Forest fire monitoring method, device, equipment and storage medium

Abstract

The invention provides a forest fire monitoring method and device, equipment and a storage medium. Relates to the technical field of fire monitoring. The method comprises the following steps: acquiring radar echo data; wherein the radar echo data is data obtained by scanning a forest area by using an X-band radar; preprocessing and repairing the radar echo data to obtain repaired echo data; constructing a machine learning model based on the restored echo data and the extracted feature data, training the machine learning model by using the feature data to obtain a smoke echo recognition model, and in response to the feature data of the input radar echo data, judging whether the current radar echo data is smoke echoes generated by the forest fire; and under the condition of determining that the radar echo data are smoke echoes generated by the forest fire, determining the occurrence position and range of the fire and the fire spreading speed according to the spatial and temporal distribution characteristics of the radar echo data. According to the invention, the forest fire monitoring precision and efficiency can be improved.

Description

Technical Field

[0001] The present application relates to the field of fire monitoring technology, and in particular to a forest fire monitoring method, device, equipment and storage medium. Background Art

[0002] Forest fires pose a significant threat to the ecological environment, biodiversity, and human life and property. Yunnan, with its rich forest resources, complex terrain, and variable climate, is prone to frequent forest fires. Traditional forest fire monitoring methods, such as satellite monitoring, video surveillance, manual patrols, and aerial patrols, each have their limitations. Satellite monitoring, while covering a wide area, has low temporal resolution, making it difficult to capture subtle changes in the early stages of a fire. Video surveillance, limited by geographical conditions and scope, has numerous blind spots. Manual patrols are inefficient and struggle to provide comprehensive coverage in complex terrain. Aerial patrols are costly, making regular monitoring impossible.

[0003] Current radar research on forest fires is mostly based on operational S-band and C-band radars. Limited by technical systems and detection modes, these radars primarily rely on the radar reflectivity characteristics of forest fire echoes. This makes it difficult to effectively distinguish small, weak, or mixed with ground objects or light precipitation, making automatic identification even more challenging. Furthermore, operational radars are widely spaced. Due to the curvature of the Earth and radar detection methods, the radar beam rises with distance from the radar. Forest fires are generally located at altitudes below 3 km, resulting in a large low-altitude blind spot, which can easily lead to missed fire reports. This radar blind spot is particularly pronounced due to the unique terrain (fragmented plateau terrain).

[0004] X-band radar has high resolution and sensitivity, offering unique advantages in detecting small-scale targets. However, currently, there is no method specifically designed to effectively monitor and infer smoke echoes from forest fires using X-band radar in complex environments, particularly methods for inferring fire area, fire spread rate, and real-time location of fires. Summary of the Invention

[0005] The present application provides a forest fire monitoring method, device, equipment and storage medium to improve the monitoring accuracy and efficiency of forest fires, realize early warning of forest fires, fire development monitoring and fire behavior inversion, and provide scientific basis and technical support for forest fire prevention and control.

[0006] In a first aspect, the present application provides a forest fire monitoring method, comprising:

[0007] Acquiring radar echo data; wherein the radar echo data is data obtained by scanning the forest area using an X-band radar;

[0008] Preprocessing the radar echo data to obtain preprocessed data, performing reflectivity outlier detection and repair and radial velocity outlier detection and repair on the preprocessed data to obtain repaired echo data;

[0009] Extracting characteristic data based on the repaired echo data, the characteristic data including echo intensity, velocity, spectral width and polarization characteristics;

[0010] Constructing a machine learning model, and using the feature data to train the machine learning model to obtain a smoke echo recognition model;

[0011] Based on the echo recognition model, in response to the characteristic data of the input radar echo data, determining whether the current radar echo data is a smoke echo generated by a forest fire;

[0012] When it is determined that the radar echo data is smoke echo generated by a forest fire, the location, scope, and spread speed of the fire are determined based on the spatiotemporal distribution characteristics of the radar echo data.

[0013] In one possible design, a method for detecting and repairing reflectivity outliers on the preprocessed data includes:

[0014] Calculate the mean and standard deviation of the echo intensity of each range library, determine the intensity threshold based on the mean and standard deviation, and determine the data point whose echo intensity exceeds the intensity threshold as a reflectivity outlier. If the intensity value of the data point and the intensity value of the adjacent area exceed a first threshold, then the data point is determined to be a reflectivity outlier;

[0015] For isolated reflectivity outliers, the neighborhood data of the set window size is selected with the outlier as the center, and the median of the neighborhood data is used to replace the outlier to repair the isolated reflectivity outlier. For continuous reflectivity outliers, the intensity value I of the adjacent azimuth and the same distance library is used. j,k and I j',k , using formula I i:i+n,k =I j,k +(ij)×(I j',k -I j,k) / (j'-j) is interpolated to repair the continuous reflectivity anomalies, where j and j' are adjacent azimuths, k is the range library number, and I i:i+n,k The reflectivity values ​​of the k distance libraries from azimuth angles i to i+n.

[0016] In one possible design, a method for detecting and repairing radial velocity outliers on the preprocessed data includes:

[0017] Set the radial velocity range to [-V max ,V max], where V max The maximum radial velocity is determined according to the radar pulse repetition frequency. If the radial velocity exceeds [-V max ,V max ] and there is no support from severe convective meteorological events, the radial velocity value is determined to be the first abnormal radial velocity value; if the radial velocity value is ambiguous and the velocity change in the adjacent distance library or time series exceeds the second threshold, the radial velocity value is determined to be the second abnormal radial velocity value;

[0018] For the first abnormal radial velocity value, the spatiotemporal smoothing filtering method is used. With the first abnormal radial velocity value as the center, a window is selected in the spatiotemporal dimension, and the weighted average value of the data in the window is calculated to replace the abnormal value. For the second abnormal radial velocity value, the defuzzification process is performed on it, and the [-V max ,V max ] Determine whether the data after defuzzification is abnormal.

[0019] In one possible design, when it is determined that the radar echo data is smoke echo generated by a forest fire, the location of the fire is determined by the following formula based on the spatiotemporal distribution characteristics of the radar echo data:

[0020] P(X,Y)=(P1(X1,Y1)-V1(x1,y1)×Δt+P2(X2,Y2)-V2(x2,y2)×Δt+P3(X3,Y3)-

[0021] V3(x3,y3)×Δt+···+P n (X n ,Y n )-V n (x n ,y n )×Δt) / n

[0022] Where P(X,Y) is the fire location, X and Y represent the latitude and longitude respectively; n is the maximum elevation angle at which the X-band radar detects the smoke echo; P n (X n ,Y n ) represents the longitude and latitude position of the echo at the nth elevation angle; V n (x n ,y n ) is the ambient wind at the nth elevation angle, and Δt is the radar detection time interval.

[0023] In one possible design, when it is determined that the radar echo data is smoke echo generated by a forest fire, the fire occurrence range is determined by the following formula based on the spatiotemporal distribution characteristics of the radar echo data:

[0024] S i=Σ(a1×S P1 +a2×S P2 +a3×S P3 +···+a n ×S Pn ) / n

[0025] Where S i represents the fire area at time i, a n is the pyramid layer coefficient, S pn is the area of ​​the radar echo; n is the maximum elevation angle at which the X-band radar detects the smoke echo.

[0026] In one possible design, when it is determined that the radar echo data is smoke echo generated by a forest fire, the fire spread speed is determined by the following formula based on the spatiotemporal distribution characteristics of the radar echo data:

[0027] V f =P*G*M*R*W*V r

[0028] Where V f represents the speed of forest fire spread, P represents the terrain factor, G also represents the vegetation factor, M represents the seasonal factor, W represents the environmental wind factor, and Vr is the speed measured by the X-band radar.

[0029] In one possible design, after determining the location, scope, and spread rate of the fire, the method further includes: predicting the spread rate of the fire using the following formula:

[0030] V f (t) = V f +b t *G t *P t *W t

[0031] Where V f (t) represents the predicted speed of forest fire spread at time t, b t represents the initial velocity of the pattern or model prediction, P t Represents the terrain factor at time t, G t represents the vegetation factor at time t, W t Represents the environmental wind factor at time t.

[0032] In a second aspect, the present application provides a forest fire monitoring device, comprising:

[0033] a data acquisition module configured to acquire radar echo data; wherein the radar echo data is data obtained by scanning the forest area using an X-band radar;

[0034] a data processing module configured to preprocess the radar echo data to obtain preprocessed data, and perform reflectivity outlier detection and repair and radial velocity outlier detection and repair on the preprocessed data to obtain repaired echo data;

[0035] a feature extraction module configured to extract feature data based on the repaired echo data, wherein the feature data includes echo intensity, velocity, spectral width and polarization characteristics;

[0036] a model training module configured to construct a machine learning model and train the machine learning model using the feature data to obtain a smoke echo recognition model;

[0037] an echo recognition module configured to determine whether the current radar echo data is a smoke echo generated by a forest fire based on the echo recognition model and in response to characteristic data of the input radar echo data;

[0038] The monitoring inversion module is configured to determine the location, scope and fire spread speed of the fire based on the spatiotemporal distribution characteristics of the radar echo data when it is determined that the radar echo data is smoke echo generated by a forest fire.

[0039] In a third aspect, an embodiment of the present application provides an electronic device comprising: at least one processor and a memory; the memory stores computer-executable instructions; the at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor executes the forest fire monitoring method described in the first aspect and various possible designs of the first aspect.

[0040] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which stores computer-executable instructions. When a processor executes the computer-executable instructions, the forest fire monitoring method described in the first aspect and various possible designs of the first aspect is implemented.

[0041] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the forest fire monitoring method described in the first aspect and various possible designs of the first aspect.

[0042] The forest fire monitoring method, device, equipment, and storage medium provided in this application have at least the following beneficial effects:

[0043] 1. This application utilizes the high resolution and multi-parameter measurement capabilities of X-band radar, combined with a smoke echo recognition model, to effectively improve the monitoring accuracy of forest fires, accurately identify small-scale, weak-intensity early-stage fire smoke echoes, reduce missed and false alarms, and facilitate real-time dynamic monitoring of fire conditions, providing meteorological data support for subsequent disaster prevention, mitigation, and relief.

[0044] 2. Through real-time monitoring of forest fires and fire behavior inversion, we can timely grasp the occurrence and development of fires, predict the future spread trend of fires, provide a scientific basis for fire fighting decisions, improve the efficiency and effectiveness of fire prevention and control, minimize the losses caused by forest fires, and at the same time ensure the safety of firefighters. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0046] Figure 1 A flowchart of a forest fire monitoring method provided in an embodiment of the present application;

[0047] Figure 2 The original X-band radar echo of the forest fire smoke echo provided in the embodiment of the present application; wherein (a) is at time t0; (b) is at time t1; (c) is at time t2; (d) is at time t3;

[0048] Figure 3 This is a diagram showing the X-band radar recognition results of forest fire smoke echoes provided in an embodiment of the present application; wherein (a) is at time t0; (b) is at time t1; (c) is at time t2; and (d) is at time t3;

[0049] Figure 4 A schematic diagram of forest fire inversion using X-band radar smoke echoes based on a campfire model provided in an embodiment of the present application;

[0050] Figure 5 This is a structural diagram of the forest fire monitoring device provided in an embodiment of the present application.

[0051] The above drawings illustrate specific embodiments of the present application, which will be described in more detail below. These drawings and the textual description are not intended to limit the scope of the present application in any way, but rather to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0052] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0053] In the technical solution of this application, the collection, storage, use, processing, transmission, provision and disclosure of information such as financial data or user data involved comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0054] It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned. They should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0055] The following specific embodiments describe in detail the technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0056] The present application embodiment provides a forest fire monitoring method. Figure 1 As shown, it is a flow chart of the forest fire monitoring method provided in an embodiment of the present application, and the forest fire monitoring method includes the following steps S100-S600.

[0057] S100: Acquire radar echo data; wherein the radar echo data is data obtained by scanning a forest area using an X-band radar.

[0058] In this embodiment, an X-band radar is used to scan the forest area to obtain radar echo data. The X-band radar has high temporal and spatial resolution and multi-parameter detection capabilities, and can accurately capture the characteristic information of smoke echoes.

[0059] In a specific embodiment, X-band radars can be reasonably deployed in forest areas to ensure that the radar can cover the main forest distribution areas. According to the geographical characteristics of the forest area and the areas with high incidence of forest fires, the detection range and scanning mode of the radar are optimized. The radar performs omnidirectional scanning of the monitoring area at set time intervals to collect radar echo data. At the same time, combined with other meteorological observation equipment (such as anemometers, thermometers, hygrometers, etc.) and geographic information data (such as digital elevation models DEM), synchronized environmental wind, terrain factors, climate factors, vegetation factors and other information are obtained to provide auxiliary data for subsequent data processing and analysis. The smoke echo of a forest fire is as follows: Figure 2 shown.

[0060] S200: Preprocessing the radar echo data to obtain preprocessed data, performing reflectivity outlier detection and repair and radial velocity outlier detection and repair on the preprocessed data to obtain repaired echo data.

[0061] Figure 2 This is an example of a typical forest fire smoke echo on a radar echo intensity map. While the smoke echo's characteristics are clearly displayed, it also contains some other clutter, requiring further processing. For example, the collected raw radar echo data can be preprocessed, including noise removal, clutter suppression, and data calibration. By employing an adaptive filtering algorithm, the filter parameters are dynamically adjusted based on the complex geographical and meteorological environment of Yunnan Province, effectively removing interfering signals such as ground clutter, weather echoes, and insect echoes, improving data quality. The resulting preprocessed data can then be further processed for outlier detection and repair, ultimately yielding repaired echo data to facilitate subsequent processing steps.

[0062] During implementation, the raw radar echo data collected in step S100 is transmitted to a data processing center. First, a denoising algorithm based on the smoke echo characteristics of forest fires is used to remove noise from the data. This algorithm effectively preserves the detailed characteristics of the echo signal while suppressing noise interference. Clutter suppression is then performed using clutter map technology. By constructing a clutter map model, different types of clutter (such as ground clutter and meteorological clutter) are identified and removed. In terms of data calibration, the echo data's intensity, velocity, and other parameters are calibrated based on the radar's system parameters and actual observations to ensure data accuracy and reliability.

[0063] In some embodiments, the steps of detecting and repairing reflectivity outliers and detecting and repairing radial velocity outliers on the pre-processed data include:

[0064] S201: Detecting reflectivity outliers.

[0065] Based on the statistical analysis method, the mean μ and standard deviation ρ of the echo intensity of each distance library are calculated. The intensity threshold range is set to [μ-3ρ, μ+3ρ], and data points outside this range are determined to be outliers. Taking into account the interference of ground echoes caused by the mountainous terrain of Yunnan, the threshold coefficient is appropriately adjusted for the radar beam coverage area close to the mountains (such as adjusting 3ρ to 2.5ρ) to more accurately identify anomalies. A secondary judgment is made using the echo intensity correlation of adjacent distance libraries and azimuths. If the difference between the intensity value of a data point and the intensity value of the adjacent area exceeds the set threshold (such as the intensity difference between adjacent data points is greater than 10dBZ), and the difference cannot be explained by meteorological or geographical factors (such as non-significant precipitation or the edge of the ground object), it is considered an anomaly.

[0066] S202: Repair of reflectivity outliers.

[0067] For isolated outliers, the median filtering method is used. With the outlier as the center, select the neighborhood data of a certain window size (such as 3×3 distance library and azimuth range), and replace the outlier with the median of the neighborhood data. If there are multiple consecutive outliers (such as 5 or more consecutive distance library data are abnormal), refer to the echo intensity data of the adjacent distance library with the same azimuth angle and repair it by linear interpolation method. For example, if the data of the i-th to i+n distance library are abnormal, according to the intensity value I of the adjacent distance library with the same azimuth angle, j,k and I j',k (j and j' are adjacent azimuths, k is the range library number), using formula I i:i+n,k =I j,k +(ij)×(I j',k -I j,k) / (j'-j) for interpolation calculation.

[0068] S203: Detection of radial velocity outliers.

[0069] Considering factors such as the earth's rotation, radar system errors and atmospheric turbulence, the speed rationality range is set. For X-band radar, the radial velocity theoretical range is usually [-V max ,V max ], where V max The speed is determined based on parameters such as the radar pulse repetition frequency (PRF). Based on common meteorological conditions in Yunnan, if the speed value exceeds this range and is not supported by special meteorological events such as severe convection, it is considered abnormal.

[0070] Use velocity deblurring techniques (such as the sliding window method) to check for ambiguity in the velocity data. If velocity ambiguity is detected, mark the corresponding data point. Simultaneously, analyze the spatiotemporal continuity of the velocity data. If there are significant speed variations between adjacent distance bins or time series (e.g., a velocity gradient exceeding a set threshold of 5 m / s / km) without a reasonable meteorological explanation, consider it an anomaly.

[0071] S204: Radial velocity outlier repair.

[0072] For velocity ambiguous data, multiple PRF techniques or phase encoding techniques are used for deambiguation. If the data remains abnormal after deambiguation, it is repaired using a weighted average method, referencing velocity data from adjacent azimuth and range libraries. The weight is determined by distance and correlation (closer and more correlated neighboring data have a higher weight).

[0073] For unambiguous outlier velocity values, if they are isolated points, a spatiotemporal smoothing filter method is used. With the outlier as the center, a suitable window is selected in the spatiotemporal dimension (e.g., two time steps before and after, three azimuth angles to the left and right, and three distance bins to the top and bottom). The weighted average of the data within the window (the weight takes into account time and spatial distance) is calculated to replace the outlier.

[0074] S300: Extracting characteristic data based on the repaired echo data, where the characteristic data includes echo intensity, velocity, spectrum width, and polarization characteristics.

[0075] In this embodiment, step S300 is a feature extraction step, which can adopt existing feature extraction methods to mine information from multiple dimensions of time domain, frequency domain, and spatial domain, and ultimately serve subsequent smoke echo identification or physical parameter inversion.

[0076] S400: Constructing a machine learning model, and using the feature data to train the machine learning model to obtain a smoke echo recognition model.

[0077] In this embodiment, a smoke echo recognition model is constructed based on a machine learning algorithm. First, a large amount of known forest fire smoke echo data and other interference echo data (such as precipitation echoes and ground object echoes) is collected. Feature extraction is performed on this data, including parameters such as echo intensity, velocity, spectral width, and polarization characteristics (this can be achieved through steps S100-S300 as described above and will not be elaborated on here).

[0078] The extracted feature data is used to train a machine learning model, selecting appropriate classification algorithms such as support vector machines (SVMs) and random forests. Model parameters are continuously optimized to improve the model's accuracy in identifying smoke echoes. During training, the characteristics of forest fires in Yunnan Province are fully considered, such as the differences in smoke echoes generated by different vegetation types and the impact of terrain on echo propagation, making the model more adaptable and robust.

[0079] In a specific embodiment, step S400 includes the following steps:

[0080] S401: Collect radar echo data from historical forest fires in a specific area, along with corresponding non-fire interference echo data. This includes data from different seasons, vegetation types, and terrain conditions. This data is carefully labeled to clearly identify smoke and dust echoes and interference echoes. Furthermore, various characteristic parameters of the echo data, such as echo intensity, velocity, spectral width, and polarization characteristics, are extracted. These characteristic parameters are then normalized to improve the efficiency and accuracy of model training.

[0081] S402: Based on the characteristics of smoke echoes as a forest fire identification algorithm, a smoke echo identification model is constructed based on the campfire model by introducing environmental wind, seasonal factors, terrain factors, vegetation factors, etc. The parameters of the model are optimized by combining historical cases with artificial intelligence, such as the three-dimensional expansion factor, the three-dimensional environmental wind factor, the vegetation factor, etc. During the training process, the labeled training data is divided into a training set and a validation set. The model is trained using the training set, and the performance of the model is evaluated using the validation set. The model parameters are continuously adjusted until the model achieves a good recognition accuracy on the validation set. At the same time, in order to improve the generalization ability of the model, data enhancement technology is used to perform random transformations (such as rotation, scaling, etc.) on the training data to increase the diversity of the training data.

[0082] S500: Based on the echo recognition model, in response to the characteristic data of the input radar echo data, determine whether the current radar echo data is a smoke echo generated by a forest fire.

[0083] In practical applications, the preprocessed radar echo data is input into the trained smoke echo recognition model, and the model outputs the recognition result to determine whether it is the smoke echo generated by a forest fire.

[0084] Figure 3 This is the identification result of the echo identification model. According to the analysis of existing cases, in the initial stage of the forest fire in Yunnan Province, a scattered cluster echo of 10-20dBZ was observed at an elevation angle of 0.5°, and the horizontal diffusion range of the smoke plume was less than 10km, showing a "point source" feature; in the strong stage, the reflectivity core increased to 30-40dBZ, and the smoke plume extended by more than 30km due to the influence of high-altitude winds, with a "V"-shaped structure; in the stable stage, the reflectivity intensity decreased but the range extended to 40km, and the secondary peak indicated local re-ignition or flying fire spread.

[0085] In actual forest fire monitoring, preprocessed radar echo data is input into a trained smoke echo recognition model. The model extracts features and classifies the input data, outputting a smoke echo identification result. If a smoke echo is identified, further forest fire monitoring and inversion operations are performed (i.e., the subsequent step S600). If it is an interference echo, it is filtered out and subsequent radar echo data monitoring continues.

[0086] S600: When it is determined that the radar echo data is smoke echo generated by a forest fire, the location, scope, and spread speed of the fire are determined based on the spatiotemporal distribution characteristics of the radar echo data.

[0087] Specifically, real-time monitoring of forest fires is achieved based on the identification of smoke echoes. By analyzing the spatiotemporal distribution of smoke echoes, the fire's location, scope, and direction of spread can be determined. Using radar echo data at multiple times, the system can track the evolution of a fire and promptly detect sudden changes in its intensity.

[0088] The fire location calculation formula is:

[0089] P(X,Y)=(P1(X1,Y1)-V1(x1,y1)×Δt+P2(X2,Y2)-V2(x2,y2)×Δt+P3(X3,Y3)-

[0090] V3(x3,y3)×Δt+···+P n (X n ,Y n )-V n (x n ,y n )×Δt) / n

[0091] Where P(X,Y) is the fire location, x and y represent the latitude and longitude respectively; n is the maximum elevation angle at which the X-band radar detects the smoke echo; P n (X n ,Y n ) represents the latitude and longitude position of the echo at the nth elevation angle, starting from the low layer to the high layer; V n (x n ,y n ) is the ambient wind at the nth elevation angle, which can be converted using radial velocity, and Δt is the radar detection time interval.

[0092] The fire area calculation formula is:

[0093] S i =Σ(a1×S P1 +a2×S P2 +a3×S P3 +···+a n ×S Pn ) / n

[0094] Where S i represents the fire area at time i, a n is the pyramid layer coefficient, S pnis the area of ​​the radar echo, an is trained based on historical cases, and the Yunnan Province coefficient matrix is ​​formed based on terrain and season.

[0095] The formula for calculating the fire spread rate is:

[0096] V f =P*G*M*R*W*V r

[0097] Where V f = represents the spread of a forest fire, P represents the terrain factor, G also represents the vegetation factor, M represents the seasonal factor, W represents the ambient wind factor, and Vr is the speed measured by X-band radar. Using historical examples and artificial intelligence, we optimize the P, G, M, R, and W factors.

[0098] In some embodiments, as Figure 4 As shown, combining physical principles and mathematical models, the fire behavior of forest fires is inverted. Based on the intensity and velocity of smoke echoes, key parameters such as the fire's combustion intensity and heat release rate are estimated. By developing a fire spread model that considers the impact of factors such as terrain, wind speed, direction, and vegetation type on fire spread, future fire trends are predicted, providing a scientific basis for firefighting decisions.

[0099] Fire spread rate prediction formula:

[0100] V f (t) = V f +b t *G t *P t *W t

[0101] Where V f (t) represents the predicted speed of forest fire spread at time t, which is closely related to the spread speed at the initial time. t represents the initial velocity of the pattern or model prediction, P t Represents the terrain factor at time t, G t represents the vegetation factor t, at this time the effect of seasonal factors is relatively small (ignored), W t Represents the environmental wind factor at time t.

[0102] Furthermore, a visualization interface can be developed to present monitoring and inversion results to users in an intuitive manner. Through map overlays and dynamic charts, information such as the fire's location, scope, fire growth, and forecast results can be displayed, allowing users to promptly understand the fire situation and formulate appropriate prevention and control measures. Based on real-time monitoring and forecasts, the dynamic spread and impact area of ​​the fire in the next 30 minutes, 1 hour, 2 hours, 3 hours, and 6 hours can be generated in real time. Subsequently, based on the temporal and spatial changes of the fire, intelligent fire extinguishing plans can be generated, which will benefit forest resource protection and personnel safety.

[0103] In a specific embodiment, step S600 may be implemented by the following steps:

[0104] S601: Fire Location and Extent Determination: The location of a forest fire is determined based on the location of identified smoke echoes within the radar scanning area, combined with radar geometry and geographic coordinate conversion, taking into account factors such as ambient wind and expansion coefficients. By analyzing smoke echoes at multiple consecutive moments, the burned area and intensity are calculated, and fire extent information is updated in real time.

[0105] S602: Fire Growth Monitoring: This system compares fire extent and smoke echo intensity and velocity at different times, combining a bonfire model with X-band radar 3D spatial data to analyze fire development. For example, the system calculates the fire's expansion rate and the rate of change in smoke echo intensity using the 3D expansion coefficient, ambient wind, terrain, and seasonal factors to determine whether the fire is in a stable or rapidly spreading phase. Furthermore, based on real-time detection of the fire's initial spread velocity and combined with a fire prediction model, the system takes into account factors such as ambient wind, terrain, vegetation, and wind direction to determine the direction of fire spread and the areas potentially affected.

[0106] S603: Fire Behavior Inversion: Based on combustion theory in physics and principles of fluid mechanics, a forest fire behavior inversion model is established. Identified smoke echo parameters such as intensity and velocity, along with meteorological and geographic data, are input into the model to invert key fire behavior parameters such as combustion intensity and heat release rate. Analysis of fire behavior parameters further understands the combustion characteristics and energy release of fires, providing a scientific basis for developing fire suppression strategies.

[0107] S604: Visualization: Display forest fire monitoring and inversion results through the developed visualization interface. On the Geographic Information System (GIS) map, different colors and symbols are used to identify the location, scope, fire development, and predicted fire spread path of the fire. The monitoring results are saved in real time to the database, providing data support for subsequent forest fire research, analysis, and review, and also providing scientific and reliable decision-making support for subsequent post-disaster reconstruction and assessment. At the same time, the changes in key fire parameters (such as combustion intensity, heat release rate, etc.) over time are displayed in the form of charts, providing users with intuitive and comprehensive forest fire information.

[0108] The present application also provides a forest fire monitoring device, such as Figure 5 As shown, the forest fire monitoring device includes:

[0109] The data acquisition module 501 is configured to acquire radar echo data; wherein the radar echo data is data obtained by scanning the forest area using an X-band radar;

[0110] The data processing module 502 is configured to pre-process the radar echo data to obtain pre-processed data, and perform reflectivity outlier detection and repair and radial velocity outlier detection and repair on the pre-processed data to obtain repaired echo data;

[0111] The feature extraction module 503 is configured to extract feature data based on the repaired echo data, wherein the feature data includes echo intensity, velocity, spectral width and polarization characteristics;

[0112] The model training module 504 is configured to build a machine learning model and train the machine learning model using the feature data to obtain a smoke echo recognition model;

[0113] The echo recognition module 505 is configured to determine whether the current radar echo data is a smoke echo generated by a forest fire based on the echo recognition model and in response to the characteristic data of the input radar echo data;

[0114] The monitoring inversion module 506 is configured to determine the location, scope and fire spread speed of the fire based on the spatiotemporal distribution characteristics of the radar echo data when the radar echo data is determined to be smoke echo generated by a forest fire.

[0115] In some embodiments, the data processing module is further configured to:

[0116] Calculate the mean and standard deviation of the echo intensity of each range library, determine the intensity threshold based on the mean and standard deviation, and determine the data point whose echo intensity exceeds the intensity threshold as a reflectivity outlier. If the intensity value of the data point and the intensity value of the adjacent area exceed a first threshold, then the data point is determined to be a reflectivity outlier;

[0117] For isolated reflectivity outliers, the neighborhood data of the set window size is selected with the outlier as the center, and the median of the neighborhood data is used to replace the outlier to repair the isolated reflectivity outlier. For continuous reflectivity outliers, the intensity value I of the adjacent azimuth and the same distance library is used. j,k and I j',k , using formula I i:i+n,k =I j,k +(ij)×(I j',k -I j,k) / (j'-j) is interpolated to repair the continuous reflectivity anomalies, where j and j' are adjacent azimuths, k is the range library number, and I i:i+n,k The reflectivity values ​​of the k distance libraries from azimuth angles i to i+n.

[0118] In some embodiments, the data processing module is further configured to:

[0119] Set the radial velocity range to [-V max ,V max ], where V max The maximum radial velocity is determined according to the radar pulse repetition frequency. If the radial velocity exceeds [-V max ,V max ] and there is no support from severe convective meteorological events, the radial velocity value is determined to be the first abnormal radial velocity value; if the radial velocity value is ambiguous and the velocity change in the adjacent distance library or time series exceeds the second threshold, the radial velocity value is determined to be the second abnormal radial velocity value;

[0120] For the first abnormal radial velocity value, the spatiotemporal smoothing filtering method is used. With the first abnormal radial velocity value as the center, a window is selected in the spatiotemporal dimension, and the weighted average value of the data in the window is calculated to replace the abnormal value. For the second abnormal radial velocity value, the defuzzification process is performed on it, and the [-V max ,V max ] Determine whether the data after defuzzification is abnormal.

[0121] In some embodiments, when it is determined that the radar echo data is smoke echo generated by a forest fire, the location of the fire is determined according to the spatiotemporal distribution characteristics of the radar echo data using the following formula:

[0122] P(X,Y)=(P1(X1,Y1)-V1(x1,y1)×Δt+P2(X2,Y2)-V2(x2,y2)×Δt+P3(X3,Y3)-

[0123] V3(x3,y3)×Δt+···+P n (X n ,Y n )-V n (x n ,y n )×Δt) / n

[0124] Where P(X,Y) is the fire location, X and Y represent the latitude and longitude respectively; n is the maximum elevation angle at which the X-band radar detects the smoke echo; P n (X n ,Y n ) represents the longitude and latitude position of the echo at the nth elevation angle; V n (x n ,y n ) is the ambient wind at the nth elevation angle, and Δt is the radar detection time interval.

[0125] In some embodiments, when the radar echo data is determined to be smoke echoes generated by a forest fire, the fire occurrence range is determined using the following formula based on the spatiotemporal distribution characteristics of the radar echo data:

[0126] S i =Σ(a1×S P1 +a2×S P2 +a3×S P3 +···+a n ×S Pn ) / n

[0127] Where S i represents the fire area at time i, a n is the pyramid layer coefficient, S pn is the area of ​​the radar echo; n is the maximum elevation angle at which the X-band radar detects the smoke echo.

[0128] In some embodiments, when the radar echo data is determined to be smoke echoes generated by a forest fire, the fire spread rate is determined according to the spatiotemporal distribution characteristics of the radar echo data using the following formula:

[0129] V f =P*G*M*R*W*V r

[0130] Where V frepresents the speed of forest fire spread, P represents the terrain factor, G also represents the vegetation factor, M represents the seasonal factor, W represents the environmental wind factor, and Vr is the speed measured by the X-band radar.

[0131] In some embodiments, after determining the location, scope, and spread rate of the fire, the method further includes: predicting the spread rate of the fire using the following formula:

[0132] V f (t) = V f +b t *G t *P t *W t

[0133] Where V f (t) represents the predicted speed of forest fire spread at time t, b t represents the initial velocity of the pattern or model prediction, P t Represents the terrain factor at time t, G t represents the vegetation factor at time t, W t Represents the environmental wind factor at time t.

[0134] An embodiment of the present application provides an electronic device, which may include a processor and a memory, wherein the processor and the memory can communicate with each other; illustratively, the processor and the memory communicate with each other via a communication bus.

[0135] The processor executes the computer-executable instructions stored in the memory, so that the processor implements the solutions in the above embodiments. The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0136] The communication bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. System buses can be categorized as address buses, data buses, and control buses. Transceivers enable communication between the database access device and other computers (e.g., clients, read-write libraries, and read-only libraries). Memory may include random access memory (RAM) or non-volatile memory.

[0137] The electronic device provided in the embodiment of the present application may be the terminal device of the above embodiment.

[0138] An embodiment of the present application also provides a computer-readable storage medium, which stores computer instructions. When the computer instructions are executed on a computer, the computer executes the technical solution of the forest fire monitoring method of the above embodiment.

[0139] An embodiment of the present application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. At least one processor can read the computer program from the computer-readable storage medium. When at least one processor executes the computer program, it can implement the technical solution of the forest fire monitoring method in the above embodiment.

[0140] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of modules is only a logical function division. In actual implementation, there may be other division methods, such as multiple modules can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or module, which can be electrical, mechanical or other forms.

[0141] Modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical units, that is, they may be located in one place or distributed across multiple network elements. Some or all of these modules may be selected to implement the solution of this embodiment based on actual needs.

[0142] In addition, the functional modules in the various embodiments of the present application may be integrated into a single processing unit, or each module may exist physically separately, or two or more modules may be integrated into a single unit. The above-mentioned modules may be implemented in the form of hardware or hardware plus software functional units.

[0143] The above-mentioned integrated module implemented in the form of a software functional module can be stored in a computer-readable storage medium. The above-mentioned software functional module is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) or a processor to perform some steps of the methods of various embodiments of the present application.

[0144] It should be understood that the processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), or application-specific integrated circuits (ASICs). A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the present invention may be directly implemented by a hardware processor or implemented by a combination of hardware and software modules in the processor.

[0145] The memory may include a high-speed RAM memory, and may also include non-volatile storage NVM, such as at least one disk memory, and may also be a USB flash drive, a mobile hard disk, a read-only memory, a magnetic disk or an optical disk.

[0146] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Buses can be divided into address buses, data buses, and control buses.

[0147] The storage medium may be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The storage medium may be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0148] An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be an integral part of the processor. The processor and the storage medium can be located in an application specific integrated circuit (ASIC). Of course, the processor and the storage medium can also exist as discrete components in an electronic control unit or a main control device.

[0149] Those skilled in the art will appreciate that all or part of the steps in the above-described method embodiments can be implemented using hardware associated with program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0150] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A forest fire monitoring method, characterized in that: The method comprises: Acquiring radar echo data; wherein the radar echo data is data obtained by scanning the forest area using an X-band radar; Preprocessing the radar echo data to obtain preprocessed data, performing reflectivity outlier detection and repair and radial velocity outlier detection and repair on the preprocessed data to obtain repaired echo data; Extracting characteristic data based on the repaired echo data, the characteristic data including echo intensity, velocity, spectral width and polarization characteristics; Constructing a machine learning model, and using the feature data to train the machine learning model to obtain a smoke echo recognition model; Based on the echo recognition model, in response to the characteristic data of the input radar echo data, determining whether the current radar echo data is a smoke echo generated by a forest fire; When it is determined that the radar echo data is smoke echo generated by a forest fire, the location, scope, and spread speed of the fire are determined based on the spatiotemporal distribution characteristics of the radar echo data.

2. The forest fire monitoring method according to claim 1, characterized in that: Methods for detecting and repairing reflectivity outliers on the preprocessed data include: Calculate the mean and standard deviation of the echo intensity of each range library, determine the intensity threshold based on the mean and standard deviation, and determine the data point whose echo intensity exceeds the intensity threshold as a reflectivity outlier. If the intensity value of the data point and the intensity value of the adjacent area exceed a first threshold, then the data point is determined to be a reflectivity outlier; For isolated reflectivity outliers, the neighborhood data of the set window size is selected with the outlier as the center, and the median of the neighborhood data is used to replace the outlier to repair the isolated reflectivity outlier. For continuous reflectivity outliers, the intensity value I of the adjacent azimuth and the same distance library is used. j,k and I j',k , using formula I i:i+n,k =I j,k +(ij)×(I j',k -I j,k) / (j'-j) is interpolated to repair the continuous reflectivity anomalies, where j and j' are adjacent azimuths, k is the range library number, and I i:i+n,k The reflectivity values ​​of the k distance libraries from azimuth angles i to i+n.

3. The forest fire monitoring method according to claim 1, characterized in that: The method of detecting and repairing radial velocity outliers on the pre-processed data includes: Set the radial velocity range to [-V max ,V max ], where V max The maximum radial velocity is determined according to the radar pulse repetition frequency. If the radial velocity exceeds [-V max ,V max ] and there is no support from severe convective meteorological events, the radial velocity value is determined to be the first abnormal radial velocity value; if the radial velocity value is ambiguous and the velocity change in the adjacent distance library or time series exceeds the second threshold, the radial velocity value is determined to be the second abnormal radial velocity value; For the first abnormal radial velocity value, the spatiotemporal smoothing filtering method is used. With the first abnormal radial velocity value as the center, a window is selected in the spatiotemporal dimension, and the weighted average value of the data in the window is calculated to replace the abnormal value. For the second abnormal radial velocity value, the defuzzification process is performed on it, and the [-V max ,V max ] Determine whether the data after defuzzification is abnormal.

4. The forest fire monitoring method according to claim 1, characterized in that: When it is determined that the radar echo data is smoke echo generated by a forest fire, the location of the fire is determined according to the spatiotemporal distribution characteristics of the radar echo data using the following formula: P(X,Y)=(P1(X1,Y1)-V1(x1,y1)×Δt+P2(X2,Y2)-V2(x2,y2)×Δt+P3(X3,Y3)- V3(x3,y3)×Δt+···+P n (X n ,Y n )-V n (x n ,y n )×Δt) / n Where P(X,Y) is the fire location, X and Y represent the latitude and longitude respectively; n is the maximum elevation angle at which the X-band radar detects the smoke echo; P n (X n ,Y n ) represents the longitude and latitude position of the echo at the nth elevation angle; V n (x n ,y n ) is the ambient wind at the nth elevation angle, and Δt is the radar detection time interval.

5. The forest fire monitoring method according to claim 1, characterized in that: When it is determined that the radar echo data is smoke echo generated by a forest fire, the fire occurrence range is determined by the following formula based on the spatiotemporal distribution characteristics of the radar echo data: S i =Σ(a1×S P1 +a2×S P2 +a3×S P3 +···+a n ×S Pn ) / n Where S i represents the fire area at time i, a n is the pyramid layer coefficient, S pn is the area of ​​the radar echo; n is the maximum elevation angle at which the X-band radar detects the smoke echo.

6. The forest fire monitoring method according to claim 1, characterized in that: When it is determined that the radar echo data is smoke echo generated by a forest fire, the fire spread speed is determined by the following formula based on the spatiotemporal distribution characteristics of the radar echo data: V f =P*G*M*R*W*V r Where V f represents the speed of forest fire spread, P represents the terrain factor, G also represents the vegetation factor, M represents the seasonal factor, W represents the environmental wind factor, and Vr is the speed measured by the X-band radar.

7. The forest fire monitoring method according to claim 1, characterized in that: After determining the location, scope, and spread rate of the fire, the method further includes: predicting the spread rate of the fire using the following formula: V f (t)=V f +b t *G t *P t *W t Where V f (t) represents the predicted speed of forest fire spread at time t, b t represents the initial velocity of the pattern or model prediction, P t Represents the terrain factor at time t, G t represents the vegetation factor at time t, W t Represents the environmental wind factor at time t.

8. A forest fire monitoring device, characterized in that: The device comprises: a data acquisition module configured to acquire radar echo data; wherein the radar echo data is data obtained by scanning the forest area using an X-band radar; a data processing module configured to preprocess the radar echo data to obtain preprocessed data, and perform reflectivity outlier detection and repair and radial velocity outlier detection and repair on the preprocessed data to obtain repaired echo data; a feature extraction module configured to extract feature data based on the repaired echo data, wherein the feature data includes echo intensity, velocity, spectral width and polarization characteristics; a model training module configured to construct a machine learning model and train the machine learning model using the feature data to obtain a smoke echo recognition model; an echo recognition module configured to determine whether the current radar echo data is a smoke echo generated by a forest fire based on the echo recognition model and in response to characteristic data of the input radar echo data; The monitoring inversion module is configured to determine the location, scope and fire spread speed of the fire based on the spatiotemporal distribution characteristics of the radar echo data when it is determined that the radar echo data is smoke echo generated by a forest fire.

9. An electronic device, characterized in that: include: a processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the forest fire monitoring method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the forest fire monitoring method according to any one of claims 1 to 7.

Images