UAV wildfire wide-area monitoring method and system based on transmission space geographic mapping

Abstract

The present invention discloses a method and system for monitoring the width of wildfires using an unmanned aerial vehicle (UAV) based on geographic mapping of a power transmission space; it relates to the technical field of wildfire monitoring; the scheme improves the method on the technical basis of two types of monitoring methods, namely, traditional satellite monitoring and video monitoring; when a wildfire alarm signal in the power transmission space is collected, the flight path and mission identity of each UAV in the UAV cluster are dynamically planned based on an artificial bee colony algorithm to dynamically track the edge of the wildfire; edge detection is performed based on image data collected by the UAV cluster to obtain a wildfire edge image; finally, the wildfire edge image and the geographic location of the UAV are combined to perform a fire range analysis to obtain the wildfire width in the power transmission space, thereby realizing the detection and tracking of the wildfire edge, dynamically analyzing the wildfire width in the power transmission space, and improving the wildfire monitoring range, efficiency and accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of mountain fire monitoring, and in particular to a method and system for wide-area mountain fire monitoring using an unmanned aerial vehicle based on power transmission space geographic mapping. Background Art

[0002] In recent years, wildfires have occurred frequently around the world, posing a serious threat to power transmission lines across mountainous areas. The high-temperature hot air currents and large amounts of smoke generated by wildfires will contaminate the surface of the transmission line insulators, reducing their insulation performance. When wildfires spread to the vicinity of transmission lines, the flames and high temperatures may cause the collapse of transmission towers and the breakage of conductors, leading to short circuits or grounding faults, and ultimately causing the transmission lines to trip, causing huge losses. The key to wildfire prevention and control is to discover and extinguish wildfires in the early stages of the fire, prevent their spread, and avoid the tripping of transmission lines caused by wildfires. Wildfire monitoring, as a key link in wildfire prevention and control, has a positive significance for wildfire prevention and control in transmission line corridors.

[0003] The methods for monitoring wildfires in transmission line corridors mainly include satellite monitoring and video monitoring. The satellite monitoring method uses satellites equipped with heat source analyzers to monitor forest fires. The existing technology uses satellites equipped with high-pixel infrared imagers (advanced himawari imagers, AHI) to monitor wildfires in transmission line corridors, and combines geographic information data with satellite remote sensing data for correlation analysis to improve the accuracy of wildfire monitoring in transmission line corridors; there are also satellite remote sensing wildfire monitoring methods based on image enhancement, which improve the accuracy of satellite identification of wildfires by resisting cloud and fog interference; and the wildfire data based on the Himawari-8 geostationary satellite is used to analyze the ignition point to achieve accurate identification of wildfires. It can be seen that there are various methods for using satellites to monitor wildfires in transmission line corridors, but satellite monitoring is limited by resolution and sensitivity, and it is difficult to detect early signs of wildfires.

[0004] Video monitoring uses cameras installed on transmission towers to monitor wildfires in transmission line corridors. The images of the transmission line corridors are captured by fixed cameras on the transmission towers, and wildfires are identified using a wildfire image feature extraction method based on grid division; or the cloud and fog interference information of the transmission line corridor images is filtered out based on the Gaussian filter algorithm, which improves the accuracy of wildfires in transmission line corridors; however, the field of view of transmission line video monitoring is limited, making it difficult to survey the scope of wildfires.

[0005] When it comes to wildfires in transmission line corridors, traditional satellite monitoring and video monitoring methods have the problem of difficulty in detecting the early signs of wildfires and the specific scope of the wildfires. Summary of the invention

[0006] The technical problem to be solved by the present invention is: for wildfires in transmission line corridors, the two traditional monitoring methods of satellite monitoring and video monitoring have the problem of difficulty in surveying the early signs of wildfires and the specific scope of wildfires; the purpose of the present invention is to provide a method and system for wide-width wildfire monitoring by unmanned aerial vehicles based on geographic mapping of transmission space, and to improve the method on the technical basis of the two traditional monitoring methods of satellite monitoring and video monitoring; according to the wildfire alarm signal, a cluster of unmanned aerial vehicles conducts detection and tracking of the edge of the wildfire, and combines the geographical location of the drone cluster and the tracked wildfire edge image information to analyze the width of the wildfire in the transmission space, thereby improving the range and accuracy of wildfire monitoring.

[0007] The present invention is achieved through the following technical solutions:

[0008] This solution provides a method for UAV wildfire wide-area monitoring based on transmission space geographic mapping, including:

[0009] Real-time collection of wildfire warning signals in the power transmission space;

[0010] When a wildfire warning signal is collected in the power transmission space, the flight path and mission identity of each drone in the drone cluster are dynamically planned based on the artificial bee colony algorithm to dynamically track the edge of the wildfire; the mission identities include: reconnaissance identity, employment identity and follower identity;

[0011] Obtain image data collected by the drone cluster and pre-process the image data;

[0012] Perform edge detection based on the preprocessed image data to obtain a wildfire edge image;

[0013] The fire range analysis was performed by combining the wildfire edge image and the UAV’s geographic location to obtain the wildfire width in the transmission space.

[0014] Working principle of this scheme: The UAV wide-width wildfire monitoring method based on transmission space geographic mapping provided by this scheme makes methodological improvements on the technical basis of the two types of monitoring methods: traditional satellite monitoring and video monitoring. According to the wildfire alarm signal, the UAV cluster conducts detection and tracking of the wildfire edge, and combines the geographical location of the UAV cluster and the tracked wildfire edge image information to analyze the wildfire width in the transmission space, thereby improving the wildfire monitoring range and accuracy.

[0015] A further optimization scheme is that the artificial bee colony algorithm is used to dynamically plan the flight path and mission identity of each drone in the drone cluster to dynamically track the edge of the wildfire in the power transmission space; including methods:

[0016] S21, command the reconnaissance identity drone to conduct a flight patrol to search for the global wildfire edge;

[0017] S22, configure the reconnaissance drone to transform into a hired drone, and cooperate with the tracking drone to dynamically track the edge of the global wildfire:

[0018] Divide the global wildfire edge into multiple local wildfire edges;

[0019] Each hired drone corresponds to a local wildfire edge. The hired drone stores the distance from the current position to the nest, the flight direction from the current position to the nest, and the corresponding local wildfire edge information;

[0020] Each hired identity drone searches the area of ​​the local wildfire edge in iterations, and hires a follower identity drone to collect images of the local wildfire edge in a roulette manner;

[0021] If the local wildfire edge image collected by the follow-up identity drone has not been updated for N1 consecutive times, the current local wildfire edge will be abandoned, and the hired identity drone of the current local wildfire edge will be configured to be transformed into a reconnaissance identity drone for flight patrol to search for new local wildfire edges.

[0022] Further optimization plans also include:

[0023] The local wildfire edge images collected by all the following UAVs are combined to obtain the global wildfire edge image;

[0024] If the global wildfire edge image has not been updated for N2 consecutive times, the current global wildfire edge image is abandoned, the hired identity drone is configured to be transformed into a reconnaissance identity drone, and the process returns to step S21.

[0025] A further optimization scheme is to hire a feasible solution randomly generated by the identity drone j y j for:

[0026] ;

[0027] Where: represents the minimum feasible solution randomly generated by the hired identity drone j; represents the maximum feasible solution randomly generated by hired identity drone j; rand (*) represents a random function.

[0028] A further optimization scheme is that the probability of hiring identity drone j being selected by the following identity drone is P j for:

[0029] ;

[0030] Where: fit ( y j) represents the fire edge richness layer corresponding to the fitness value of the feasible solution of hired identity drone j. The larger the richness layer, the greater the possibility that hired identity drone j follows the selection of identity drone; n Indicates the total number of hired drones.

[0031] A further optimization scheme is that the method for preprocessing the image data includes:

[0032] Get the wildfire target pixels and environment pixels of the current image;

[0033] Calculate the brightness and darkness relationship between the wildfire target pixels and the environment pixels, and adjust the contrast of the current image according to the brightness and darkness relationship to correct the color and airspace of the current image;

[0034] Dynamically expand the corrected current image.

[0035] A further optimization scheme is that edge detection is performed based on the preprocessed image data to obtain a wildfire edge image; including the following method:

[0036] Convert the preprocessed image data into a grayscale image;

[0037] Based on the Sobel operator, the intensity and direction of each pixel in the grayscale image are calculated, and the gradient value of two adjacent pixels is calculated to perform non-maximum suppression;

[0038] Perform edge double threshold processing on the grayscale image to divide strong edge pixels, weak edge pixels and non-edge pixels;

[0039] Connect the weak edge pixels and the surrounding strong edge pixels to obtain the wildfire edge image.

[0040] The further optimization scheme is to combine the wildfire edge image and the geographic location of the drone to analyze the fire range and obtain the wildfire width in the transmission space; including methods:

[0041] Construct a fire edge polygon by combining the wildfire edge image and the geographic location of the drone: obtain the geographic coordinates of each drone, and determine the wildfire edge coordinate points by combining the wildfire edge image; connect the wildfire edge coordinate points into line segments in sequence, and close the line segments into a fire edge polygon;

[0042] The area of ​​the fire edge polygon is calculated based on the polygon fitting algorithm.

[0043] This solution also provides a UAV wildfire wide-area monitoring system based on power transmission space geographic mapping, which is used to implement the above-mentioned UAV wildfire wide-area monitoring method based on power transmission space geographic mapping; the system includes:

[0044] The dynamic tracking module is used to dynamically plan the flight path and mission identity of each drone in the drone cluster based on the artificial bee colony algorithm when a wildfire warning signal is collected in the power transmission space, so as to dynamically track the edge of the wildfire; the mission identity includes: reconnaissance identity, employment identity and follower identity;

[0045] A preprocessing module, used to obtain image data collected by the drone cluster and preprocess the image data;

[0046] An edge detection module is used to perform edge detection based on the preprocessed image data to obtain a wildfire edge image;

[0047] The range analysis module is used to analyze the fire range by combining the wildfire edge image and the geographical location of each drone to obtain the wildfire width in the transmission space.

[0048] The preprocessing module comprises:

[0049] A pixel acquisition unit, used to acquire wildfire target pixels and environmental pixels of the current image;

[0050] A correction unit, used to calculate the light-dark relationship between the wildfire target pixel and the environment pixel, and adjust the contrast of the current image according to the light-dark relationship to correct the color and airspace of the current image;

[0051] The expansion unit is used to dynamically expand the corrected current image.

[0052] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0053] 1. The invention provides a method and system for monitoring wide-width wildfires using drones based on geographic mapping of power transmission space. The invention improves the method based on the technical foundation of two types of monitoring methods: traditional satellite monitoring and video monitoring. According to the wildfire warning signal, the drone cluster detects and tracks the edge of the wildfire, and analyzes the width of the wildfire in the power transmission space by combining the geographical location of the drone cluster and the image information of the tracked wildfire edge, thereby improving the range and accuracy of wildfire monitoring.

[0054] 2. The present invention provides a method and system for monitoring wide-area wildfires using drones based on power transmission space geographic mapping; the invention dynamically plans the flight path and mission identity of each drone in the drone cluster based on an artificial bee colony algorithm to dynamically track the edge of the wildfire; the dynamic tracking of the edge of the wildfire is achieved, and in combination with the artificial bee colony algorithm, there is no need to understand other special information about wildfires in the transmission line corridor, only a comparison of the advantages and disadvantages of wildfires in the transmission line corridor is required, and the local optimization of each individual bee is performed to achieve global optimization of the group. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other relevant drawings can be obtained based on these drawings without creative work. In the drawings:

[0056] Figure 1 It is a flowchart of the method for wide-area monitoring of wildfires using UAVs based on spatial geographic mapping of power transmission;

[0057] Figure 2 This is a schematic diagram of the structure of the UAV wildfire wide-area monitoring system based on transmission space geographic mapping. DETAILED DESCRIPTION

[0058] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments and drawings. The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention.

[0059] For wildfires in transmission line corridors, the traditional satellite monitoring and video monitoring methods have the problem of difficulty in detecting the initial signs of wildfires and the specific scope of wildfires; in view of this, this solution provides the following embodiments to solve this technical problem:

[0060] Embodiment 1: This embodiment provides a method for monitoring wide-area wildfires using a drone based on geographic mapping of power transmission space, such as Figure 1 As shown, including:

[0061] Step 1: Real-time collection of wildfire warning signals in the transmission space; wildfires in the transmission line corridor are sudden and random, and infrared cameras can be installed on the transmission tower to detect the fire heat source in real time and activate the wildfire warning; when the wildfire warning signal is collected, the transmission line operation and maintenance personnel drive the vehicle to quickly transport the drone's nest to the vicinity of the fire point for follow-up inspections. Unmanned aerial vehicles (UAVs) have shown great potential in the field of wildfire monitoring due to their flexibility, rapid response, and low cost, and have become an important means to improve the accuracy of wildfire monitoring range.

[0062] Step 2: When a wildfire warning signal is collected in the transmission space, the flight path and mission identity of each drone in the drone cluster are dynamically planned based on the artificial bee colony algorithm to dynamically track the edge of the wildfire; the mission identities include: reconnaissance identity, employment identity and follower identity;

[0063] In step 2, the method of dynamically planning the flight path and mission identity of each drone in the drone cluster based on the artificial bee colony algorithm to dynamically track the edge of the wildfire in the power transmission space includes:

[0064] S21, command the reconnaissance identity drone to conduct a flight patrol to search for the global wildfire edge;

[0065] S22, configure the reconnaissance drone to transform into a hired drone, and cooperate with the tracking drone to dynamically track the edge of the global wildfire:

[0066] Divide the global wildfire edge into multiple local wildfire edges;

[0067] Each hired drone corresponds to a local wildfire edge. The hired drone stores the distance from the current position to the nest, the flight direction from the current position to the nest, and the corresponding local wildfire edge information;

[0068] Each hired identity drone searches the area of ​​the local wildfire edge in the iteration, and hires the follower identity drone to collect local wildfire edge images in a roulette manner; specifically, in the roulette method, the probability of each individual being selected is proportional to the size of its fitness function Q value, so as to prevent individuals with smaller fitness values ​​from being directly eliminated.

[0069] If the local fire edge image collected by the follow-up identity drone has not been updated for N1 consecutive times, the current local fire edge is abandoned, and the hired identity drone of the current local fire edge is configured to be transformed into a reconnaissance identity drone for flight inspection to search for a new local fire edge; the update of the new local fire edge b j for:

[0070] ;

[0071] Where: ϕ j For the j A random number between [-1,1] is selected for each hired identity drone; To randomly generate j A feasible solution for hiring identity drones.

[0072] It also includes S23: obtaining a combination of local wildfire edge images collected by all following identity drones to obtain a global wildfire edge image; in order to prevent the artificial bee colony algorithm of the hired identity drone from having a local optimal solution, if the global wildfire edge image has not been updated for N2 consecutive times, the current global wildfire edge image is abandoned, the hired identity drone is configured to be transformed into a reconnaissance identity drone, and the process returns to step S21.

[0073] The feasible solution of hiring identity drone j randomly generated yj for:

[0074] ;

[0075] Where: represents the minimum feasible solution randomly generated by hiring identity drone j; represents the maximum feasible solution randomly generated by hired identity drone j; rand (*) represents a random function.

[0076] The probability that the hired identity drone j is selected by the follower identity drone P j for:

[0077] ;

[0078] Where: fit ( y j ) represents the fire edge richness layer corresponding to the fitness value of the feasible solution of hired identity drone j. The larger the richness layer, the greater the possibility that hired identity drone j follows the choice of identity drone; n Indicates the total number of hired drones.

[0079] The artificial bee colony algorithm is a cluster optimization algorithm that imitates the honey-collecting behavior of bees. The algorithm does not need to understand the special information of wildfires in the transmission line corridor. It only needs to compare the advantages and disadvantages of wildfires in the transmission line corridor and achieve global optimization of the group through local optimization of each individual bee. The artificial bee colony algorithm has the ability of parallel processing, which is suitable for handling large-scale cluster optimization problems and has strong global search capabilities. The artificial bee colony algorithm consists of three types of elements: food sources, employed bees, and unemployed bees, and includes two behaviors: food source recruitment and bees abandoning a single food source. In the wildfire flight inspection of the transmission line corridor, the food source is wildfire, hired bees are hired drones and unhired bees are unhired drones; the value of the wildfire point in the artificial bee colony algorithm is composed of the distance from the drone to the wildfire point, the flight difficulty level, and the rate of return of discovering the wildfire edge; hired drones are also called leading drones, which correspond one-to-one to the collected local wildfire edges. Hired drones store the distance from the current position to the machine nest, the flight direction from the current position to the machine nest, and the corresponding local wildfire edge information, and share this information with other drones; unhired drones are divided into reconnaissance drones and follower drones. Reconnaissance drones search for wildfire edges in the transmission line corridor near the machine nest; follower drones find the wildfire edge points through the information shared by the reconnaissance drones, hired drones show the rate of return of the wildfire edge points through the duration of the swing dance, and the dance area of ​​the hired drones is the area for exchanging wildfire edge monitoring information. The follower drones can observe a large number of rates of return to determine which wildfire edge point to go to for shooting images.

[0080] Step 3: Obtain image data collected by the drone cluster and pre-process the image data; Wildfires in transmission line corridors change quickly, and the forms of wildfires in different locations and types of vegetation vary greatly. The video camera devices carried by drones are easily affected by fog in forest areas and smoke emitted by farmers' chimneys, which can easily cause false alarms and edge recognition errors for wildfires in transmission line corridors. Therefore, before conducting wildfire edge monitoring, the transmission line images taken by drones are filtered to eliminate the influence of fog and smoke emitted by chimneys.

[0081] The main difference between wildfires in transmission line corridors and clouds and smoke emitted by farmers' chimneys is the contrast. Therefore, by filtering the transmission line corridor image, an accurate image of wildfires in transmission line corridors can be obtained.

[0082] The method for preprocessing image data comprises:

[0083] S31, obtaining wildfire target pixels and environment pixels of the current image;

[0084] S32, calculating the light-dark relationship between the wildfire target pixel and the environment pixel, and adjusting the contrast of the current image according to the light-dark relationship to correct the color and airspace of the current image;

[0085] S33, dynamically expand the corrected current image.

[0086] The specific preprocessing process can be implemented based on automatic color balancing technology. Automatic color balancing technology is an image processing technology that enhances the local contrast of images. It processes the image data locally and processes the contrast of the transmission line corridor by calculating the brightness and darkness and relationship between the wildfire target pixels and the surrounding pixels. It also forms a cumulative contrast balance of the human eye retina, retains the details of the wildfire image in the transmission line corridor, and reduces the misjudgment rate of clouds and smoke emitted from farmers' chimneys.

[0087] Automatic color balancing technology includes color adjustment, spatial adjustment and dynamic image adjustment of the transmission line corridor wildfire image, completes image correction, and then performs dynamic image expansion. The final output data after preprocessing x ACE for:

[0088] ;

[0089] Where: w It represents the adjustment weight of the current image and , and the farther away from the wildfire shape, the smaller its value; g Represents the contrast adjustment function of the current image; z Represents the current image of the input; z 0 represents the minimum contrast value of the current input image;I The current image range.

[0090] Step 4: Perform edge detection based on the preprocessed image data to obtain a wildfire edge image; the edge detection based on the preprocessed image data to obtain the wildfire edge image can be implemented based on the Canny algorithm, which has good anti-noise ability and can accurately detect the edges in the wildfire image of the transmission line corridor; this step 4 specifically includes the following method:

[0091] S41, converting the preprocessed image data into a grayscale image;

[0092] S42, based on the Sobel operator, calculate the intensity and direction of each pixel in the grayscale image, and calculate the gradient value of two adjacent pixels to perform non-maximum suppression; specifically, for each adjacent pixel, check whether its gradient value is the maximum value among the adjacent pixels in its gradient direction. If not, suppress the gradient value of the pixel (set it to 0).

[0093] S43, performing edge double threshold processing on the grayscale image to divide strong edge pixels, weak edge pixels and non-edge pixels; the edge double threshold processing process is:

[0094] ;

[0095] Where: x strong Indicates the strong edge of the grayscale image in the Canny algorithm; x weak Indicates the weak edge of the grayscale image in the Canny algorithm; x non It is the non-edge of grayscale image in Canny algorithm; x my Represents the output data calculated by the Canny algorithm, x mh Represents the high threshold of the gradient strength of the grayscale image; x ml Represents the low threshold of the gradient strength of the grayscale image.

[0096] S44, connecting the weak edge pixel points and the surrounding strong edge pixel points to obtain a wildfire edge image.

[0097] Step 5: Combine the wildfire edge image and the drone’s geographic location to analyze the fire range and obtain the wildfire width in the transmission space. Step 5 specifically includes the following methods:

[0098] S51, construct a fire edge polygon by combining the wildfire edge image and the geographic location of the drone: obtain the geographic coordinates of each drone, and determine the wildfire edge coordinate points by combining the wildfire edge image; connect each wildfire edge coordinate point into a line segment in sequence, and close the line segment into a fire edge polygon; during the process, it is necessary to determine whether the fire edge points have been collected completely. If not, the drone will perform another flight inspection until the fire edge points are collected completely.

[0099] S52, calculating the area of ​​the fire edge polygon based on a polygon fitting algorithm.

[0100] This scheme calculates the area of ​​the fire edge polygon based on the Douglas-Peucker algorithm, which approximates the curve as a series of points and reduces the number of points, thereby simplifying the complex multi-curve and improving the detection efficiency of the geographical range. The specific process is:

[0101] Assume that there are m wildfire edge coordinate points, the wildfire edge coordinate point set A DPA for:

[0102] ;

[0103] Where: a m Indicates m The horizontal coordinates of the fire edge coordinate points; b m Indicates m The vertical coordinate of the fire edge coordinate point.

[0104] The formula for fitting the straight line of the wildfire edge coordinates is: b m = ca m + d ;

[0105] Where: c It represents the intercept of the horizontal axis of the wildfire edge coordinate; d Represents the intercept of the vertical axis of the wildfire edge coordinates.

[0106] The best fitting straight line is solved based on the minimum sum of squared errors of the edge coordinates of the wildfires. E SS for:

[0107] ;

[0108] Where: b k Indicates k The vertical coordinate of the fire edge; a k Indicatesk The horizontal coordinate of the edge of the wildfire.

[0109] The derivative is c and d The values ​​are:

[0110] ;

[0111] ;

[0112] The transmission space in this plan includes the space containing transmission lines, including transmission line corridors, etc.

[0113] Embodiment 2: This embodiment provides a UAV wildfire wide-area monitoring system based on power transmission space geographic mapping, which is used to implement the UAV wildfire wide-area monitoring method based on power transmission space geographic mapping described in Embodiment 1; Figure 2 As shown, the system comprises:

[0114] The dynamic tracking module is used to dynamically plan the flight path and mission identity of each drone in the drone cluster based on the artificial bee colony algorithm when a wildfire warning signal is collected in the power transmission space, so as to dynamically track the edge of the wildfire; the mission identity includes: reconnaissance identity, employment identity and follower identity;

[0115] A preprocessing module, used to obtain image data collected by the drone cluster and preprocess the image data;

[0116] An edge detection module is used to perform edge detection based on the preprocessed image data to obtain a wildfire edge image;

[0117] The range analysis module is used to analyze the fire range by combining the wildfire edge image and the geographical location of each drone to obtain the wildfire width in the transmission space.

[0118] Embodiment 3: This embodiment provides a computer-readable medium having a computer program stored thereon, and the computer program is executed by a processor to implement a method for monitoring wide-area wildfires using a drone based on power transmission space geographic mapping as described in Embodiment 1; specifically, the following steps are performed:

[0119] Step 1: Real-time collection of wildfire warning signals in the power transmission space;

[0120] Step 2: When a wildfire warning signal is collected in the transmission space, the flight path and mission identity of each drone in the drone cluster are dynamically planned based on the artificial bee colony algorithm to dynamically track the edge of the wildfire; the mission identities include: reconnaissance identity, employment identity and follower identity;

[0121] Step 3: Obtain image data collected by the drone cluster and pre-process the image data;

[0122] Step 4: Perform edge detection based on the preprocessed image data to obtain a wildfire edge image;

[0123] Step 5: Combine the wildfire edge image and the UAV’s geographic location to analyze the fire range and obtain the wildfire width in the transmission space.

[0124] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for monitoring wide-area wildfires using drones based on power transmission space geographic mapping, characterized in that: include: When a wildfire warning signal is collected in the transmission space, the flight path and mission identity of each drone in the drone cluster are dynamically planned based on the artificial bee colony algorithm to dynamically track the edge of the wildfire; The mission identities include: reconnaissance identity, employment identity and follower identity; the artificial bee colony algorithm is used to dynamically plan the flight path and mission identity of each drone in the drone cluster to dynamically track the edge of the wildfire in the power transmission space; including methods: S21, command the reconnaissance identity drone to conduct a flight patrol to search for the global wildfire edge; S22, configure the reconnaissance drone to transform into a hired drone, and cooperate with the tracking drone to dynamically track the edge of the global wildfire: Divide the global wildfire edge into multiple local wildfire edges; Each hired drone corresponds to a local wildfire edge. The hired drone stores the distance from the current position to the nest, the flight direction from the current position to the nest, and the corresponding local wildfire edge information; Each hired identity drone searches the area of ​​the local wildfire edge in iterations, and hires a follower identity drone to collect images of the local wildfire edge in a roulette manner; If the local wildfire edge image collected by the follow-up identity drone has not been updated for N1 consecutive times, the current local wildfire edge is abandoned, and the hired identity drone of the current local wildfire edge is configured to be transformed into a reconnaissance identity drone for flight inspection to search for a new local wildfire edge; Obtain image data collected by the drone cluster and pre-process the image data; Perform edge detection based on the preprocessed image data to obtain a wildfire edge image; The fire range analysis was performed by combining the wildfire edge image and the UAV’s geographic location to obtain the wildfire width in the transmission space.

2. The method for monitoring wide-area wildfires using unmanned aerial vehicles based on power transmission space geographic mapping according to claim 1, characterized in that: Also includes: The local wildfire edge images collected by all the following UAVs are combined to obtain the global wildfire edge image; If the global wildfire edge image has not been updated for N2 consecutive times, the current global wildfire edge image is abandoned, the hired identity drone is configured to be transformed into a reconnaissance identity drone, and the process returns to step S21.

3. The method for monitoring wide-area wildfires using unmanned aerial vehicles based on power transmission space geographic mapping according to claim 1, characterized in that: The feasible solution of hiring identity drone j randomly generated y j for: ； Where: represents the minimum feasible solution randomly generated by hiring identity drone j; represents the maximum feasible solution randomly generated by hired identity drone j; rand (*) represents a random function.

4. The method for monitoring wide-area wildfires using unmanned aerial vehicles based on power transmission space geographic mapping according to claim 1, characterized in that: The probability that the hired identity drone j is selected by the follower identity drone P j for: ； Where: fit ( y j ) represents the fire edge richness layer corresponding to the fitness value of the feasible solution of the hired identity drone j; n Indicates the total number of hired drones.

5. The method for monitoring wide-area wildfires using unmanned aerial vehicles based on power transmission space geographic mapping according to claim 1, characterized in that: The method for preprocessing image data comprises: Get the wildfire target pixels and environment pixels of the current image; Calculate the brightness and darkness relationship between the wildfire target pixels and the environment pixels, and adjust the contrast of the current image according to the brightness and darkness relationship to correct the color and airspace of the current image; Dynamically expand the corrected current image.

6. The method for monitoring wide-area wildfires using unmanned aerial vehicles based on power transmission space geographic mapping according to claim 1, characterized in that: The method performs edge detection based on the preprocessed image data to obtain a wildfire edge image; comprising: method: Convert the preprocessed image data into a grayscale image; Based on the Sobel operator, the intensity and direction of each pixel in the grayscale image are calculated, and the gradient value of two adjacent pixels is calculated to perform non-maximum suppression; Perform edge double threshold processing on the grayscale image to divide strong edge pixels, weak edge pixels and non-edge pixels; Connect the weak edge pixels and the surrounding strong edge pixels to obtain the wildfire edge image.

7. The method for monitoring wide-area wildfires using unmanned aerial vehicles based on power transmission space geographic mapping according to claim 1 or 3, characterized in that: The fire range is analyzed by combining the wildfire edge image and the UAV’s geographic location to obtain the wildfire width in the transmission space; Included methods: Construct the fire edge polygon by combining the wildfire edge image and the geographic location of the drone: obtain the geographic coordinates of each drone, and determine the wildfire edge coordinate points by combining the wildfire edge image; Connect the edge coordinate points of each wildfire into line segments in sequence, and close the line segments into a fire edge polygon; The area of ​​the fire edge polygon is calculated based on the polygon fitting algorithm.

8. The UAV wildfire wide-area monitoring system based on power transmission space geographic mapping is characterized by: Used to implement the UAV wildfire wide-area monitoring method based on power transmission space geographic mapping as described in any one of claims 1 to 7; the system includes: The dynamic tracking module is used to dynamically plan the flight path and mission identity of each drone in the drone cluster based on the artificial bee colony algorithm when a wildfire warning signal is collected in the power transmission space, so as to dynamically track the edge of the wildfire; the mission identity includes: reconnaissance identity, employment identity and follower identity; A preprocessing module, used to obtain image data collected by the drone cluster and preprocess the image data; An edge detection module is used to perform edge detection based on the preprocessed image data to obtain a wildfire edge image; The range analysis module is used to analyze the fire range by combining the wildfire edge image and the geographical location of each drone to obtain the wildfire width in the transmission space.

9. The UAV wildfire wide-area monitoring system based on power transmission space geographic mapping according to claim 8 is characterized in that: The preprocessing module comprises: A pixel acquisition unit, used to acquire wildfire target pixels and environmental pixels of the current image; A correction unit, used to calculate the light-dark relationship between the wildfire target pixel and the environment pixel, and adjust the contrast of the current image according to the light-dark relationship to correct the color and airspace of the current image; The expansion unit is used to dynamically expand the corrected current image.

Images