Artificial intelligence-driven typhoon path dynamic tracking method and system

Abstract

The invention provides an artificial intelligence-driven typhoon path dynamic tracking method and system, the system comprises an operation environment, a data control program and a typhoon positioning program, the data control program is responsible for managing data flow input and output, and the data control program comprises an initial coordinate processing process, a meteorological data preprocessing process and a data post-processing process; path tracking coordinates are updated through loop iteration, complete typhoon tracking data are finally output, and a typhoon positioning program adopts a three-layer progressive processing architecture and comprises dynamic window tracking, multi-scale feature fusion network reasoning and path coordinate generation. The method has the advantages that based on the deep learning multi-scale feature fusion network, numerical weather forecast data can be accurately and rapidly utilized to complete typhoon path tracking, the operation environment is compatible with operation in GPU and CPU environments, and the operation efficiency is far higher than that of a traditional positioning algorithm.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of meteorological forecasting and artificial intelligence, and in particular relates to an artificial intelligence-driven method and system for dynamic tracking of typhoon paths. Background Technology

[0002] Typhoon track tracking, as a core task of meteorological forecasting, has long been a focus of attention for academic and operational institutions both domestically and internationally. Traditional methods mainly rely on numerical weather prediction models (such as WRF and ECMWF) and statistical models, achieving predictions by solving atmospheric dynamic equations or fitting historical data. In recent years, with the advancement of international collaborative projects and technological iterations, the accuracy of typhoon track predictions has gradually improved. For example, the Typhoon Landfall Forecasting Demonstration Project (TLFDP), led by the World Meteorological Organization (WMO), has developed 14 new technologies for forecasting typhoon tracks, intensities, and precipitation by integrating the technological achievements of multiple national institutions, significantly improving the reliability of forecasts. At the specific technical level, the vortex tracking technique (GVT) proposed by the US GFDL (Geophysical Fluid Dynamics Laboratory) has become one of the representative methods. This technique, by locating the core region of the typhoon vortex, effectively corrects the "sawtooth" oscillation problem that easily occurs in track predictions using traditional methods (such as the minimum sea level pressure method and the maximum 500 hPa vorticity method). Studies have shown that the application of the GVT regional model in Zhejiang Province makes typhoon track predictions closer to actual observations, especially in controlling the positioning error of the minimum pressure at the typhoon center.

[0003] While traditional methods have made progress in business applications, their limitations are becoming increasingly apparent: 1. Insufficient timeliness: Traditional methods take a relatively long time to generate typhoon path tracking data and consume a lot of resources; 2. Limited ability to capture nonlinear features: Changes in structure and properties and sudden changes in path during the dynamic evolution of typhoons often lead to positioning errors. Summary of the Invention

[0004] In view of this, the present invention aims to propose an artificial intelligence-driven dynamic tracking method and system for typhoon paths, in order to solve the problems of insufficient timeliness and limited nonlinear feature capture capability of existing technologies.

[0005] To achieve the above objectives, the technical solution of the present invention is implemented as follows: In a first aspect, the present invention provides an artificial intelligence-driven method for dynamic tracking of typhoon paths, comprising the following steps: S1. Train the MSF-Net model, a multi-scale feature fusion network; The MSF-Net model includes the following modules: a primary feature encoding module, used to extract texture features of the input meteorological field step by step through convolutional operations; a deep feature enhancement module, used to introduce Gaussian noise into the convolutional layers and add the input features and output features through cross-layer connections; a global context fusion module, used to integrate multi-scale features using convolutional kernels and generate a global feature vector; a feature compression module, used to compress the global feature vector through a fully connected network; and a relative position index prediction module, used to output the relative position index of the typhoon center. S2. Based on the original numerical weather forecast data, the meteorological data processing layer performs meteorological data fusion and grid resampling, and outputs the processed numerical weather forecast data. S3. Based on the initial typhoon information, the initial coordinate processing layer determines the grid index and boundary range of the typhoon center; the geographic coordinate mapping layer establishes a two-way conversion relationship between geographic coordinates and grid index. S4. Iterate and update the typhoon path tracking coordinates in a loop; S5. Output typhoon path tracking data and perform Gaussian smoothing and quality control processing on the typhoon path tracking data. In step S3, based on the initial typhoon information, the initial coordinate processing layer determines the grid index and boundary range of the typhoon center, including: Based on the typhoon's starting location, the initial coordinate processing layer generates the initial latitude and longitude index of the typhoon center and the initial latitude and longitude boundary index. In step S4, the typhoon path tracking coordinates are updated iteratively, including: S41. Based on the processed numerical weather forecast data, the latitude and longitude index of the typhoon center, and the latitude and longitude boundary index, a sliding window extraction layer is used to extract the local meteorological field. S42. Based on the extracted local meteorological field, the MSF-Net inference layer of the multi-scale feature fusion network predicts the relative position shift of the typhoon center. S43. Based on the predicted relative position of the typhoon center, the path coordinate generation layer updates the latitude and longitude index and latitude and longitude boundary index of the typhoon center. S44. The updated latitude and longitude indexes of the typhoon center and the latitude and longitude boundary indexes are fed back into the sliding window extraction layer for the next round of iteration.

[0006] Furthermore, in step S1, the multi-scale feature fusion network MSF-Net model is trained, including: S11. Divide the typhoon path data into training set, validation set and test set; S12. Using the training set, validation set, and test set, the MSF-Net model of the multi-scale feature fusion network is learned, optimized, and evaluated. S13. Save the MSF-Net model of the multi-scale feature fusion network.

[0007] Furthermore, in step S2, the meteorological data processing layer performs meteorological data fusion and grid resampling, including: S21. Meteorological data fusion: Integrating meteorological parameter channels covering near-surface meteorological element fields and upper-air meteorological element fields; S22. Grid resampling processing: The meteorological data is resampled to a fixed grid of a set size using a bilinear interpolation algorithm, covering the specified latitude and longitude area.

[0008] Furthermore, in step S41, the sliding window extraction layer extracts the local meteorological field, including: S411. Set window size; S412. Dynamically adjust window offset based on the predicted location of the typhoon center; S413. Boundary constraint check: When any boundary of the window is less than 10 grid units away from the boundary of the global data domain, automatically adjust the offset to ensure that the window is completely within the data boundary; S414. Extract local meteorological field: Output a three-dimensional tensor and use it as the input unit of the MSF-Net deep learning model. The input unit of the MSF-Net deep learning model includes meteorological field information in both spatial and temporal dimensions.

[0009] In step S42, the MSF-Net inference layer of the multi-scale feature fusion network predicts the relative position shift of the typhoon center, including: S421. Based on the input unit of the MSF-Net deep learning model, the MSF-Net network backbone of the multi-scale feature fusion network MSF-Net inference layer performs the inference process; S422, The MSF-Net inference layer of the multi-scale feature fusion network outputs the relative position index of the typhoon center; In step S421, the MSF-Net network backbone performs the inference process, including: S4211, Primary Feature Encoding: Extracting meteorological field texture features step by step through multi-level convolution operations; S4212, Deep Feature Enhancement: Introduce Gaussian noise into the convolutional layer; add the input features and output features through cross-layer connections; S4213, Global Context Fusion: Convolutional kernels are used to integrate multi-scale features and generate a global feature vector; S4214. Feature Compression: Compress the global feature vector through a fully connected network.

[0010] Furthermore, in step S43, the path coordinate generation layer updates the latitude and longitude indexes of the typhoon center and the latitude and longitude boundary indexes, including: S431. Decode relative position index: Convert the position index within the relative window into a latitude and longitude index; S432. Update typhoon center coordinates: Use the decoded latitude and longitude index as the new typhoon center coordinates, and calculate the latitude and longitude boundary index of the typhoon center based on the latitude and longitude index of the typhoon center.

[0011] Secondly, based on the same concept, the present invention also provides an artificial intelligence-driven dynamic typhoon path tracking system, including an operating environment, a data control program, and a typhoon positioning program. The operating environment is either a CPU operating environment or a GPU operating environment; The data control program includes an initial coordinate processing layer, a meteorological data processing layer, and a geographic coordinate mapping layer. The initial coordinate processing layer is used to determine the grid index and boundary range based on the typhoon's starting location. The meteorological data processing layer is used to fuse multi-source meteorological data and perform grid resampling. The geographic coordinate mapping layer is used to establish a bidirectional conversion relationship between geographic coordinates and grid indexes. The typhoon location procedure includes a sliding window extraction layer, a multi-scale feature fusion network MSF-Net inference layer, and a path coordinate generation layer. The sliding window extraction layer is used to extract the local meteorological field, the multi-scale feature fusion network MSF-Net inference layer is used to predict the relative position shift of the typhoon center, and the path coordinate generation layer is used to convert the relative position into absolute coordinates and update the latitude and longitude index and latitude and longitude boundary index.

[0012] Furthermore, the inference layer of the multi-scale feature fusion network MSF-Net includes a primary feature encoding module, a deep feature enhancement module, a global context fusion module, a feature compression module, and a relative position index prediction module; The primary feature encoding module is used to extract the texture features of the meteorological field step by step. The deep feature enhancement module is used to introduce Gaussian noise into the convolutional layer and add the input features and output features through cross-layer connections. The global context fusion module is used to integrate multi-scale features through convolutional kernels to generate a global feature vector. The feature compression module is used to compress the global feature vector through a fully connected network. The relative position index prediction module is used to predict the relative position index.

[0013] Compared with existing technologies, the artificial intelligence-driven dynamic tracking method and system for typhoon paths described in this invention has the following advantages: (1) A spatiotemporal sliding window tracking mechanism is adopted to focus on the core area of ​​the typhoon by dynamically adjusting the window position, effectively capturing spatiotemporal local features.

[0014] (2) Using convolutional neural networks (CNN) to extract meteorological field features, compared with traditional methods, it can more accurately predict the sudden changes and turning points of typhoon paths.

[0015] (3) Utilize GPU to accelerate the training process, while ensuring feature integrity and significantly improving computational efficiency to meet real-time requirements.

[0016] (4) For the typhoon path tracking task, a special network structure including a Gaussian noise enhancement layer, a multi-level downsampling module and residual connections is designed to inject Gaussian noise to enhance the robustness of the model to data noise and enhance the generalization ability.

[0017] (5) Progressive feature compression uses 5 levels of convolution to gradually compress the 11-channel input to a 2-channel offset output, reducing computation while retaining the feature information that is crucial for typhoon path prediction.

[0018] (6) The geographic coordinate inverse mapping algorithm converts the grid index into latitude and longitude coordinates, supports non-uniform grid mapping, and covers the area from latitude 60°N to 0° and longitude 60°E to 180°E, meeting the needs of typhoon tracking in the Northwest Pacific. Attached Figure Description

[0019] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the overall system flow according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the overall system architecture as described in an embodiment of the present invention; Figure 3 This is a schematic diagram of the quality control system process described in an embodiment of the present invention; Figure 4 This is a schematic diagram of the path coordinate generation layer process according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the MSF-Net inference layer process according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the sliding window extraction layer process according to an embodiment of the present invention; Figure 7 This is a schematic diagram of the data post-processing layer flow according to an embodiment of the present invention; Figure 8 This is a schematic diagram of the geographic coordinate mapping layer process according to an embodiment of the present invention; Figure 9 This is a schematic diagram of the meteorological data processing layer flow according to an embodiment of the present invention; Figure 10This is a schematic diagram of the initial coordinate processing layer process according to an embodiment of the present invention; Figure 11 This is a schematic diagram of dynamic window tracking as described in an embodiment of the present invention.

[0020] Figure 12 This is a schematic diagram of the model structure of the present invention. Detailed Implementation

[0021] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0022] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0023] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0024] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0025] like Figures 1 to 2 As shown, an AI-driven typhoon path dynamic tracking system includes an operating environment, a data control program, and a typhoon positioning program; the operating environment is a CPU operating environment or a GPU operating environment; the data control program includes an initial coordinate processing layer, a meteorological data processing layer, and a geographic coordinate mapping layer; the typhoon positioning program includes a sliding window extraction layer, a multi-scale feature fusion network (MSF-Net) inference layer, and a path coordinate generation layer. An AI-driven typhoon path dynamic tracking method includes the following steps: S1. Train the MSF-Net model, a multi-scale feature fusion network; S2. Based on the original numerical weather forecast data, the meteorological data processing layer performs meteorological data fusion and grid resampling, and outputs the processed numerical weather forecast data. S3. Based on the initial typhoon information, the initial coordinate processing layer determines the grid index and boundary range of the typhoon center; the geographic coordinate mapping layer establishes a two-way conversion relationship between geographic coordinates and grid index. S4. Iterate and update the typhoon path tracking coordinates in a loop; S5 outputs typhoon path tracking data and performs Gaussian smoothing and quality control processing on the typhoon path tracking data.

[0026] The specific implementation method is as follows: I. Training the MSF-Net Model (Multi-Scale Feature Fusion Network) Data source: 2020-2024 Northwest Pacific typhoon track data (CMA optimal track set), including typhoon center latitude and longitude coordinates and meteorological field parameters at the corresponding time.

[0027] Division ratio: Training set: Data from 2020-2023 (80%), used for model parameter learning; Validation set: Data from January to August 2020-2023 (20%), used for hyperparameter tuning; Test set: Data from 2024, used for final performance evaluation.

[0028] Data format: Each sample includes spatiotemporal sequence data of 11 meteorological parameter channels, including near-surface meteorological elements (such as sea level pressure and 10-meter UV wind field) and upper-air meteorological elements (such as 700hPa / 850hPa UV wind field, geopotential height, and relative vorticity), along with the corresponding typhoon center label. Based on the above division, a total of 2920 data points were obtained for training, validation, and testing. After training and optimization, the model was saved as a pth format file.

[0029] II. System Architecture The system inputs include two types of data sources: raw numerical weather forecast data and initial typhoon information. The system adopts a dual-program collaborative architecture, with the data control program and the typhoon location program processing in parallel, and dynamically tracking the typhoon path through a loop iteration mechanism.

[0030] The data control program is primarily responsible for data input and output management, comprising three core processing layers: an initial coordinate processing layer that determines the grid index and boundary range based on the typhoon's starting position; a meteorological data processing layer that fuses multi-source meteorological data and performs grid resampling; and a geographic coordinate mapping layer that establishes a bidirectional conversion relationship between latitude / longitude and the grid index. Through a cyclic update mechanism, the program ultimately outputs complete time-series typhoon path tracking data.

[0031] The typhoon location program employs a three-layer progressive processing architecture: a sliding window extraction layer dynamically extracts a 40×40 grid area centered on the typhoon's current location; a multi-scale feature fusion network (MSF-Net) inference layer extracts features based on a deep learning model and predicts relative position offsets; and a path coordinate generation layer converts the relative position into absolute coordinates and updates the boundary index. The program uses a closed-loop feedback mechanism, taking the updated coordinates and boundary indexes obtained from the current prediction as input for the next time step, iterating until path predictions for all forecast lead times are completed.

[0032] The functions and interaction logic of each level of the data control program are as follows: 2.1 Initial coordinate processing layer, such as Figure 10 As shown: Function: First, using the relationships provided by the geographic coordinate mapping layer (see paragraph 2.3), the latitude and longitude coordinates (lat_obs, lon_obs) of the typhoon's initial data are converted into initial latitude and longitude indices (lat_idx_init, lon_idx_init) on the 241×481 fixed grid. The initial latitude and longitude index refers to the row and column number of the grid point in the global data array, both of which are integers starting from 0.

[0033] Subsequently, a dynamic tracking area centered on the typhoon's initial position is established. The boundary of this area is defined by the initial latitude and longitude boundary index, calculated using the following formula: Latitude boundary index: lat_idx_min = lat_idx_init - 20, lat_idx_max = lat_idx_init + 20; Longitude boundary index: lon_idx_min = lon_idx_init - 20, lon_idx_max = lon_idx_init + 20; `lat_idx_init`: Initial latitude grid row number for the typhoon center. `lon_idx_init`: Initial longitude grid column number for the typhoon center. `lat_idx_min`: Minimum latitude boundary row number for the tracking area. `lat_idx_max`: Maximum latitude boundary row number for the tracking area. `lon_idx_min`: Minimum longitude boundary column number for the tracking area. `lon_idx_max`: Maximum longitude boundary column number for the tracking area.

[0034] The value "20" is set based on the fact that the sliding window size used in this system is 40×40 grid units (see section 2.5.1). The formula means that, with the initial typhoon location as the center, expanding outwards by 20 grid points in each direction creates a square area with sides of 40 grid units, which serves as the initial search range for subsequent sliding window extraction. This ensures that the initial window can completely cover the core vortex structure of the typhoon.

[0035] 2.2 Meteorological data processing layer, such as Figure 9 As shown: Input: Receive raw numerical weather forecast data (6-hour interval numerical weather forecast model forecast data).

[0036] The processing includes: 2.2.1 Meteorological data fusion: Content: Integrates 11 key meteorological parameter channels, specifically covering near-surface meteorological elements (sea level pressure, 10-meter UV wind field) and upper-air meteorological elements (700hPa / 850hPa UV wind field, geopotential height, relative vorticity).

[0037] Resolution: The original data resolution is 0.25°×0.25° latitude and longitude grid.

[0038] 2.2.2 Grid resampling processing: Method: Bilinear interpolation algorithm is used.

[0039] Target grid: Data is resampled to a fixed grid of 241×481, covering the latitude range of 60°N to 0° and the longitude range of 60°E to 180°E.

[0040] Objective: To ensure spatial consistency of meteorological data from different sources and resolutions, and to output processed numerical weather forecast data.

[0041] 2.3 Geographic coordinate mapping layer, such as Figure 8 As shown: Method: Establish a bidirectional mapping relationship between latitude and longitude and grid index.

[0042] Implementation method: The conversion between latitude and longitude and grid index is completed quickly by pre-compiling a lookup table.

[0043] Performance: A single query takes less than 0.1ms, meeting real-time requirements.

[0044] 2.4 Data post-processing layer, such as Figure 7 As shown: 2.4.1 Reverse Geographic Mapping Function: Establishes a two-way conversion relationship between geographic coordinates (latitude and longitude) and grid index.

[0045] The fixed grid covers a latitude and longitude range of 60°N to 0° and a longitude range of 60°E to 180°E. The grid resolution is 0.25° × 0.25°, therefore there are (60° / 0.25°) = 241 points in the latitude direction and (120° / 0.25°) = 481 points in the longitude direction.

[0046] The formula for calculating geographic coordinates (lat, lon) from grid index (lat_idx, lon_idx) is as follows: Latitude: lat = 60° - (lat_idx / 240) × 60°; Longitude: lon = 60° + (lon_idx / 480) × 120°; Explanation of the meaning of formula parameters: lat_idx and lon_idx: are the index values ​​of the grid points in the latitude and longitude directions, respectively.

[0047] The denominators 240 and 480: This is because there are 241 points from index 0 to index 240, but they span 240 intervals. The same applies to longitude. This calculation represents dividing the total number of intervals equally.

[0048] Latitude formula: The starting point is 60°N (when lat_idx=0). As lat_idx increases, the latitude decreases linearly towards the equator (0°).

[0049] Longitude formula: The starting point is 60°E (when lon_idx=0), and as lon_idx increases, the longitude increases linearly towards 180°E.

[0050] Correspondingly, the formula for calculating the grid index (lat_idx, lon_idx) from the geographic coordinates (lat, lon) can be derived by reversing the above formula, specifically: lat_idx = floor( (60° - lat) / 60° × 240 ); lon_idx = floor( (lon - 60°) / 120 ° × 480 ); where floor(·) is the floor function.

[0051] 2.4.2 Gaussian smoothing filter: Parameter: σ = 1.5 grid units.

[0052] Objective: To smooth the predicted path, eliminate jitter that may occur in single-step prediction, and improve the consistency and reliability of the path.

[0053] 2.4.3 Quality control system, such as Figure 3 As shown: Reasonableness check: Coastline buffer zone: When the predicted location is more than 50km from the coastline, an early warning mechanism is triggered.

[0054] Abnormal movement speed: When the typhoon's movement speed exceeds 40m / s, it is considered abnormal data and will be marked or corrected.

[0055] 2.4.4 Output: Output the final typhoon path data.

[0056] The functions and interaction logic of each level of the typhoon location program are as follows: 2.5 Sliding window extraction layer, such as Figure 6 As shown: Input: Processed numerical weather forecast data, current typhoon center coordinates (initial latitude and longitude index or updated latitude and longitude index), and boundary index.

[0057] 2.5.1 Window Size Setting: Size: 40×40 grid units.

[0058] Spatial range: Corresponding to an area of ​​approximately 1000km × 1000km, capable of covering the main impact area of ​​typhoons.

[0059] 2.5.2 Dynamic window offset adjustment: Basis: The window offset is dynamically calculated based on the predicted positions of the typhoon centers from the previous two typhoons.

[0060] Objective: To ensure that the window is always focused on the core area of ​​the typhoon and to improve the targeting of feature extraction.

[0061] 2.5.3 Boundary constraint check: Range restrictions: Ensure that the grid index of the window does not exceed the data range, i.e., 0 ≤ lat_idx < 241, 0 ≤ lon_idx < 481.

[0062] Handling method: When the window approaches the data boundary, the offset is automatically adjusted to avoid exceeding the boundary.

[0063] 2.5.4 Construction of the Spacetime Cube: Output: Forms an 11×40×40 three-dimensional tensor, which serves as the input unit for the MSF-Net deep learning model.

[0064] Significance: The input units of the MSF-Net deep learning model include information in both spatial and temporal dimensions, which helps the model capture the dynamic evolution characteristics of typhoons.

[0065] 2.6 Multi-scale feature fusion network MSF-Net inference layer, such as Figure 5 As shown: This layer receives input units from the MSF-Net deep learning model, performs inference through the MSF-Net network backbone, and outputs a relative position index.

[0066] 2.6.1 Primary Feature Encoding: Structure: The texture features of the meteorological field are extracted step by step through three convolutional operations (Conv0-Conv2).

[0067] Parameter changes: The number of output channels gradually increases from 11 to 128, while the spatial resolution gradually decreases to 1 / 8 of the original size.

[0068] Activation function: The ReLU activation function is used to increase the nonlinear expressive power of the network.

[0069] 2.6.2 Deep Feature Enhancement: Gaussian noise injection: Gaussian noise (σ=0.1) is introduced in the 3rd and 4th convolutional layers (Conv3-Conv4).

[0070] Residual connection: By adding input features to output features through cross-layer connections, gradient vanishing is prevented and training convergence is accelerated.

[0071] Objective: To enhance the robustness of the model to noise in the input data and improve the stability of feature extraction.

[0072] 2.6.3 Global Context Fusion: Method: The final convolutional layer (Conv5) uses a 1×1 convolutional kernel to integrate multi-scale features.

[0073] Output: Generates a 1024-dimensional global feature vector to capture the overall features of the typhoon.

[0074] 2.6.4 Feature Compression: Structure: The 1024-dimensional features are compressed through a 4-layer fully connected network (FC1-FC4).

[0075] Target dimension: Ultimately compressed to 128 dimensions, reducing computational load while retaining key information.

[0076] Activation function: Each layer is followed by a ReLU activation function to maintain the nonlinear characteristics of the network.

[0077] 2.6.5 Relative Position Index Prediction: Output layer design: Instead of using an activation function, we directly regress the typhoon center within the relative window using MSF-Net deep learning, and output the relative position index (lat_idx_rel, lon_idx_rel).

[0078] Advantages: It avoids the nonlinear distortion that activation functions may introduce, thus improving the accuracy of location prediction.

[0079] 2.7 Path coordinate generation layer, such as Figure 4 As shown: Input: relative position index, boundary index (initial latitude and longitude boundary index or updated latitude and longitude boundary index).

[0080] 2.7.1 Relative Position Index Decoding: formula: Latitude: lat_idx_pre = lat_idx_rel + lat_idx_min; Longitude: lon_idx_pre = lon_idx_rel + lon_idx_min; Wherein, lat_idx_rel: The relative latitude row number of the typhoon center predicted by the model within the sliding window. lon_idx_rel: The relative longitude column number of the typhoon center predicted by the model within the sliding window. lat_idx_min: The minimum latitude boundary row number of the current sliding window in the global grid. lon_idx_min: The minimum longitude boundary column number of the current sliding window in the global grid. lat_idx_pre: The calculated predicted latitude row number of the typhoon center in the global grid. lon_idx_pre: The calculated predicted longitude column number of the typhoon center in the global grid.

[0081] Objective: To convert the position index within a relative window into an updated latitude and longitude index.

[0082] 2.7.2 Typhoon Center Coordinates Updated: Operation: Use the updated latitude and longitude indices lat_idx_pre and lon_idx_pre obtained from decoding as the new typhoon center coordinates, and calculate the updated latitude and longitude boundary indices accordingly.

[0083] Looping Update: The updated coordinates and boundary indices will be used as input for the next loop update, enabling continuous path tracking.

[0084] The advantages and beneficial effects of this invention are as follows: (1) A spatiotemporal sliding window tracking mechanism is adopted to focus on the core area of ​​the typhoon by dynamically adjusting the window position, effectively capturing spatiotemporal local features.

[0085] (2) Using convolutional neural networks (CNN) to extract meteorological field features, compared with traditional methods, it can more accurately predict the sudden changes and turning points of typhoon paths.

[0086] (3) Utilize GPU to accelerate the training process, while ensuring feature integrity and significantly improving computational efficiency to meet real-time requirements.

[0087] (4) For the typhoon path tracking task, a special network structure including a Gaussian noise enhancement layer, a multi-level downsampling module and residual connections is designed to inject Gaussian noise to enhance the robustness of the model to data noise and enhance the generalization ability.

[0088] (5) Progressive feature compression uses 5 levels of convolution to gradually compress the 11-channel input to a 2-channel offset output, reducing computation while retaining the feature information that is crucial for typhoon path prediction.

[0089] (6) The geographic coordinate inverse mapping algorithm converts the grid index into latitude and longitude coordinates, supports non-uniform grid mapping, and covers the area from latitude 60°N to 0° and longitude 60°E to 180°E, meeting the needs of typhoon tracking in the Northwest Pacific.

[0090] Example 1 like Figure 12As shown, a deep learning model architecture specifically designed for processing meteorological element data is presented. The processing flow begins with multi-channel meteorological element input data organized in a two-dimensional grid, with different colors representing the numerical distribution of different meteorological elements. The data first enters the deep feature encoding module, which uses carefully designed downsampling ratios of 11, 32, 64, 128, 256, 512, and 1024 to gradually reduce spatial resolution while extracting multi-level feature representations using operations such as convolutional and pooling layers, achieving hierarchical capture from local details to global features. Subsequently, the primary feature encoding module further refines and reduces the dimensionality of the deep features, laying the foundation for subsequent processing. The processing flow then enters the crucial global context fusion module, which fuses local features with overall meteorological model information through mechanisms such as global average pooling, effectively enhancing the model's understanding of the macroscopic structure of the atmospheric system. Finally, the feature compression module uses techniques such as multilayer perceptrons to compress high-dimensional features into compact latent representations and generates the typhoon center relative position index (lat_idx_rel, lon_idx_rel).

[0091] like Figure 11 As shown in the figure, the complete closed-loop process of the deep learning-based intelligent typhoon path tracking method in this invention is clearly illustrated. Its technical logic is as follows: First, the system takes the known initial state of the typhoon (center_init) as input, which includes precise latitude and longitude coordinates and the corresponding meteorological element observation area bounding box. Then, the multi-channel meteorological element field within the bounding box is input into a pre-trained multi-scale feature fusion network (MSF-Net). This network, through its unique encoder-decoder structure and feature pyramid design, can simultaneously capture the macroscopic environmental field features and local eye region structural features of the typhoon system, and outputs the predicted typhoon center position and range bounding box (center_next) for the next time period. This prediction result is then fed back to the model input as a new initial state, forming a recursive inference chain of "state input - feature extraction - position prediction - state update". Through this sequential iterative prediction mechanism, the system can automatically complete the continuous tracking of the typhoon's movement path until the end of the entire observation period, ultimately generating a continuous, smooth, and physically meaningful typhoon path trajectory.

[0092] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An artificial intelligence-driven method for dynamic tracking of typhoon paths, characterized in that: Includes the following steps: S1. Train the MSF-Net model, a multi-scale feature fusion network; The MSF-Net model includes the following modules: a primary feature encoding module, used to extract texture features of the input meteorological field step by step through multi-level convolutional operations; a deep feature enhancement module, used to introduce Gaussian noise into the convolutional layers and add the input features and output features through cross-layer connections; a global context fusion module, used to integrate multi-scale features using convolutional kernels and generate a global feature vector; a feature compression module, used to compress the global feature vector through a fully connected network; and a relative position index prediction module, used to output the relative position index of the typhoon center. S2. Based on the original numerical weather forecast data, the meteorological data processing layer performs meteorological data fusion and grid resampling, and outputs the processed numerical weather forecast data. S3. Based on the initial typhoon information, the initial coordinate processing layer determines the grid index and boundary range of the typhoon center; the geographic coordinate mapping layer establishes a two-way conversion relationship between geographic coordinates and grid index. S4. Iterate and update the typhoon path tracking coordinates in a loop; S5. Output typhoon path tracking data and perform Gaussian smoothing and quality control processing on the typhoon path tracking data. In step S3, based on the initial typhoon information, the initial coordinate processing layer determines the grid index and boundary range of the typhoon center, including: Based on the typhoon's starting location, the initial coordinate processing layer generates the initial latitude and longitude index of the typhoon center and the initial latitude and longitude boundary index. In step S4, the typhoon path tracking coordinates are updated iteratively, including: S41. Based on the processed numerical weather forecast data, the latitude and longitude index of the typhoon center, and the latitude and longitude boundary index, a sliding window extraction layer is used to extract the local meteorological field. S42. Based on the extracted local meteorological field, the MSF-Net inference layer of the multi-scale feature fusion network predicts the relative position shift of the typhoon center. S43. Based on the predicted relative position of the typhoon center, the path coordinate generation layer updates the latitude and longitude index and latitude and longitude boundary index of the typhoon center. S44. The updated latitude and longitude indexes of the typhoon center and the latitude and longitude boundary indexes are fed back into the sliding window extraction layer for the next round of iteration.

2. The artificial intelligence-driven dynamic tracking method for typhoon paths according to claim 1, characterized in that: In step S1, the multi-scale feature fusion network MSF-Net model is trained, including: S11. Divide the typhoon path data into training set, validation set and test set; S12. Using the training set, validation set, and test set, the MSF-Net model of the multi-scale feature fusion network is learned, optimized, and evaluated. S13. Save the MSF-Net model of the multi-scale feature fusion network.

3. The method for dynamic tracking of typhoon paths driven by artificial intelligence according to claim 1, characterized in that: In step S2, the meteorological data processing layer performs meteorological data fusion and grid resampling, including: S21. Meteorological data fusion: Integrating meteorological parameter channels covering near-surface meteorological element fields and upper-air meteorological element fields; S22. Grid resampling processing: The meteorological data is resampled to a fixed grid of a set size using a bilinear interpolation algorithm, covering the specified latitude and longitude area.

4. The artificial intelligence-driven dynamic tracking method for typhoon paths according to claim 1, characterized in that: In step S41, the sliding window extraction layer extracts the local meteorological field, including: S411, Set window size; S412. Dynamically adjust the window offset based on the predicted location of the typhoon center; S413 Boundary constraint check: When the distance between any boundary of the window and the boundary of the global data area is less than 10 grid units, the offset is automatically adjusted; S414. Extract local meteorological field: Output a three-dimensional tensor and use it as the input unit of the MSF-Net deep learning model. The input unit of the MSF-Net deep learning model includes meteorological field information in spatial and temporal dimensions.

5. The artificial intelligence-driven dynamic tracking method for typhoon paths according to claim 1, characterized in that: In step S42, the MSF-Net inference layer of the multi-scale feature fusion network predicts the relative position shift of the typhoon center, including: S421. Based on the input unit of the MSF-Net deep learning model, the MSF-Net network backbone of the multi-scale feature fusion network MSF-Net inference layer performs the inference process; S422, The MSF-Net inference layer of the multi-scale feature fusion network outputs the relative position index of the typhoon center; In step S421, the MSF-Net network backbone performs the inference process, including: S4211, Primary Feature Encoding: Extracting meteorological field texture features step by step through multi-level convolution operations; S4212, Deep Feature Enhancement: Introduce Gaussian noise into the convolutional layer; add the input features and output features through cross-layer connections; S4213, Global Context Fusion: Convolutional kernels are used to integrate multi-scale features and generate a global feature vector; S4214. Feature Compression: Compress the global feature vector through a fully connected network.

6. The artificial intelligence-driven dynamic tracking method for typhoon paths according to claim 1, characterized in that: In step S43, the path coordinate generation layer updates the latitude and longitude indexes of the typhoon center and the latitude and longitude boundary indexes, including: S431. Decode relative position index: Convert the position index within the relative window into a latitude and longitude index; S432. Update typhoon center coordinates: Use the decoded latitude and longitude index as the new typhoon center coordinates, and calculate the latitude and longitude boundary index of the typhoon center based on the latitude and longitude index of the typhoon center.

7. An AI-driven typhoon path dynamic tracking system, applied to the AI-driven typhoon path dynamic tracking method described in any one of claims 1-6, characterized in that: This includes the operating environment, data control program, and typhoon location program; The operating environment is either a CPU operating environment or a GPU operating environment; The data control program includes an initial coordinate processing layer, a meteorological data processing layer, and a geographic coordinate mapping layer. The initial coordinate processing layer is used to determine the grid index and boundary range based on the typhoon's starting location. The meteorological data processing layer is used to fuse multi-source meteorological data and perform grid resampling. The geographic coordinate mapping layer is used to establish a bidirectional conversion relationship between geographic coordinates and grid indexes. The typhoon location procedure includes a sliding window extraction layer, a multi-scale feature fusion network MSF-Net inference layer, and a path coordinate generation layer. The sliding window extraction layer is used to extract the local meteorological field, the multi-scale feature fusion network MSF-Net inference layer is used to predict the relative position shift of the typhoon center, and the path coordinate generation layer is used to convert the relative position into absolute coordinates and update the latitude and longitude index and latitude and longitude boundary index.

Images