Forest fire extinguishing robot path planning method and system based on air-ground cooperation

Abstract

The application discloses a forest fire extinguishing robot path planning method and system based on air-ground cooperation, wherein the method uses a UAV to carry a laser radar and infrared and visible light cameras to perform air scanning on a disaster area, and acquires spatial information and fire intensity of a fire point in real time; the UAV is connected in communication with a wind power fire extinguishing robot, shares a three-dimensional point cloud model, and performs environment calibration and positioning; an improved A* algorithm is applied to path planning, and a path planning mathematical model with a minimum disaster degree as a target is constructed; a global satellite navigation system and an inertial navigation system are fused to dynamically update a planned operation path of the wind power fire extinguishing robot; after reaching a target fire point, the wind power fire extinguishing robot identifies a fire source, calculates a safe fire extinguishing distance by using a flame detection model result, and dynamically adjusts a wind cylinder angle according to a flame state to perform fire extinguishing operation. Through the cooperative work between the UAV and the fire extinguishing robot, the efficiency and intelligent degree of fire extinguishing operation are improved.

Description

Technical Field

[0001] This invention belongs to the field of forest fire fighting robot technology, and relates to a path planning method for forest fire fighting robots, specifically a path planning method and system for forest fire fighting robots based on ground-air collaboration. Background Technology

[0002] Forest fires are a major disaster that seriously threatens the ecological environment and human life and property. Traditional firefighting methods mainly rely on manual firefighting or ground machinery, but in large-scale fire environments with complex terrain, these methods have many drawbacks, such as slow response speed, low firefighting efficiency, and limited coverage. With the rapid development of drone and robotic technologies, intelligent firefighting systems that combine ground and air capabilities have gradually become a research hotspot.

[0003] In existing technologies, many studies and technical solutions have attempted to use drones and robots in collaboration for firefighting. For example, some drones are equipped with sensors such as lidar and infrared cameras to detect fire locations, fire spread, and forest terrain; some firefighting robots use sensors to acquire environmental data and plan routes to perform firefighting tasks. However, these technologies still have some limitations, especially in route planning, fire point identification, fire spread prediction, and dynamic adjustment of firefighting strategies, and have not yet achieved truly intelligent collaboration.

[0004] While existing technologies offer new approaches to fire suppression, several technical bottlenecks remain. For example, current path planning methods often overlook the dynamic changes of fire points and the speed of fire spread, resulting in suboptimal planned paths that cannot respond in real time to changes caused by fire spread. Although existing fire point identification and extinguishing operations employ various sensors and algorithms, they largely rely on static judgments, failing to fully consider the dynamic changes of flames and the flexibility of extinguishing strategies, thus limiting the improvement of extinguishing efficiency. The collaborative work between robots and drones in most existing technologies is still imperfect, lacking effective information sharing and cooperation mechanisms, making it impossible to efficiently complete fire suppression tasks in complex environments. Therefore, there is an urgent need for a new technological solution that can plan optimal paths in real time in dynamically changing fire environments, accurately locate fire sources, and intelligently adjust extinguishing strategies to improve extinguishing efficiency and success rate. Summary of the Invention

[0005] The purpose of this invention is to provide a path planning method and system for forest fire fighting robots based on ground-air collaboration, which improves the efficiency and intelligence of fire fighting operations through the collaborative work between drones and fire fighting robots.

[0006] The objective of this invention is achieved through the following technical solution:

[0007] A path planning method for forest fire fighting robots based on ground-air cooperation includes the following steps:

[0008] Step S1: The drone uses a lidar to scan the forest disaster area from the air, generating a 3D point cloud model of the drone. Infrared and visible light cameras are used to obtain real-time information on the latitude and longitude of each fire point, the size of the fire point, and the intensity of the fire.

[0009] Step S2: The wind-powered firefighting robot establishes a communication connection with the drone, acquires a three-dimensional spatial model of the surrounding environment through lidar, compares it with the three-dimensional point cloud model of the drone, and locates the information of the wind-powered firefighting robot.

[0010] Step S3: Calculate the route length and time information of the wind-powered fire extinguishing robot to reach the fire point. Based on the above information, the size of the fire point area and the fire spread speed, construct a path planning mathematical model with the goal of minimizing the degree of damage.

[0011] Step S4: By integrating the global satellite navigation system and the inertial navigation system, obtain the speed, position, and attitude information of the wind-powered fire extinguishing robot, and dynamically update and plan the operation path of the wind-powered fire extinguishing robot based on the fire point information;

[0012] Step S5: After reaching the target fire point, the fire source is identified by the camera on the wind-powered fire extinguishing robot, the safe fire extinguishing distance is calculated using the flame detection model results, and the angle of the wind duct is dynamically adjusted according to the flame state to carry out the fire extinguishing operation.

[0013] A path planning system for a forest wind-powered firefighting robot based on ground-air collaboration, implementing the above method, includes a ground-air communication module, an identification and positioning module, a navigation and planning module, and a fire detection and extinguishing module, wherein:

[0014] The ground-to-air communication module is used to enable wireless communication between the ground control station and the drone and wind-powered firefighting robot, ensuring real-time data transmission and command reception.

[0015] The identification and positioning module is used to identify the surrounding environment in real time and determine the position of the wind-powered fire extinguishing robot itself through the sensors carried by the wind-powered fire extinguishing robot.

[0016] The navigation planning module is used to calculate and optimize the optimal path for the fire extinguishing robot to reach the target fire point based on the wind-powered fire extinguishing robot's own position and surrounding environment information provided by the identification and positioning module, combined with the target fire point position and fire spread speed parameters, and adjust the path in real time according to the dynamic changes of the fire point.

[0017] The fire detection and extinguishing module is used to identify the position, shape, and height of the flame in real time through the camera and flame detection model mounted on the wind-powered fire extinguishing robot, and dynamically adjust the angle of the wind duct and the wind force according to the changes in the flame to perform fire extinguishing operations.

[0018] Compared with the prior art, the present invention has the following advantages:

[0019] (1) Based on the path planning theory of UAVs and vehicles, a path planning model with dynamic fire data, multi-factor weights of robots and the goal of minimizing the degree of disaster is constructed. It can dynamically evaluate and select the optimal path to minimize the damage of fire to forests.

[0020] (2) By adopting the improved A* algorithm, the computational efficiency and accuracy of path planning are further improved by optimizing the cost function and adjusting the heuristic weights. The optimal path can be generated more quickly, ensuring the timely execution of firefighting tasks.

[0021] (3) The wind-powered fire extinguishing robot monitors the rate of change of flame height in real time and dynamically adjusts the angle of the wind duct at different stages of fire to ensure the flexibility and effectiveness of fire extinguishing operations, thereby improving the success rate of fire extinguishing and effectively suppressing the spread of fire. Attached Figure Description

[0022] Figure 1 A flowchart of a path planning method for a forest wind-powered firefighting robot based on ground-air collaboration;

[0023] Figure 2 A schematic diagram illustrating the air duct angle adjustment strategy for a wind-powered fire extinguishing robot.

[0024] Figure 3 This is a schematic diagram of the path planning system for a forest wind-powered firefighting robot based on ground-air collaboration. Detailed Implementation

[0025] The technical solution of the present invention will be further described below with reference to the accompanying drawings, but it is not limited thereto. Any modifications or equivalent substitutions to the technical solution of the present invention that do not depart from the spirit and scope of the technical solution of the present invention should be covered within the protection scope of the present invention.

[0026] This invention provides a path planning method for forest fire fighting robots based on ground-air collaboration. The method utilizes a drone equipped with lidar and infrared / visible light cameras to perform aerial scanning of the disaster area, acquiring real-time spatial information and fire intensity of fire points, and transmitting the data to a control center. The drone communicates with a wind-powered fire fighting robot, sharing a 3D point cloud model and performing environmental calibration and positioning. An improved A* algorithm is applied for path planning, constructing a mathematical model with the goal of minimizing damage based on the distance to the fire point, time, fire area, and fire spread speed. The wind-powered fire fighting robot's operational path is dynamically updated through the fusion of a global satellite navigation system and an inertial navigation system. Upon reaching the target fire point, the wind-powered fire fighting robot identifies the fire source, calculates the safe firefighting distance using the flame detection model results, and dynamically adjusts the wind tunnel angle according to the flame state to perform firefighting operations. Figure 1 As shown, the specific steps include the following:

[0027] Step S1: The drone, equipped with a lidar sensor, performs an aerial scan of the affected forest area to generate a 3D point cloud model. Infrared and visible light cameras are used to acquire real-time information on the latitude and longitude of each fire point, its area, and the intensity of the fire. The specific steps are as follows:

[0028] Step S11: The UAV is equipped with a high-precision lidar and performs aerial scanning of the forest disaster area according to the predetermined flight path to generate UAV three-dimensional point cloud data.

[0029] Step S12: Use an infrared camera to detect areas of abnormal temperature and mark the location of the fire point, and use a visible light camera to confirm the characteristics of the fire point.

[0030] Step S13: Transmit information such as the location, area, and intensity of the fire to the control center in real time to provide input for subsequent fire extinguishing planning.

[0031] Step S2: The wind-powered firefighting robot establishes a communication connection with the drone, acquires a 3D spatial model of the surrounding environment using lidar, compares it with the drone's 3D point cloud model, and locates the robot's information. The specific steps are as follows:

[0032] Step S21: The wind-powered fire extinguishing robot activates its lidar and scans the surrounding environment multiple times to acquire multiple sets of local 3D point cloud data;

[0033] Step S22: Use rigid transformation to align the local 3D point cloud data obtained from different scans to obtain a unified local 3D point cloud set. Downsample the local 3D point cloud set to reduce the density of the point cloud and remove isolated and abnormal noise points. Establish a coordinate system based on the current position of the wind-powered fire extinguishing robot.

[0034] Step S23: Based on the pre-provided global point cloud model of the UAV and the preliminary positioning of the wind-powered fire extinguishing robot, each point in the current local three-dimensional point cloud set is paired with the nearest point in the UAV's three-dimensional point cloud data. The rotation and translation transformations are calculated based on the matching points, and the local point cloud is aligned with the global point based on the point cloud features to locate the position of the wind-powered fire extinguishing robot.

[0035] Step S3: Calculate the route length and time information for the wind-powered fire extinguishing robot to reach the fire point. Based on the above information, the size of the fire point, and the fire spread speed, construct a path planning mathematical model with the goal of minimizing the damage. The shortest path from position i to fire point j can be obtained through this function, and the path distance is L. ij The specific steps are as follows:

[0036] Step S31: Apply the improved A* path planning algorithm to generate the optimal paths to each fire point, where:

[0037] The cost function of the improved A* path planning algorithm is as follows:

[0038] F n =G n +(H n +1) k ·H n

[0039] Among them, F n G represents the estimated cost from the starting node to the current node n; n H represents the actual cost from the starting node to the current node n, i.e., the distance traveled or the actual cost incurred from the starting node to the current node; n The heuristic cost estimate from the current node n to the target node is represented by k, which is the predicted distance between the current node and the target node. k is the weighting coefficient of the heuristic function, used to adjust the influence of the heuristic cost estimate on the total cost. The improved A* path planning algorithm introduces a dynamic weighting function (H... n +1) k When a node is far from the target, the heuristic function weights can be increased to make the algorithm more dependent on H. n This accelerates the search process, allowing for faster progress towards the target; as nodes approach the target, it reduces H... n The weighting coefficients are adjusted to reduce reliance on heuristic functions, prevent over-search, and ensure the algorithm focuses more on the actual path cost G. n This increases the probability of finding the optimal path;

[0040] Heuristic function H n The formula is as follows:

[0041]

[0042] Where (x1, y1) are the coordinates of the current node n; (x2, y2) are the coordinates of the target node; the heuristic function H n Based on Euclidean distance calculation, it represents the straight-line distance from the current node to the target node, ensuring that path planning can select the shortest path possible;

[0043] Step S32: Based on the parameters of the wind-powered fire extinguishing robot's route length and time to reach the fire point, the fire point area, and the fire spread rate, a target function for the wind-powered fire extinguishing robot is constructed with the goal of minimizing the damage. This target function determines which fire point to prioritize for extinguishing when multiple fire points occur, and plans the order of extinguishing multiple fire points to achieve the minimum damage, based on these parameters.

[0044] The objective function formula is as follows:

[0045] F min =min{F ij |i∈N1,j∈N2}

[0046] Among them, F ij Let be the cost for the wind-powered fire extinguishing robot to reach the fire location j from its current position i, where i∈N1, N1 is the set of current positions of the wind-powered fire extinguishing robot, and j∈N2, N2 is the set of target fire locations.

[0047]

[0048] Among them, L ij T represents the path length of the wind-powered firefighting robot from position i to fire point j; ij The time taken for the wind-powered firefighting robot to travel from location i to fire point j; A j V represents the area of ​​fire point j; j Let ε1, ε2, ε3, and ε4 represent the fire spread rate at fire point j; L represents the fire spread rate at fire point j. ij T ij A j and V j The weights; λ and μ are the penalty coefficients for path time and fire spread rate;

[0049] Calculate the fire spread rate V j The formula is as follows:

[0050]

[0051] Among them, A j (t1) represents A at time t1. j The size of the area, A j (t2) represents time t2, A jThe size of the area.

[0052] Step S4: By fusing the global satellite navigation system and the inertial navigation system, the speed, position, and attitude information of the wind-powered fire extinguishing robot are obtained. The operation path of the wind-powered fire extinguishing robot is dynamically updated and planned based on the fire point information. The specific steps are as follows:

[0053] Step S41: Use the Global Navigation Satellite System and the Inertial Navigation System to fuse position, velocity and attitude, and combine real-time positioning and mapping technology to complete high-precision navigation;

[0054] Step S42: Recalculate and adjust the operation path based on the real-time updated fire point data from the drone and the path deviation status of the wind-powered fire extinguishing robot.

[0055] Step S5: Upon reaching the target fire point, the fire source is identified using the camera mounted on the wind-powered fire extinguishing robot. The safe fire extinguishing distance is calculated using the flame detection model results, and the angle of the wind duct is dynamically adjusted according to the flame condition to carry out the fire extinguishing operation. The specific steps are as follows:

[0056] Step S51: The wind-powered firefighting robot acquires real-time video of the flame area using a camera, and extracts the flame area and spatial location information using the YOLO flame detection model, wherein:

[0057] The formula for calculating the spatial position (X, Y, Z) of a flame is as follows:

[0058]

[0059] Where X and Y represent the horizontal and vertical offsets of the flame in the camera coordinate system; Z represents the depth distance between the flame and the camera; f x and f y For camera focal length; x c and y c x represents the pixel coordinates of the flame in the image; i and y i d represents the coordinates of the image center; d is the depth value, i.e., the actual distance of the flame.

[0060] Step S52: Based on the pixel coordinates and depth information obtained in step S51, construct... Figure 2 The illustrated air duct angle adjustment strategy describes how the wind-powered fire extinguishing robot adjusts the angle and wind force of the air duct according to different stages of the fire. When the flame height does not increase significantly or the flame begins to decrease, the air duct angle remains close to 0° parallel to the horizontal direction. The air duct is aimed at the top of the flame to suppress it, spraying strong wind upwards and swinging the air duct to dissipate heat. When the flame height continues to decrease or the fire weakens, the air duct will gradually move to the bottom of the flame to extinguish it. The air duct angle will gradually increase from 0°, with a maximum angle limited to 45°, and finally point to the bottom of the flame to cut and quickly extinguish the fire source.

[0061] The air duct angle adjustment strategy is as follows:

[0062]

[0063] Where, θ ft Δh represents the angle between the ventilation duct and the horizontal direction at time t. fire h represents the rate of change of flame height, that is, the amount of change in flame height per unit time. fire d1 represents the flame height at time t; d1 is the horizontal distance between the duct and the flame; α is the sensitivity coefficient, used to adjust the sensitivity of the duct angle in response to changes in flame height.

[0064] This invention also provides a path planning system for a forest wind-powered firefighting robot based on ground-air collaboration, such as... Figure 3 As shown, the system includes an air-to-ground communication module, an identification and positioning module, a navigation and planning module, and a detection and extinguishing module, wherein:

[0065] The ground-to-air communication module is used to enable wireless communication between the ground control station and the drone and wind-powered firefighting robot, ensuring real-time data transmission and command reception.

[0066] The identification and positioning module is used to identify the surrounding environment in real time and determine the position of the wind-powered fire extinguishing robot itself through the sensors carried by the wind-powered fire extinguishing robot.

[0067] The identification and positioning module includes a lidar unit and an inertial measurement unit. The lidar unit is used to scan the surrounding environment to generate three-dimensional point cloud data, and the inertial measurement unit is used to provide attitude and motion state information of the wind-powered fire extinguishing robot. The precise positioning of the wind-powered fire extinguishing robot and its surrounding environment is achieved through a data fusion algorithm.

[0068] The navigation planning module is used to calculate and optimize the optimal path for wind-powered fire suppression to reach the target fire point based on the location and surrounding environment information provided by the identification and positioning module, combined with the target fire point location and fire spread speed parameters, and adjust the path in real time according to the dynamic changes of the fire point.

[0069] The fire detection and extinguishing module is used to identify the position, shape, and height of the flame in real time through the camera and flame detection model mounted on the wind-powered fire extinguishing system, and dynamically adjust the angle and wind force of the wind duct of the wind-powered fire extinguishing system according to the changes in the flame to perform fire extinguishing operations.

Claims

1. A path planning method for forest fire fighting robots based on ground-air collaboration, characterized in that... The method includes the following steps: Step S1: The drone uses a lidar to scan the forest disaster area from the air, generating a 3D point cloud model of the drone. Infrared and visible light cameras are used to obtain real-time information on the latitude and longitude of each fire point, the size of the fire point, and the intensity of the fire. Step S2: The wind-powered firefighting robot establishes a communication connection with the drone, acquires a three-dimensional spatial model of the surrounding environment through lidar, compares it with the three-dimensional point cloud model of the drone, and locates the information of the wind-powered firefighting robot. Step S3: Calculate the route length and time information for the wind-powered firefighting robot to reach the fire point. Based on the above information, the size of the fire point, and the fire spread speed, construct a path planning mathematical model with the goal of minimizing the damage. The specific steps are as follows: Step S31: Application Improvement The path planning algorithm generates the optimal path to each fire point, where: improve The cost function of the path planning algorithm is as follows: in, This represents the distance from the starting node to the current node. Estimated costs; From the starting node to the current node The actual cost, that is, the distance traveled or the actual cost incurred from the starting node to the current node; To start from the current node The heuristic cost estimate to the target node is the expected distance from the current node to the target node; These are the weighting coefficients of the heuristic function, used to adjust the degree of influence of the heuristic estimation cost on the total cost; Step S32: Based on the parameters of the wind-powered fire extinguishing robot's route length and time to reach the fire point, the fire area, and the fire spread rate, construct the objective function for the wind-powered fire extinguishing robot with the goal of minimizing the damage, where: The objective function formula is as follows: in, For the wind-powered fire extinguishing robot from its current location Reaching the fire location The value of the price, among which , This is a collection of the current locations of the wind-powered firefighting robots. , The set of target fire locations; The The formula is as follows: in, For wind-powered firefighting robots from location To the fire point Path length; For wind-powered firefighting robots from location Reaching the fire point The time spent; For fire point The size of the area; For fire point The speed at which the fire spreads; , , , for , , and The weights; and The coefficients for the penalty terms of path time and fire spread rate; Step S4: By integrating the global satellite navigation system and the inertial navigation system, obtain the speed, position, and attitude information of the wind-powered fire extinguishing robot, and dynamically update and plan the operation path of the wind-powered fire extinguishing robot based on the fire point information; Step S5: Upon reaching the target fire point, the fire source is identified using the camera mounted on the wind-powered fire extinguishing robot. The safe fire extinguishing distance is calculated using the flame detection model results, and the angle of the wind duct is dynamically adjusted according to the flame condition to carry out the fire extinguishing operation. The specific steps are as follows: Step S51: The wind-powered fire extinguishing robot acquires real-time video of the flame area by carrying a camera, and uses the YOLO flame detection model to extract the flame area and flame spatial location information; Step S52: Based on the pixel coordinates and depth information obtained in step S51, a wind tunnel angle adjustment strategy is constructed. The wind-powered fire extinguishing robot adjusts the angle and wind force of the fire extinguishing wind tunnel according to different stages of the flame. When the flame height does not increase significantly or the flame begins to decrease, the wind tunnel angle remains close to 0° parallel to the horizontal direction. The wind tunnel is aimed at the top of the flame to suppress it, spraying strong wind upwards and swinging the wind tunnel to dissipate heat. When the flame height continues to decrease or the fire weakens, the wind tunnel will gradually move to the bottom of the flame to extinguish the fire. The wind tunnel angle will gradually increase from 0°, with a maximum angle limited to 45°, and finally point to the bottom of the flame to cut and quickly extinguish the fire source. The air duct angle adjustment strategy is as follows: in, This represents the angle between the ventilation duct and the horizontal direction at time t; This represents the rate of change of flame height, that is, the amount of change in flame height per unit time. This represents the flame height at time t; The horizontal distance between the air duct and the flame; This is the sensitivity coefficient, used to adjust the sensitivity of the duct angle in response to changes in flame height.

2. The path planning method for forest fire fighting robots based on ground-air collaboration according to claim 1, characterized in that... The specific steps of step S1 are as follows: Step S11: The UAV is equipped with a high-precision lidar and performs aerial scanning of the forest disaster area according to the predetermined flight path to generate UAV three-dimensional point cloud data. Step S12: Use an infrared camera to detect areas of abnormal temperature and mark the location of the fire point, and use a visible light camera to confirm the characteristics of the fire point. Step S13: Transmit the location, area, and intensity of the fire to the control center in real time to provide input for subsequent fire extinguishing planning.

3. The path planning method for forest fire fighting robots based on ground-air collaboration according to claim 1, characterized in that... The specific steps of step S2 are as follows: Step S21: The wind-powered fire extinguishing robot activates its lidar and scans the surrounding environment multiple times to acquire multiple sets of local 3D point cloud data; Step S22: Use rigid transformation to align the local 3D point cloud data obtained from different scans to obtain a unified local 3D point cloud set. Downsample the local 3D point cloud set to reduce the density of the point cloud and remove isolated and abnormal noise points. Establish a coordinate system based on the current position of the wind-powered fire extinguishing robot. Step S23: Based on the pre-provided global point cloud model of the UAV and the preliminary positioning of the wind-powered fire extinguishing robot, each point in the current local three-dimensional point cloud set is paired with the nearest point in the UAV's three-dimensional point cloud data. The rotation and translation transformations are calculated based on the matching points, and the local point cloud is aligned with the global point based on the point cloud features to locate the position of the wind-powered fire extinguishing robot.

4. The path planning method for forest fire fighting robots based on ground-air collaboration according to claim 1, characterized in that... The heuristic cost estimation The formula is as follows: in,( , ) is the current node The coordinates; , ) represents the coordinates of the target node.

5. The path planning method for forest fire fighting robots based on ground-air cooperation according to claim 1, characterized in that... The speed of fire spread The calculation formula is as follows: in, for time The size of the area, for time The size of the area.

6. The path planning method for forest fire fighting robots based on ground-air collaboration according to claim 1, characterized in that... The specific steps of step S4 are as follows: Step S41: Use the Global Navigation Satellite System and the Inertial Navigation System to fuse position, velocity and attitude, and combine real-time positioning and mapping technology to complete high-precision navigation; Step S42: Recalculate and adjust the operation path based on the real-time updated fire point data from the drone and the path deviation status of the wind-powered fire extinguishing robot.

7. The path planning method for forest fire fighting robots based on ground-air cooperation according to claim 1, characterized in that... The spatial location of the flame ( , , The calculation formula for ) is as follows: in, , The horizontal and vertical offsets of the flame in the camera coordinate system; The depth distance between the flame and the camera; and The focal length of the camera; and These are the pixel coordinates of the flame in the image; and The coordinates of the image center; This represents the depth value, i.e., the actual distance of the flame.

8. A path planning system for a forest wind-powered firefighting robot based on ground-air collaboration, implementing the method of any one of claims 1-7, characterized in that... The system includes an air-to-ground communication module, an identification and positioning module, a navigation and planning module, and a detection and extinguishing module, wherein: The ground-to-air communication module is used to enable wireless communication between the ground control station and the drone and wind-powered firefighting robot, ensuring real-time data transmission and command reception. The identification and positioning module is used to identify the surrounding environment in real time and determine the position of the wind-powered fire extinguishing robot itself through the sensors carried by the wind-powered fire extinguishing robot. The navigation planning module is used to, based on the wind-powered fire extinguishing self-position and surrounding environment information provided by the identification and positioning module, and combined with the target fire location and fire spread rate parameters, improve... The path planning algorithm calculates and optimizes the best path for the fire-fighting robot to reach the target fire point, and adjusts the path in real time according to the dynamic changes of the fire point. The fire detection and extinguishing module is used to identify the position, shape, and height of the flame in real time through the camera and flame detection model mounted on the wind-powered fire extinguishing robot, and dynamically adjust the angle of the wind duct and the wind force according to the changes in the flame to perform fire extinguishing operations.

Images