning and avoiding unnecessary expansion; expanding each neighbor node of the current coordinate node to obtain the corresponding expanded node of the neighbor node, updating the estimated value of the distance coordinate node in real time, processing all neighbor nodes of the current node, obtaining the actual cost from the current node to the neighbor node and the estimated cost from the neighbor node to the target node; obtained by adding the edge weight (or movement cost) from the current node to the neighbor node; when the heuristic estimate of the neighbor node to the target node.

[0184] Update the extended neighbor nodes of the neighbor nodes and obtain the path cost value in real time, so that the path cost of each node is accurately evaluated, thus providing a basis for selecting the next extended node; when the neighbor node is not in the open list, add it to the open list and prepare for further expansion; ensure that all possible paths are considered and all potential shortest path candidate nodes can be visited in time; when the neighbor node is already in the list, the obtained path distance cost is smaller, and the path cost value is updated in real time;

[0185] When the neighbor node already exists in the open list and the path through the current node is shorter (that is, the new distance cost is smaller), the path distance is updated to reflect the shorter path; this update process ensures the optimality of the path. By continuously optimizing the path cost of the expanded nodes, the final path cost is ensured to be minimum; the current coordinate node is moved to the closed list to avoid repeated expansion. After the current node is processed, it is added to the closed list (ClosedList). The closed list contains all nodes that have been expanded to avoid repeated expansion.

[0186] By moving the processed nodes to the closed list, repeated calculations are avoided and the efficiency of the algorithm is improved. In addition, it prevents dead loops in path planning and ensures that nodes are effectively processed; repeat the operation until the target coordinate node is found or the open list is empty; select the best node from the open list to expand until the target node is found or the open list is empty; it means that the path planning is completed and the path can be traced back from the target node to obtain the shortest path; it means that no path is found, indicating that the target is unreachable; by repeatedly expanding nodes and updating the path cost, the algorithm ensures that the best path is eventually found. If the open list is empty, it means that the path is unreachable, which helps to avoid the generation of errors or invalid paths.

[0187] In a preferred embodiment of the present invention, the low-altitude flight strategy is adjusted in a timely manner based on the early warning information of various meteorological change trends, and further includes:

[0188] Based on the low-altitude flight strategy, To obtain the low-altitude flight yaw scheduling coordinates; the low-altitude flight yaw scheduling coordinates are the target coordinates for regulating the low-altitude flight strategy;

[0189] Where C(path) is the coordinate point of the low-altitude flight path after yaw scheduling, x i ,y i 、z i They represent the control parameters of the current low-altitude flight, which include the coordinate position, flight direction and flight angle of the current low-altitude flight, T i represents the flight time, cost represents the flight risk or energy consumption cost based on weather changes, and n represents the number of low-altitude flight control parameters;

[0190] Based on the low-altitude flight yaw scheduling coordinates, a plurality of the low-altitude flight yaw scheduling coordinates are obtained to obtain a low-altitude flight route after regulation, The flight strategy is gradually optimized to obtain the optimal flight path strategy; the interference of weather changes is avoided according to the optimal flight path strategy; where V(t) represents the optimal strategy cost at time t, Indicates from time The flight cost to t;

[0191] In the implementation of the present invention, according to the low-altitude flight strategy, the yaw scheduling coordinates are first calculated by obtaining the control parameters of the low-altitude flight. The control parameters of the low-altitude flight include the current coordinate position (x i , y i , z i ), flight direction, flight angle and flight time (T i ); when these parameters together define the flight path and the dynamic state of the flight; xi , y i and z i The three-dimensional coordinates representing the current position of the aircraft, which determine the heading and tilt of the aircraft; t i Represents time, that is, the state of the aircraft at a given point in time.

[0192] Through the above control parameters, the yaw scheduling coordinates of low-altitude flight can be obtained, that is, the updated flight path or position. These coordinate points are the target coordinates that the aircraft needs to reach in the next step; through precise control parameters, dynamic and real-time path adjustment can be provided for low-altitude aircraft, so that the aircraft can flexibly respond to real-time changing environmental conditions;

[0193] Based on the acquired low-altitude flight yaw scheduling coordinates, the system gradually optimizes the low-altitude flight path by calculating multiple scheduling coordinate points. Each newly calculated flight path is based on the result of the previous path optimization, gradually reducing energy consumption, flight time and flight risk; the goal of the optimization is to avoid adverse weather conditions, reduce flight risks, or optimize flight energy efficiency by adjusting the path coordinates. The optimization process is usually based on a cost function, where the cost function may include factors such as risk assessment of the flight path, energy consumption, and time cost; thus, the optimized strategy can effectively reduce the risk of flight, especially in low-altitude flight, which is affected by factors such as weather changes, obstacles, and the performance of the aircraft itself. Gradually optimizing the path can help the aircraft avoid danger and improve flight safety and efficiency.

[0194] Eventually, the system will calculate an optimal flight path that can minimize the flight cost while ensuring flight safety. It usually involves factors such as energy consumption and flight time of the flight path; at each moment t, the system will adjust the flight path according to real-time data such as flight status and weather changes to optimize the flight cost. Through this process, the optimization of the flight path can be updated in real time to adapt to changes in the flight environment; through the optimization of this optimal path, the aircraft can not only effectively avoid the risks brought by weather changes, but also maximize flight energy efficiency, reduce energy consumption, and extend flight time. At the same time, since the path optimization process is adjusted in real time, the system can flexibly respond to various uncertainties that arise during the flight process to ensure the stability and safety of the flight.

[0195] A key goal of the optimal flight path strategy is to avoid interference from weather changes. The optimization of the flight path will adjust the route in combination with real-time weather data (such as wind speed, air pressure, temperature, etc.). Through this adjustment, the aircraft can avoid encountering areas with severe weather and reduce the risks during flight; since meteorological factors are one of the key factors affecting low-altitude flight, being able to effectively avoid meteorological interference can greatly improve flight safety and avoid accident risks caused by abnormal weather, such as storms, lightning, strong air currents, etc.

[0196] Throughout the flight, the calculation of cost functions and flight costs is the basis of the step-by-step optimization process. These costs reflect the impact of cost, energy consumption, time, etc. during the flight process. By optimizing these cost functions, the flight path can be adjusted to the optimal solution to ensure that the aircraft completes the flight mission at the lowest cost (including energy, time and risk); minimizing the flight cost can not only improve flight efficiency and reduce energy consumption, but also provide the best decision at every moment of the flight, ensuring that the aircraft can operate stably in complex environments and avoid unnecessary delays and additional costs.

[0197] like Figure 2 As shown, a low-altitude meteorological early warning system based on the Internet of Things includes:

[0198] Data collection module: It is used to collect a variety of historical meteorological data to obtain a meteorological data set; according to the meteorological data set, to obtain a variety of environmental parameters corresponding to each meteorological data item, and to compile an environmental parameter set; to obtain the low-altitude flight space coordinates of the current time node to obtain flight coordinate data;

[0199] Model training module: used to construct and train the meteorological data set and the environmental parameter set to obtain a meteorological prediction model; based on the meteorological prediction model, various environmental parameters in the low-altitude flight meteorological data at the current time node are obtained in real time, and the meteorological data environmental parameters at multiple time nodes are input into the meteorological prediction model to output the meteorological data prediction result representing the next time node;

[0200] Early warning response module: It is used to judge the meteorological change trend according to the meteorological data prediction results of the meteorological forecast model; set multiple change trend early warning thresholds, respond to the preset early warning thresholds according to different meteorological change trends, and output early warning information.

[0201] Therefore, the purpose of the present invention can also be achieved by running a program or a group of programs on any computing device. The computing device can be a well-known general intelligent device. Therefore, the purpose of the present invention can also be achieved by simply providing a program product containing a program code that implements the method or system. It should also be pointed out that in the device and method of the present invention, it is obvious that each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. In addition, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order. Some steps can be performed in parallel or independently of each other.

Claims

1. A low-altitude meteorological early warning method based on the Internet of Things, characterized in that: The following methods are included: S100, collecting a variety of historical meteorological data to obtain a meteorological data set; obtaining a variety of environmental parameters corresponding to each meteorological data item according to the meteorological data set, and assembling the environmental parameter set; obtaining the low-altitude flight space coordinates of the current time node to obtain flight coordinate data; S200, constructing and training the meteorological data set and the environmental parameter set to obtain a meteorological prediction model; S300, based on the meteorological prediction model, real-time acquisition of low-altitude flight meteorological data environment parameters at the current time node, inputting the meteorological data environment parameters at multiple time nodes into the meteorological prediction model to output a meteorological data prediction result representing the next time node; S400, judging the weather change trend according to the weather data prediction result of the weather prediction model; Set a variety of change trend warning thresholds, respond to the preset warning thresholds according to different meteorological change trends, and output warning information; Plan various low-altitude flight strategies based on weather changes; The low-altitude flight strategy obtains the flight path coordinate nodes according to the warning information; by adding the original coordinate nodes to the open list and continuously expanding the neighboring nodes, the path cost and the estimated value are updated in real time until the target node is found or the open list is empty, that is, the safe coordinate node is obtained.

2. The low-altitude meteorological early warning method based on the Internet of Things according to claim 1 is characterized in that: According to various meteorological changes, low-altitude flight strategies corresponding to various meteorological changes are planned; based on the low-altitude flight strategies, according to the early warning information of various meteorological change trends, the low-altitude flight strategies are adjusted in time.

3. The low-altitude meteorological early warning method based on the Internet of Things according to claim 1 is characterized in that: The meteorological prediction model comprises: The meteorological data set and the environmental parameter set are used to construct a data set to generate structured data, and the structured data is encoded into sequence data to train the meteorological prediction model; The sequence data is input into the weather forecast model; the weather forecast model comprises an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output layer, the intermediate representation data of multiple hidden layers are transmitted to the output layer, and the output layer outputs the weather data recognition results corresponding to the multiple environmental parameters; At least two environmental parameter data items at the current time node are input into the meteorological prediction model, and the output layer output of the meteorological prediction model represents the current meteorological data prediction result.

4. The low-altitude meteorological early warning method based on the Internet of Things according to claim 3 is characterized in that: According to the meteorological data prediction results of the meteorological prediction model, the meteorological change trend is judged, including: According to the meteorological data prediction results, meteorological data and environmental parameters of multiple time nodes are obtained; based on the environmental parameters of meteorological data at different time nodes, Obtain the change rates of various environmental parameters at different time nodes; In the formula, H represents the change rate between adjacent time nodes, and Respectively represent the time node t i and t i+1 At least one environment parameter variable value A, represents the time difference between the i-th and i+1-th adjacent time nodes, and t i+1 Indicates t i The time node after the time node; Obtain the change rates between at least three different adjacent time nodes to obtain a change rate control group; compare at least two of the change rate control groups to obtain a comparison result; and judge the meteorological change trend based on the existence of a plurality of different comparison results in different change rate control groups.

5. The low-altitude meteorological early warning method based on the Internet of Things according to claim 4 is characterized in that: Set a variety of change trend warning thresholds, and output warning information in response to preset warning thresholds according to different meteorological change trends, including: Based on various different comparison results of different change rate control groups, various change trend warning mechanisms are set; the various change trend warning mechanisms set change trend warning threshold ranges [G i , G j ], when the change rate of various environmental parameters at different time nodes of meteorological data changes significantly within the warning threshold range, the first meteorological warning information is output; The first meteorological warning information includes upward trend warning information and downward trend warning information; when the comparison result of the change rate control group of different environmental parameters is in an obvious upward trend, the change rate is gradually increasing to output the upward trend warning information; when the comparison result of the change rate control group of different environmental parameters is in an obvious downward trend, the change rate is gradually decreasing to output the downward trend warning information.

6. The low-altitude meteorological early warning method based on the Internet of Things according to claim 5 is characterized in that: The warning information also includes: Based on the meteorological data set and the environmental parameter set, the environmental parameter characteristics of different meteorological data are obtained; based on the first meteorological warning information, the impact results of the increase or decrease of the environmental parameters that have changed significantly within the warning threshold range are summarized to obtain the final meteorological change trend information, so as to output the second meteorological warning information; The second meteorological warning information includes warning information for improved meteorological changes and warning information for worse meteorological changes; when the change trend of each environmental parameter at the same time node meets the environmental parameter characteristics of improved meteorological changes, the warning information for improved meteorological changes is output; when the change trend of each environmental parameter at the same time node meets the environmental parameter characteristics of worse meteorological changes, the warning information for worse meteorological changes is output; Compare the predicted meteorological change trend results with the actual meteorological changes to obtain the comparison results; and according to the comparison results, The meteorological change trend results predicted by calibration are: MSE represents the mean square error calibration algorithm, N represents the type of meteorological data, represents the predicted value; Represents the true value.

7. The low-altitude meteorological early warning method based on the Internet of Things according to claim 2 is characterized in that: According to various weather changes, we plan low-altitude flight strategies corresponding to various weather changes, including: Based on the flight coordinate data, obtain the low-altitude flight coordinate position of the current node; according to the meteorological data, set a variety of low-altitude flight strategies; The low-altitude flight strategy includes: in normal weather conditions, the low-altitude flight is smoothly flown along a preset path; when abnormal weather conditions occur, the environmental parameters based on the abnormal weather data are To obtain the estimated values ​​of the coordinate nodes involved in the abnormal weather flight path; In the formula, f(x) represents the estimated value of the planned path coordinate node, g(x) represents the distance cost from the original coordinate node to the current coordinate node, h(x) represents the distance cost from the current coordinate node to the target coordinate node, and x represents a coordinate node in the path planning.

8. The low-altitude meteorological early warning method based on the Internet of Things according to claim 7 is characterized in that: Add the original coordinate node to the open list and set g(x)=0; h(s) is the heuristic estimate of the original coordinate node, where s represents the original coordinate node; Select the node with the smallest distance from the open list, which is the current coordinate node S; expand each neighbor node S' of S to obtain g(S') and h(S'), where g(S') represents the update distance cost from the original coordinate node to the current coordinate node, and h(S') is the estimated value of the real-time update distance coordinate node; If the neighbor node S' has not been added to the open list, add it to the open list; if the neighbor node S' is already in the list, the obtained path distance cost is smaller, and the path cost value is updated in real time; move the current coordinate node S to the closed list to avoid repeated expansion; Repeat the operation until the target coordinate node is found or the open list is empty.

9. The low-altitude meteorological early warning method based on the Internet of Things according to claim 8, characterized in that: Based on the low-altitude flight strategy, the low-altitude flight strategy is adjusted in a timely manner according to the early warning information of various meteorological change trends, including: Based on the low-altitude flight strategy, To obtain the low-altitude flight yaw scheduling coordinates; the low-altitude flight yaw scheduling coordinates are the target coordinates for regulating the low-altitude flight strategy; Where C(path) is the coordinate point of the low-altitude flight path after yaw scheduling, x i ,y i and z i They represent the control parameters of the current low-altitude flight, which include the coordinate position, flight direction and flight angle of the current low-altitude flight, T i represents the flight time, cost represents the flight risk or energy consumption cost based on weather changes, and n represents the number of low-altitude flight control parameters; Based on the low-altitude flight yaw scheduling coordinates, a plurality of the low-altitude flight yaw scheduling coordinates are obtained to obtain a low-altitude flight route after regulation, The flight strategy is gradually optimized to obtain the optimal flight path strategy; the interference of weather changes is avoided according to the optimal flight path strategy; where V(t) represents the optimal strategy cost at time t, Indicates from time The flight cost to t.

10. A low-altitude meteorological early warning system based on the Internet of Things, characterized in that: The system is provided with an electronic device including a memory, a processor, and a low-altitude meteorological early warning method program based on the Internet of Things stored in the memory and executable on the processor. When the low-altitude meteorological early warning method program based on the Internet of Things is executed by the processor, the steps of the low-altitude meteorological early warning method based on the Internet of Things are implemented as described in any one of claims 1 to 9. The system includes: Data collection module: It is used to collect a variety of historical meteorological data to obtain a meteorological data set; according to the meteorological data set, to obtain a variety of environmental parameters corresponding to each meteorological data item, and to compile an environmental parameter set; to obtain the low-altitude flight space coordinates of the current time node to obtain flight coordinate data; Model training module: it is used to construct and train the meteorological data set and the environmental parameter set to obtain a meteorological prediction model; based on the meteorological prediction model, various environmental parameters in the low-altitude flight meteorological data at the current time node are obtained in real time, and the meteorological data environmental parameters of multiple time nodes are input into the meteorological prediction model to output the meteorological data prediction result representing the next time node; Early warning response module: It is used to judge the meteorological change trend according to the meteorological data prediction results of the meteorological prediction model; set multiple change trend early warning thresholds, and output early warning information according to different meteorological change trends in response to the preset early warning thresholds.

Images