A storm disaster early warning method and system based on storm disaster bearing coefficient

Abstract

This invention relates to the field of disaster forecasting and early warning technology, and discloses a method and system for early warning of rainstorm disasters based on the rainstorm disaster bearing capacity coefficient. The method includes the following steps: acquiring multi-source data and basic geographic information of the warning area; extracting spatiotemporal features based on the multi-source data and basic geographic information; and calculating the gridded forecast rainfall within a predetermined time period based on the spatiotemporal features; acquiring historical rainstorm disaster data of the warning area; calculating the gridded rainstorm disaster bearing capacity coefficient based on the historical rainstorm disaster data and basic geographic information; establishing an empirical model based on the historical rainstorm disaster data and basic geographic information to calculate the gridded rainfall threshold; and obtaining the rainstorm disaster prediction result of the warning area based on the gridded forecast rainfall, the gridded rainstorm disaster bearing capacity coefficient, the gridded rainfall threshold, and the on-site measured rainfall, combined with the algorithm model. This method solves the problem of weak monitoring capabilities in existing disaster forecasting and early warning systems.

Description

Technical Field

[0001] This invention relates to the field of weather forecasting technology, and in particular to a method for early warning of rainstorm disasters based on the rainstorm disaster bearing coefficient. Background Technology

[0002] Rainfall is a perfectly normal meteorological phenomenon. However, during the rainy season, due to its extreme nature and high intensity in a short period of time, it can easily trigger torrential rain and floods. Torrential rain and floods are often accompanied by geological disasters such as flash floods, landslides, and mudslides, resulting in huge economic losses.

[0003] Existing disaster early warning technologies mainly rely on weather forecasts for protection and early warning. Current weather forecasts primarily predict common weather conditions. However, when facing extreme weather events, because weather forecasts are based solely on meteorological data, they cannot accurately target areas where disasters may occur. Their monitoring and forecasting capabilities are weak and lack precision. Consequently, they cannot effectively and objectively assess whether a region will experience flooding, and therefore cannot effectively provide early warnings for flooding disasters. Summary of the Invention

[0004] This invention provides a method and system for early warning of rainstorm disasters based on the rainstorm disaster bearing coefficient, in order to solve the problem of weak monitoring capabilities in existing disaster forecasting and early warning systems.

[0005] To achieve the above objectives, the present invention employs the following technical solution:

[0006] In a first aspect, the present invention provides a method for early warning of rainstorm disasters based on the rainstorm disaster bearing coefficient, comprising the following steps:

[0007] Step 1: Obtain multi-source data and basic geographic information of the warning area, extract spatiotemporal features based on the multi-source data and basic geographic information, and calculate the gridded forecast rainfall within the predetermined time based on the spatiotemporal features;

[0008] Step 2: Obtain historical rainstorm disaster data for the warning area, calculate the grid point rainstorm disaster bearing coefficient based on the historical rainstorm disaster data and basic geographic information, and establish an empirical model based on the historical rainstorm disaster data and basic geographic information to calculate the grid point rainfall threshold;

[0009] Step 3: Based on the grid-based forecast rainfall, grid-based rainstorm disaster bearing capacity coefficient, grid-based rainfall threshold, and on-site measured rainfall, the algorithm model is combined to obtain the rainstorm disaster prediction results for the warning area.

[0010] Furthermore, the multi-source data includes correction data and measurement data;

[0011] The measurement data includes satellite data, ground observation data, radar data, and sounding data.

[0012] Furthermore, the extraction of spatiotemporal features based on multi-source data and basic geographic information includes: extracting spatial sequences, time sequences, and image data of rainfall based on multi-source data combined with spatial sequence features; encoding the spatial sequences, time sequences, and image data; and gridding the basic geographic information to obtain gridded geographic information.

[0013] Multi-source vector data is obtained by encoding gridded geographic information, coded spatial sequences, time series, image data, and forecast correction data, and spatiotemporal features are extracted from the multi-source vector data.

[0014] Furthermore, the extraction of spatiotemporal features based on multi-source vector data includes: obtaining corresponding tensor data through algorithmic encoding based on multi-source vector data, and then optimizing the tensor data to obtain spatiotemporal features by combining a temporal attention mechanism.

[0015] Furthermore, the multi-source vector data employs multimodal feature extraction, and the multi-source vector data includes a vector matrix, a physical constraint matrix, and an image feature matrix.

[0016] Furthermore, the step of calculating the grid-point forecast rainfall within a predetermined time period based on spatiotemporal characteristics includes: performing feature decoding based on spatiotemporal characteristics using a deep temporal neural network to calculate the grid-point forecast rainfall within a predetermined time period.

[0017] Furthermore, in step 2, the calculation of the gridded rainstorm disaster bearing coefficient based on historical rainstorm disaster data and basic geographic information includes: analyzing historical rainstorm disaster data, obtaining rainstorm disaster factors through machine learning, gridding the basic geographic information, obtaining gridded geographic information, fusing the rainstorm disaster factors and gridded geographic information, and calculating the gridded rainstorm disaster bearing coefficient using the deterministic coefficient method.

[0018] Furthermore, the step of establishing an empirical model based on historical rainstorm disaster data and basic geographic information to calculate the grid point rainfall threshold includes: obtaining rainfall data of rainstorm meteorological disasters based on the historical disaster data, gridding the basic geographic information to obtain grid point geographic information, establishing an empirical model based on the rainfall data and grid point geographic information, and calculating the grid point rainfall threshold.

[0019] Furthermore, in step 3, the algorithm model calculates the probability of disaster at grid points within the warning area based on the grid point forecast rainfall, the grid point rainstorm disaster bearing coefficient, the grid point rainfall threshold, and the on-site measured rainfall, and issues a warning signal.

[0020] Secondly, the present invention provides a rainstorm disaster early warning system based on the rainstorm disaster bearing coefficient, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method described above.

[0021] Beneficial effects:

[0022] This invention provides a rainstorm disaster early warning method and system based on the rainstorm disaster bearing coefficient. By calculating gridded forecast rainfall based on basic geographic information and multi-source data, and combining the gridded disaster bearing coefficient calculated from historical rainstorm disaster data, the gridded rainfall threshold, and the rainfall measured on-site, the method uses an empirical model from multiple data perspectives to conduct disaster early warning, thereby improving the precision and accuracy of disaster early warning. Attached Figure Description

[0023] Figure 1 This is a flowchart of a rainstorm disaster early warning method based on the rainstorm disaster bearing coefficient according to the present invention;

[0024] Figure 2 This is a schematic diagram of the structure of the AutoEncoder network used in this invention. Detailed Implementation

[0025] The technical solution of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms "an" or "a" and similar terms do not indicate a quantity limitation, but rather indicate the presence of at least one. The terms "connected" or "linked" and similar terms are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up," "down," "left," "right," etc., are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship also changes accordingly.

[0027] Please see Figure 1 This invention provides a method for early warning of rainstorm disasters based on the rainstorm disaster bearing coefficient, comprising the following steps:

[0028] Step 1: Obtain multi-source data and basic geographic information of the warning area, extract spatiotemporal features based on the multi-source data and basic geographic information, and calculate the gridded forecast rainfall within the predetermined time based on the spatiotemporal features;

[0029] Multi-source data includes correction data and measurement data;

[0030] The measurement data includes satellite data, ground observation data, radar data, and radiosonde data;

[0031] The correction data from multi-source data mainly consists of grid forecast correction products, which include precipitation, phase, temperature, UV wind, relative humidity, thunderstorms, short-duration heavy precipitation, hail, and thunderstorm winds. The data is primarily obtained from hourly precipitation meteorological element products over 24 hours output from a high-resolution rapid cyclic assimilation numerical weather prediction system, multi-source observation data, and kilometer-level multi-source data from radar and satellite. In this embodiment, 14-channel radar data and geographic information were collected at four time points (1, 2, 3, and 4) at four locations (A, B, C, and D) in the warning area. Please refer to Tables 1 and 2 for details.

[0032] Table 1: Collected Radar Data

[0033]

[0034] Table 2: Collected Geographic Information

[0035]

[0036] The measurement data is integrated based on meteorological geographic information data metadata files, including: vector format meteorological map base maps, topographic data at different scales, 30-meter land cover data, township boundary data, global topographic data, global 90-meter DEM data, and comprehensive thematic data. For insights into the correspondence between meteorological data and geographic locations, please refer to [link to relevant documentation]. Figure 2 The AutoEncoder network is used to encode the absolute geographic location, and the specific geographic location is used as an important feature factor. An efficient encoding and semantic extraction model is obtained, which maps high-precision terrain information into high-dimensional feature vectors, constructs a high-dimensional terrain feature vector dataset, and forms a reusable feature vector dataset.

[0037] Specifically, based on multi-source data and spatial sequence feature extraction, spatial, temporal, and image data of rainfall are obtained. The spatial, temporal, and image data are encoded, and the basic geographic information is gridded to obtain gridded geographic information.

[0038] Based on gridded geographic information, encoded spatial sequences, time series, and image data, forecast correction data are encoded to obtain multi-source vector data. Based on the multi-source vector data, corresponding tensor data are obtained through algorithmic encoding. Then, combined with a temporal attention mechanism, the tensor data is optimized to obtain spatiotemporal features.

[0039] The multi-source vector data employs a multimodal feature extraction method, which includes a vector matrix, an image feature matrix, and a physical constraint representation matrix. The vector matrix is ​​obtained by encoding measurement data, the image feature matrix is ​​mainly derived from encoding weather radar mosaic data, and the physical constraint representation matrix is ​​a probability mask matrix between 0 and 1 formed by encoding the physical constraints between surface runoff, reservoir capacity, water level, precipitation, and river runoff.

[0040] Spatiotemporal feature extraction simultaneously acquires feature information from spatial sequence data in both the temporal and spatial dimensions. This information refers to the tensor data describing the spatiotemporal features obtained after encoding each set of multi-source fused data through an algorithm. The hybrid encoder is divided into a spatial transformer block and a temporal converter block. The spatial transformer block focuses on calculating the self-attention between data points, aiming to learn the relationships between data at each time point and promote the correlation between geographic information and meteorological data. The temporal converter block calculates the self-attention between time points, focusing on learning the temporal correlation information of the fused data, making the acquired data more accurate and effectively used for disaster early warning.

[0041] Based on spatiotemporal characteristics, feature decoding is performed using a deep temporal neural network to calculate the grid-point forecast rainfall within a predetermined time period.

[0042] A hybrid spatiotemporal coding approach is used to construct a feature extractor, which considers both the intrinsic relationships between different data types at each time point and the temporal relationships between the data. Based on this spatiotemporal feature extractor, the accuracy of rainfall prediction is improved.

[0043] Spatial correlation learning using hybrid spatiotemporal coding: First, a linear embedding layer projects various data at each time point onto a high-dimensional feature. Then, the spatial representation of this node is fed back into a spatial self-attention mechanism to simulate the correlation between all data points and output a high-dimensional representation.

[0044] Temporal Correlation Learning with Hybrid Spatiotemporal Encoding: To inject effective motion trajectories into the learned representation, this embodiment considers the temporal correspondence of data to model the correlation of the same joints over long time series. Different categories of data are separated along the temporal dimension, making the evolution of each node's data a separate label, and different categories of data in the fused data are modeled in parallel. From a temporal perspective, the different evolutions of various types of data in the fused data are modeled separately to better represent temporal correlation. Furthermore, treating each category of data in the fused data as a separate label reduces the model's dimensionality from N×dim to dim, and also allows for the processing of longer sequences within the model.

[0045] The attention calculation for the query Q, keyword K, and value matrix V in each header of the transformer block in MixSTE is represented by the following formula;

[0046] ;

[0047] in , Represents the attention mechanism, The function represents the normalization exponential function, and N represents the types of data to be fused. Here, T represents the dimension of each data element, and T stands for matrix transpose. Multi-head attention is defined as follows:

[0048] ;

[0049] ;

[0050] Where the linear projection weight is MSA represents multi-head self-attention mechanism, Concat represents merging and concatenation, and h represents the number of independent groups. In the Transformer encoder of this embodiment, each joint token... From low dimension The Union Projection. Joint label p through matrix Embedded location information:

[0051] ;

[0052] The above The matrix dimensions are transformed according to the different task requirements of modeling spatial correlation and temporal correlation. Representative level standardization, This represents a linear embedding layer.

[0053] The feature decoding structure uses the deep temporal model LSTNet, which adds a traditional autoregressive linear model to the nonlinear neural network part, making the nonlinear deep learning model more robust to time series with scale violations.

[0054] The structure consists of five main modules: convolutional components, recursive components, recursive-skip layers, temporal attention layers, and autoregressive components.

[0055] Convolutional Component: This component is a convolutional network without pooling, designed to extract short-term patterns over time and local dependencies between variables.

[0056] Recursive Component: A recursive component is a gated recurrent unit (GRU) that uses the ReLU function as the activation function for updating the hidden state. Its purpose is to extract the temporal features from the features extracted by the convolutional layers using the GRU.

[0057] Recursive-skipped layers: Traditional GRUs have difficulty capturing long-term patterns, so skipped connection layers can be used. That is, by sampling at intervals, we can look back at a longer time while keeping the sampling sequence length unchanged, so as to capture long-term features.

[0058] Temporal Attention Layer: Recursive-skip layers require a predefined hyperparameter, which is disadvantageous for aperiodic or time series with dynamically changing period lengths. To mitigate this issue, this embodiment employs an alternative approach: an attention mechanism. It learns a weighted combination of hidden representations at each window position of the input matrix.

[0059] Autoregressive Component: Due to the non-linear nature of convolutional and recursive components, a major drawback of neural network models is that the output scale is insensitive to the input scale. In specific real-world datasets, the scale of the input signal changes continuously in a non-periodic manner, significantly reducing the prediction accuracy of neural network models. The final prediction of LSTNet can be decomposed into a linear part and a non-linear part, where the linear part mainly focuses on local scale issues, while the non-linear part focuses on recursive patterns. In the LSTNet architecture, a classic autoregressive (AR) model is used as the linear component.

[0060] Essentially, it's a linear layer. Finally, the output of the neural network is added to the AR component to obtain the final prediction of LSTNet.

[0061] In this embodiment, based on the above, the grid-based forecast rainfall at 30 minutes, 60 minutes, 90 minutes and 120 minutes is calculated according to the spatiotemporal characteristics. Please refer to Table 3 for details.

[0062] Table 3: Forecast rainfall at grid points within the predetermined time period

[0063]

[0064] Step 2: Obtain historical rainstorm disaster data for the warning area, calculate the grid point rainstorm disaster bearing coefficient based on the historical rainstorm disaster data and basic geographic information, and establish an empirical model based on the historical rainstorm disaster data and basic geographic information to calculate the grid point rainfall threshold;

[0065] Specifically, rainfall data for rainstorm meteorological disasters is obtained based on historical rainstorm disaster data. The basic geographic information is gridded to obtain gridded geographic information. An empirical model is established based on the rainfall data and the gridded geographic information to calculate the rainfall threshold at each grid point. Please refer to Table 4 for details.

[0066] Table 4: Calculated Grid Point Rainfall Thresholds

[0067]

[0068] Meteorological disasters such as floods, landslides, and debris flows are strongly correlated with precipitation. Utilizing regional rainfall characteristics to forecast and warn of large-scale rainstorm meteorological disasters is an important approach to preventing such disasters. The most critical issue is determining the critical rainfall threshold that triggers these disasters. Historical statistical methods are used to establish empirical models for studying the critical rainfall threshold for rainstorm meteorological disasters.

[0069] The critical rainfall threshold for disasters is determined by classifying the terrain and landforms. Therefore, gridded geographic information is incorporated into the model calculation. Based on the precipitation data of all short-term historical rainstorm meteorological disasters, the correlation between disasters and precipitation is analyzed, and the critical rainfall threshold for disasters is assigned to the gridded geographic information according to the precipitation and correlation at the disaster site.

[0070] The calculation of gridded rainstorm disaster bearing capacity based on historical rainstorm disaster data and basic geographic information includes: analyzing historical rainstorm disaster data, obtaining rainstorm disaster factors through machine learning, gridding basic geographic information, obtaining gridded geographic information, fusing rainstorm disaster factors and gridded geographic information, and calculating gridded rainstorm disaster bearing capacity based on the deterministic coefficient method.

[0071] Specifically, by analyzing historical rainstorm disaster data, rainstorm disaster factors are obtained through machine learning Apriori correlation analysis algorithm;

[0072] The coefficient of determination (CF) method is a probability function, generally used to determine the contribution rate of various indicators when an event occurs. The calculation formula is as follows:

[0073] ;

[0074] In the formula: It represents the conditional probability of a certain disaster occurring in influence factor a, usually expressed as the ratio of the number of disasters occurring at a certain grid point in influence factor a to the area of ​​that grid point. It represents the probability of a certain disaster occurring in a region, usually calculated by dividing the total number of disasters occurring in the entire region by the total area of ​​the region. The value ranges from [-1, 1]. Positive values ​​indicate that the factor is positively correlated with disasters, while negative values ​​indicate the opposite.

[0075] Determine each factor After the value, all factors are... The total contribution value is calculated by summing the values ​​at each level. .

[0076] Note: Step-by-step stacking refers to sequentially stacking each factor. The value is added to the total contribution to reflect the combined effect among factors.

[0077] ;

[0078] in, Indicates total contribution. Indicates the number of impact factors. Represents the CF values ​​for different factors;

[0079] Then, the independent contribution of each factor is quantified using an "exclusion method." First, for each factor, the total contribution value after removing that factor is calculated. .

[0080] ;

[0081] Then the relative contribution value of this factor for:

[0082] ;

[0083] Finally, the relative contribution values ​​are normalized into weights.

[0084] ;

[0085] in, This represents the weights after normalization. Indicates the weights of different factors. This represents the CF value for different factors.

[0086] The regional disaster occurrence coefficient H is obtained by multiplying the CF values ​​of each factor by their weights and then summing the results.

[0087] ;

[0088] The coefficient ranges from [−1, 1], and the larger the value, the higher the probability of a disaster occurring.

[0089] Step 3: Based on the grid-based forecast rainfall, grid-based rainstorm disaster bearing capacity coefficient, grid-based rainfall threshold, and on-site measured rainfall, the algorithm model is combined to obtain the rainstorm disaster prediction results for the warning area.

[0090] Specifically, please refer to Table 5, which uses the algorithm model to obtain the rainstorm disaster prediction results for the four locations A, B, C, and D in the warning area;

[0091] Table 5: Rainstorm Disaster Forecast Results

[0092]

[0093] The algorithm model calculates the probability of disaster at grid points within the warning area based on grid point forecast rainfall, grid point rainstorm disaster bearing coefficient, grid point rainfall threshold, and on-site measured rainfall, and issues a warning signal.

[0094] In this embodiment, the model combines the on-site measured rainfall at a grid point with the rolling output value of the precipitation forecast model to calculate and determine the probability that the total expected rainfall within the grid point will reach the grid point rainfall threshold.

[0095] ;

[0096] ;

[0097] Among them, indicator function Defined as:

[0098] ;

[0099] In the above formula, The probability of precipitation-related disasters. This represents the sum of the measured precipitation and the predicted precipitation for each time period, where T represents the grid point precipitation threshold. For on-site measurement points On-site measurement of rainfall, The grid point rainstorm disaster bearing coefficient, The hourly rainfall forecast is generated starting from the current time, where M is the total number of field observation points within the grid.

[0100] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A method for early warning of rainstorm disasters based on the rainstorm disaster bearing coefficient, characterized in that, Includes the following steps: Step 1: Obtain multi-source data and basic geographic information of the warning area, extract spatiotemporal features based on the multi-source data and basic geographic information, and calculate the gridded forecast rainfall within the predetermined time based on the spatiotemporal features; Step 2: Obtain historical rainstorm disaster data for the warning area, calculate the grid point rainstorm disaster bearing coefficient based on the historical rainstorm disaster data and basic geographic information, and establish an empirical model based on the historical rainstorm disaster data and basic geographic information to calculate the grid point rainfall threshold; Step 3: Based on the grid-based forecast rainfall, grid-based rainstorm disaster bearing capacity coefficient, grid-based rainfall threshold, and on-site measured rainfall, the algorithm model is combined to obtain the rainstorm disaster prediction results for the warning area; The algorithm model calculates the probability of disaster at grid points within the warning area based on grid point forecast rainfall, grid point rainstorm disaster bearing coefficient, grid point rainfall threshold, and on-site measured rainfall, and issues a warning signal. This can be expressed by the following formula: ； ； in, Indicates the on-site measurement points On-site measurement of rainfall; The grid point represents the disaster tolerance coefficient for heavy rain. Indicates the total duration of rainfall; This indicates the hourly grid-based rainfall forecast starting from the current time. The probability of precipitation-related disasters. This represents the sum of the measured precipitation and the predicted precipitation for each time period. Indicates the threshold for rainfall at each grid point; The indicator function is represented by the following formula: 。 2. The rainstorm disaster early warning method based on the rainstorm disaster bearing coefficient according to claim 1, characterized in that, The multi-source data includes correction data and measurement data; The measurement data includes satellite data, ground observation data, radar data, and sounding data.

3. The rainstorm disaster early warning method based on the rainstorm disaster bearing coefficient according to claim 2, characterized in that, The extraction of spatiotemporal features based on multi-source data and basic geographic information includes: obtaining spatial, temporal, and image data of rainfall based on multi-source data combined with spatial sequence feature extraction methods; encoding the spatial, temporal, and image data; and gridding the basic geographic information to obtain gridded geographic information. Multi-source vector data is obtained by encoding gridded geographic information, coded spatial sequences, time series, image data, and forecast correction data, and spatiotemporal features are extracted from the multi-source vector data.

4. The rainstorm disaster early warning method based on the rainstorm disaster bearing coefficient according to claim 3, characterized in that, The extraction of spatiotemporal features based on multi-source vector data includes: obtaining corresponding tensor data through algorithmic encoding based on multi-source vector data, and then optimizing the tensor data to obtain spatiotemporal features by combining a temporal attention mechanism.

5. The rainstorm disaster early warning method based on the rainstorm disaster bearing coefficient according to claim 4, characterized in that, The multi-source vector data employs multimodal feature extraction, and the multi-source vector data includes a vector matrix, a physical constraint matrix, and an image feature matrix.

6. The rainstorm disaster early warning method based on the rainstorm disaster bearing coefficient according to any one of claims 1-5, characterized in that, The calculation of grid-point forecast rainfall within a predetermined time period based on spatiotemporal characteristics includes: performing feature decoding based on spatiotemporal characteristics using a deep temporal neural network to calculate the grid-point forecast rainfall within a predetermined time period.

7. The rainstorm disaster early warning method based on the rainstorm disaster bearing coefficient according to any one of claims 1-5, characterized in that, In step 2, the calculation of the gridded rainstorm disaster bearing coefficient based on historical rainstorm disaster data and basic geographic information includes: analyzing historical rainstorm disaster data, obtaining rainstorm disaster factors through machine learning, gridding the basic geographic information, obtaining gridded geographic information, fusing the rainstorm disaster factors and gridded geographic information, and calculating the gridded rainstorm disaster bearing coefficient using the deterministic coefficient method.

8. The rainstorm disaster early warning method based on the rainstorm disaster bearing coefficient according to any one of claims 1-5, characterized in that, The step of establishing an empirical model based on historical rainstorm disaster data and basic geographic information to calculate the grid point rainfall threshold includes: obtaining rainfall data of rainstorm meteorological disasters based on the historical rainstorm disaster data, gridding the basic geographic information to obtain grid point geographic information, establishing an empirical model based on the rainfall data and grid point geographic information, and calculating the grid point rainfall threshold.

9. A rainstorm disaster early warning system based on a rainstorm disaster bearing coefficient, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

Images