Forest steppe fire line following mobile automatic fire extinguishing method and system

Abstract

The application discloses a forest and grassland fire line following type mobile automatic fire extinguishing method and system. The method comprises the following steps: acquiring fire field information through a binocular camera and an infrared thermal imaging camera, extracting a fire line three-dimensional position, a temperature area and a moving direction by adopting a YOLOv5 segmentation and a frame difference method, and forming a fire line state model; constructing a prior environment map by using a laser radar and a double RTK, superimposing the fire line state and the high temperature area, and establishing a dynamic fire field environment map containing a comprehensive passing cost; a fire extinguishing task planner divides the continuous fire line into fire line segment sub-tasks, calculates discrete spraying angles of the fire gun through coordinate transformation, and dynamically adjusts the chassis speed according to the remaining fire degree; in the spraying process, the suppression effect is judged according to the visible light flame area and the infrared temperature change, the next fire line segment is switched to after the suppression is completed, and otherwise the map and the task are updated. The application realizes the autonomous following and continuous suppression of the continuous fire line of the grassland fire, and improves the automation degree and the adaptability of the fire extinguishing.

Description

Technical Field

[0001] This application relates to the field of automatic fire extinguishing technology for forest and grassland fires, specifically to a method and system for automatic fire extinguishing that follows the fire line in forest and grassland fires. Background Technology

[0002] Fire is one of the most common types of disasters. For fires involving general buildings, warehouses, or equipment, the burning area is usually concentrated in a relatively fixed location, allowing firefighters or firefighting equipment to approach, aim, and spray firefighting equipment based on the location of the fire source. However, forest and grassland fires are significantly different from the above-mentioned fires. Forests, grasslands, wastelands, shrublands, forest edges, and pastures typically have characteristics such as continuous distribution of combustibles, large spatial range, and complex terrain conditions. Once a fire breaks out, the flames will spread rapidly along continuous fuels such as grasslands, shrubs, and fallen leaves, and will continuously change their spread direction and burning range under the influence of factors such as wind speed, wind direction, slope, and vegetation type.

[0003] In forest and grassland fires, the fire target is often not an isolated hot spot, but a fire line formed by multiple continuous burning areas. This fire line usually has a certain length, direction of extension, and tendency to move, and it will continue to advance into unburned areas over time. If only a local hot spot is sprayed, it may only suppress a small section of the fire line, while the uncovered fire line may continue to spread.

[0004] Currently, forest and grassland fire fighting mainly relies on manual firefighting, firebreak construction, fire truck spraying, and aerial firefighting. Manual firefighting carries high safety risks; fire trucks have limited mobility in complex, unstructured terrain; and aerial firefighting struggles to continuously suppress residual high temperatures on the ground. With the development of mobile robot technology, utilizing mobile firefighting platforms to replace personnel in dangerous areas for firefighting operations is becoming an important trend. However, existing firefighting robots or automatic fire monitors primarily target localized fire points. Their typical technical approach involves: collecting fire scene environmental information—identifying the fire source or high-temperature target—constructing an obstacle map—planning a path to the vicinity of the fire source—controlling the chassis to approach the fire source—controlling the fire monitor to aim at the center of the fire source. This approach is ill-suited to the characteristics of forest and grassland fires: large fire lines, continuously changing locations, and linear distribution of fire targets. It is prone to problems such as discontinuous fire coverage, mismatch between the platform's travel path and changes in the fire line, and delayed target spraying by the fire monitor.

[0005] Therefore, there is an urgent need for a mobile automatic fire extinguishing method and system capable of autonomously identifying, following, and continuously suppressing continuous fire lines in forest and grassland fires. Summary of the Invention

[0006] In view of this, this application aims to provide a method and system for automatic fire extinguishing that follows the fire line in forest and grassland fires, enabling the fire extinguishing platform to maintain a safe distance from the fire line in complex grassland fire environments, move autonomously along the direction of the fire line, and continuously complete fire extinguishing operations during the movement.

[0007] To achieve the above objectives, according to one aspect of this application, a method for automatically extinguishing forest and grassland fires by following the fire line is provided. The method includes the following steps:

[0008] Step S1: Obtain fire scene perception information through the multi-source perception unit mounted on the mobile fire extinguishing platform, segment and identify the fire line area based on the visible light image in the perception information, combine infrared thermal imaging information to verify authenticity and identify high temperature areas, and extract the fire line movement direction through continuous inter-frame motion analysis to form fire line status information including fire line three-dimensional information, fire line temperature area and fire line movement direction.

[0009] Step S2: Construct a priori environment map using lidar and dual RTK positioning and orientation modules. The priori environment map includes terrain structure, obstacle distribution, and passable area information. Overlay the fire line status information and high-temperature area obtained in step S1 onto the priori environment map, and introduce safety distance constraints and comprehensive passage costs to generate a dynamic fire scene environment map.

[0010] Step S3: The fire extinguishing task planner divides the continuous fire line into one or more fire line segments based on the dynamic fire scene environment map, determines the safe following path of the fire line and the target of the fire monitor spray, and through the coordinated control of the chassis and the fire monitor, enables the mobile fire extinguishing platform to move autonomously along the safe following path, while controlling the fire monitor to continuously spray and suppress the current fire line segment.

[0011] Step S4: During the spray suppression process, the suppression effect of the current fire line segment is judged in real time based on the changes in visible light flame and infrared temperature. If the suppression completion conditions are met, the process switches to the next fire line segment sub-task. If the conditions are not met or the fire line status changes, the fire line status information, dynamic fire scene environment map, and fire extinguishing task are dynamically updated to achieve closed-loop control.

[0012] Preferably, step S1 specifically includes:

[0013] The YOLOv5 image segmentation model was used to segment the flame region in the visible light image acquired by the binocular camera to obtain a fire line segmentation mask.

[0014] The displacement changes of the fire line segmentation mask in adjacent frames are analyzed using the frame difference method. Combined with the spatial information obtained from the binocular camera, the three-dimensional spatial position of the fire line is determined. and direction of movement ;

[0015] The high-temperature area in the infrared thermal imaging image is matched with the visible light segmentation result. When the visible light flame area and the infrared high-temperature area overlap, it is determined to be a real fire line. The infrared temperature distribution is used to distinguish between the burned side and the unburned side of the fire line.

[0016] Through the above steps, fire line status information including three-dimensional fire line information, fire line temperature range, and fire line movement direction is generated.

[0017] Preferably, in step S2 above, the method for constructing the prior environment map includes:

[0018] Point cloud information of the surrounding environment of the mobile platform is collected by lidar, and then filtered, segmented, and extracted to obtain basic terrain structure information.

[0019] The real-time position and vehicle heading of the mobile platform are obtained through dual RTK positioning and orientation modules, and the local point cloud data collected by the lidar is converted into a unified map coordinate system.

[0020] Based on changes in ground elevation, slope, and obstacle distribution, the environment is divided into passable areas, potentially passable areas, and impassable areas, generating a navigation map in the form of a cost map. Areas with obstacles, ditches, and excessive slopes are assigned higher passable costs, while flat ground or areas suitable for chassis passage are assigned lower passable costs.

[0021] Preferably, in step S2, the method for constructing the dynamic fire scene environment map specifically includes:

[0022] Let the prior environment map be ,in, Indicates the first in the map A grid cell is represented as:

[0023] in, Indicates the spatial position of the grid cell. This indicates information about slope or terrain undulation. Indicates the state of the obstacle. This indicates the basic accessibility of the grid;

[0024] Let the current fire status be... ,in The three-dimensional position of the fire line. The direction of fire line movement. Temperature distribution;

[0025] For any grid Calculate its position relative to the fire line. distance ,definition

[0026]

[0027] in The distance affected by the fire line, among which, This indicates that the grid belongs to the fire zone or the fire-affected zone. This indicates that the grid does not belong to the fire-affected area;

[0028] Let the first The temperature value corresponding to each grid cell is The high temperature threshold is The high-temperature region is represented as:

[0029]

[0030] in, This indicates that the grid is a high-temperature region. This indicates that the grid has not met the high temperature judgment condition;

[0031] Build a grid Comprehensive traffic costs ,

[0032]

[0033] in, Indicates the first The total cost of each grid cell at the current moment. Indicates the cost of traversing the terrain. Indicates the cost of obstacles. This indicates the costs associated with maintaining a safe distance from the line of fire. , These are the weighting coefficients for each cost item;

[0034] Candidate grids also need to meet the following requirements ,in For minimum safe distance, This refers to the effective spray distance of the fire monitor.

[0035] Finally, a dynamic fire scene environment map is generated. .

[0036] Preferably, in step S3, the specific method for dividing the continuous fireline into fireline segment sub-tasks includes:

[0037] Let the current three-dimensional point set of the fire line be... ;

[0038] Based on the continuity of the fire line space, the length of the fire line, the temperature distribution, and the effective spray coverage of the fire monitor, Divided into Each of the following is a separate, unrelated sentence: "In each of the following sub-missions..." This includes the three-dimensional point set of the fire line segment, the temperature region corresponding to the fire line segment, and the direction of movement;

[0039] The fire extinguishing task planner queries the terrain and accessibility information near each fire line segment in the dynamic fire scene environment map, and selects the side with lower access cost and a distance from the fire line that meets the requirements of safe distance and effective range of fire monitors as the fire line following side.

[0040] Preferably, in step S3, the specific method by which the fire extinguishing task planner converts the three-dimensional fire line information into the fire monitor control angle includes:

[0041] Establish map coordinate system Vehicle coordinate system Binocular camera coordinate system Fire monitor coordinate system ;

[0042] Let the coordinates of a point on the fire line in the camera coordinate system be... Then its position in the map coordinate system is ,in For the vehicle pose transformation matrix provided by dual RTK, The mounting matrix of the camera relative to the vehicle body;

[0043] set up ,in The mounting matrix of the fire monitor relative to the vehicle body;

[0044] set up The required horizontal rotation angle for the fire monitor to be aimed at that point. and pitch angle for

[0045]

[0046]

[0047] in, The pitch compensation angle is calculated to account for the drop of the extinguishing medium jet, the injection pressure, and the injection distance.

[0048] Preferably, in step S3, the chassis-fire monitor coordinated control includes a method for dynamically adjusting the chassis driving speed: defining the remaining fire intensity of the current fire segment. :

[0049]

[0050] in, Indicates the first The area of ​​the visible flame region of each fire segment at the current moment. This indicates the area of ​​the flame zone when the fire line segment begins to be suppressed. This indicates the average infrared temperature of the current fire segment. This indicates the average infrared temperature at the start of pressing. Indicates the safe temperature threshold. and These are the weighting coefficients;

[0051] Based on the remaining fire intensity and the vehicle's heading deviation and safety distance deviation Generate chassis driving speed , ,

[0052] in, This indicates the permitted higher driving speed. Indicates the minimum following speed when the jet is executed. and These are the weighting coefficients;

[0053] when The current fire segment suppression is completed at the time of determination. To suppress the completion threshold, the task planner marks the fire segment as completed and increases the chassis speed.

[0054] Preferably, in step S4, the method for determining the pressing completion condition includes:

[0055] Record the visible flame area when the current fire line segment begins to suppress. and average infrared temperature ;

[0056] Real-time calculation of the current flame area during fire extinguishing. and current average infrared temperature If both conditions are met:

[0057] and ,

[0058] Then it is determined that the fire line segment has been suppressed, where The flame area attenuation threshold. The temperature safety threshold;

[0059] Once a fire line segment is determined to be suppressed, the fire suppression task planner removes the fire line segment from the current task queue or marks it as completed, while updating the fire line status and high-temperature zone status in the dynamic fire scene environment map.

[0060] According to another aspect of this application, a forest and grassland fire-following mobile automatic fire extinguishing system is provided for implementing the above-mentioned method, characterized in that the system comprises:

[0061] The mobile chassis is used to carry the various components of the system and drive the system to move autonomously and safely along the fire line.

[0062] Fire monitors, mounted on mobile chassis, are capable of horizontal rotation and pitch adjustment, and are used to spray extinguishing media onto fire-prone areas.

[0063] The multi-source sensing unit includes a binocular camera, an infrared thermal imaging camera, a lidar, and a gimbal. The binocular camera is used to acquire visible light images of the fire line, the infrared thermal imaging camera is used to acquire the temperature distribution of the fire line and the surrounding area, the lidar is used to acquire terrain structure and obstacle information, and the gimbal is used to support and adjust the observation direction of the binocular camera and the infrared thermal imaging camera so that they are continuously facing the fire line area.

[0064] The positioning and orientation unit uses a dual RTK module, with two RTK antennas arranged at different positions on the top of the vehicle body to obtain high-precision position and heading information of the mobile platform.

[0065] The control computing unit is electrically connected to the mobile chassis, fire monitor, multi-source sensing unit, and positioning and orientation unit, respectively. The control computing unit includes:

[0066] The fire status identification and modeling module is used to execute step S1;

[0067] The dynamic fire scene environment map construction module is used to execute step S2;

[0068] Firefighting task planner, used to execute task planning in step S3;

[0069] The chassis motion control module is used to execute the chassis motion control in step S3.

[0070] The fire monitor end control module is used to execute the fire monitor spray control in step S3;

[0071] The feedback update module is used to execute step S4.

[0072] Preferably, the fire monitor end control module receives the discrete spray angle points output by the fire extinguishing task planner as follows:

[0073] ,

[0074] And a smooth, continuous desired trajectory is generated through linear interpolation or spline interpolation.

[0075] , , ,

[0076] And limits are imposed on the angular velocity and angular acceleration of the fire monitor.

[0077] , ,

[0078] The fire monitor control module performs closed-loop control based on the deviation between the current actual angle and the desired angle. Let the current actual horizontal angle and pitch angle be... , The expected angles are respectively , Then the angle error is , ,

[0079] During the fire monitor's spraying process, the feedback update module continuously uses visible light flame changes collected by the binocular camera and temperature changes collected by the infrared camera to determine the current fire suppression effect, and feeds back the judgment results to the fire line status recognition and modeling module, the dynamic fire scene environment map construction module, and the fire extinguishing task planner.

[0080] The forest and grassland fire-fighting method and system provided in this application, which follows the fire line, utilizes binocular vision, infrared thermal imaging, and inter-frame motion analysis to structurally describe the fire line from three aspects: spatial location, temperature characteristics, and spread trend. It also breaks down continuous fire lines into executable sub-tasks, overcoming the limitations of traditional methods that simplify fires into single fire points. By deeply fusing the fire line status, high-temperature areas, and a priori environmental map constructed using lidar / RTK, a dynamic fire scene environment map incorporating comprehensive travel costs is generated. This allows path planning to simultaneously consider terrain accessibility, safe distance, and fire monitor range, significantly improving driving safety and firefighting reliability in complex grassland terrain.

[0081] Furthermore, the fire-fighting task planner converts fire targets into discrete angle points for fire monitors and dynamically adjusts the chassis speed based on fire suppression effectiveness, achieving coordinated control of movement and spraying. When the fire is still burning, low-speed movement ensures sufficient suppression; after suppression, accelerated movement ensures continuity and complete coverage of the fire-fighting process. The system continuously uses visible light and infrared feedback to assess suppression effectiveness and dynamically updates the map and tasks based on fire changes, forming an intelligent closed loop of perception-planning-execution-feedback. This effectively addresses dynamic changes such as fire advance, turning, and reignition, demonstrating a high degree of automation and strong on-site adaptability. Attached Figure Description

[0082] Figure 1 This is a schematic diagram of the composition of a forest and grassland fire-following mobile automatic fire extinguishing system according to an embodiment of this application.

[0083] Figure 2 This is a flowchart illustrating the automatic fire extinguishing method for forest and grassland fires that follows the fire line, according to an embodiment of this application. Detailed Implementation

[0084] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0085] like Figure 1 As shown, in one embodiment, the forest and grassland fire line following mobile automatic fire extinguishing system is installed on a tracked mobile platform, mainly including: mobile chassis, fire monitor (1), vision system (2), infrared thermal imaging camera, lidar (3), dual RTK positioning and orientation module and control computing unit.

[0086] The mobile chassis employs a tracked structure to enhance its mobility in unstructured terrain such as grasslands, wastelands, and slopes. The chassis houses the vehicle body and equipment mounting platform. A fire monitor is installed at the front or top of the vehicle body to spray water, foam, or other extinguishing agents onto the fire-prone area. The fire monitor has horizontal rotation and elevation adjustment capabilities, allowing it to adjust its spray direction according to the target area provided by the fire suppression planner.

[0087] The multi-source sensing unit mainly includes a vision system, an infrared camera, and a lidar. The vision system comprises a binocular camera mounted on a pan-tilt unit. This camera acquires visible light images and binocular video information of the fire line area, obtaining flame morphology, fire line outline, fire line boundary, and changes in the fire line across consecutive image frames. An infrared camera is mounted in the center to acquire temperature distribution information of the fire line area and its surrounding area, assisting in determining whether the fire line identified by visible light is a real combustion area and identifying residual high-temperature areas and suspected reignition areas. The pan-tilt unit supports and adjusts the observation direction of the vision system and infrared camera, ensuring they are continuously pointed towards the fire line or suspected fire line area, improving the continuity of fire line identification.

[0088] The lidar is installed on the upper sides and rear of the vehicle body to collect point cloud information around the mobile platform, acquiring environmental information such as terrain undulations, ditches, rocks, shrubs, obstacles, and passable areas. The data collected by the lidar is used to construct a priori environmental map, providing basic terrain and obstacle information for chassis path planning and dynamic fire scene environmental mapping.

[0089] The positioning and orientation unit employs a dual RTK module, with two RTK antennas positioned at different locations on the top of the vehicle to acquire high-precision position and heading information of the mobile platform. The position and orientation information provided by the dual RTK serves as a unified spatial coordinate reference, enabling data acquired by the binocular camera, infrared camera, and LiDAR to be converted into the same map coordinate system. Through this positioning and orientation unit, the system can uniformly represent 3D fire line information, terrain structure information, obstacle information, and vehicle pose information, providing a foundation for subsequent fire line modeling, environmental map construction, and follow-path planning.

[0090] The control and computing unit is located inside the mobile platform or the vehicle's equipment compartment and is used to run the automatic fire extinguishing method according to this application. In one embodiment, the control and computing unit includes a fire line status recognition and modeling module, a dynamic fire scene environment map construction module, a fire extinguishing task planner, a chassis motion control module, a fire monitor end control module, and a feedback update module. The fire line status recognition and modeling module receives information collected by a binocular camera and an infrared camera, extracts the three-dimensional information of the fire line, temperature regions, and movement direction, and converts continuous fire lines into fire line segments that can be extinguished. The dynamic fire scene environment map construction module overlays fire line status information and high-temperature regions on the prior environment map established by lidar and dual RTK, forming a dynamic fire scene environment map that includes terrain structure, obstacles, passable areas, fire line areas, high-temperature regions, safety distances, and vehicle pose.

[0091] The fire suppression task planner receives fire line status information and a dynamic fire scene environment map. Based on the fire line segment location, temperature zone, fire line movement direction, vehicle current position, terrain accessibility, and fire monitor spraying capability, it determines the fire line segment that needs to be prioritized and generates a fire line following path, chassis speed, and fire monitor spray target. The chassis motion control module, based on the following path and speed commands output by the fire suppression task planner, controls the moving chassis to move at a low speed along the safe side of the fire line, dynamically adjusting the speed according to the fire suppression situation. The fire monitor end control module, based on the discrete spray angle points output by the fire suppression task planner, generates a smooth fire monitor control trajectory, controlling the fire monitor to complete horizontal rotation, pitch adjustment, and continuous spraying.

[0092] The feedback update module continuously assesses the fire suppression effectiveness during firefighting. Based on visible flame changes captured by a binocular camera and temperature changes captured by an infrared camera, this module determines whether the current fire segment has been successfully suppressed and feeds the results back to the fire line status recognition and modeling module, the dynamic fire scene environment map construction module, and the firefighting task planner. Thus, the system forms a closed-loop control process from fire line perception, map construction, task planning, chassis following, fire monitor spraying, to effect feedback, enabling the mobile firefighting platform to autonomously follow and continuously suppress the grassland fire.

[0093] Reference Figure 2 , Figure 2 A schematic diagram illustrating the specific process of an automatic fire extinguishing method for moving the fire line of a forest or grassland fire according to an embodiment of this application is shown.

[0094] In the first phase, the system identifies the fire status.

[0095] After the system starts up, the binocular cameras acquire a video stream of the fire front at a frame rate of 30fps. The fire line status recognition and modeling module in the control computing unit processes each frame of the image.

[0096] First, a YOLOv5-based image segmentation model is used to segment the flame region and suspected fire line region in the current frame image, obtaining the fire line segmentation mask at the current moment. The YOLOv5 model has been pre-trained and can effectively identify flame patterns in grassland environments.

[0097] To determine the direction of fire movement, the system performs inter-frame motion analysis using continuous video frames captured by a binocular camera. Specifically, it compares the fire line segmentation results at adjacent times or several frames apart, analyzes the displacement changes of the fire line region in the image using the frame difference method, and obtains the motion trend of the leading edge of the fire line in continuous frames. Then, combined with the spatial position information obtained by the binocular camera (calculated through a disparity map), the fire line displacement in the image plane is converted into spatial direction changes, thereby determining the direction of fire movement or spread.

[0098] Temperature information acquired by an infrared thermal imaging camera is used to verify and supplement the visible light recognition results. The system matches high-temperature areas in the infrared image with fireline areas segmented by YOLOv5. If an area is identified as a fireline in the visible light image and also appears as a high-temperature area in the infrared image (e.g., temperature exceeding 150°C), the confidence level of identifying that area as a real fireline is high. If an area shows a flame-like color or shape in the visible light image, but its infrared temperature does not meet the fireline characteristics (e.g., below 80°C), the confidence level of identifying that area as a real fireline is reduced, thereby reducing false identifications caused by factors such as background color, sunlight reflection, and smoke interference.

[0099] In addition, infrared thermal imaging cameras are used to identify high-temperature areas near the fire line. In grassland fires, the burned areas after the fire line has passed usually retain residual high temperatures, while the temperature characteristics of unburned areas differ. Therefore, the system can combine the direction of fire line movement and infrared temperature distribution to determine which side of the area on either side of the fire line is more likely to be burned and which side is more likely to be unburned. This determination can serve as the basis for subsequent construction of dynamic fire scene environmental maps, distinguishing between high-temperature hazard areas, burned areas, and unburned areas, and assisting in determining which side of the fire line the mobile platform should follow for firefighting.

[0100] After the above processing, the system forms a simplified fire state model. This model mainly includes three types of information:

[0101] First, Fireline 3D Information P t Fireline 3D point set, 3D boundary, or 3D centerline obtained from a binocular camera and a parallax map;

[0102] Second, the fire wire temperature zone T t : High-temperature areas along the fire line, residual high-temperature areas, and suspected reignition high-temperature areas obtained from infrared thermal imaging cameras;

[0103] Third, the direction of fire movement is V. t The fire spread direction was obtained by analyzing the inter-frame difference of YOLOv5 segmentation results and continuous frames of stereo video.

[0104] The output of this step is not simply a flame detection frame, but a complete fireline status information containing three-dimensional spatial position, temperature range, and direction of movement. .

[0105] In the second stage, the system constructs a priori environment map.

[0106] While completing the initial fire-line awareness, the system uses LiDAR and dual RTK positioning and orientation modules to construct a priori environment map. This priori environment map is mainly used to describe the relatively stable terrain structure, obstacle distribution, and navigation space around the mobile platform.

[0107] The lidar is used to collect point cloud information of the environment surrounding the mobile platform, obtaining the spatial location of ground undulations, slope changes, ditches, rocks, shrubs, trees, and other obstacles. The system filters, segments the ground, and extracts obstacles from the lidar point cloud, distinguishing ground points, non-ground obstacle points, and areas with significant elevation changes, thereby obtaining basic terrain structure information in the grassland fire area.

[0108] The dual RTK positioning and orientation module is used to acquire the real-time location and heading information of the mobile platform. Using the high-precision positioning and orientation information provided by dual RTK, the system can determine the mobile platform's position in the global coordinate system and convert the local point cloud data collected by the LiDAR to a unified map coordinate system. In this way, as the mobile platform moves, the system can stitch together and update terrain and obstacle information collected at different times.

[0109] During the navigation map construction process, the system first establishes a local 3D environment model based on LiDAR point clouds. Then, combining dual RTK pose information, it transforms the local environment model into a global or local navigation coordinate system. Subsequently, based on changes in ground elevation, slope, and obstacle distribution, the system divides the environment into passable, potentially passable, and impassable areas, generating a navigation map for chassis motion planning. This navigation map is represented in the form of a cost map, where obstacles, ditches, and areas with excessively steep slopes are assigned higher passability costs (e.g., set to 100), while flat ground or areas suitable for chassis passage are assigned lower passability costs (e.g., set to 0). Areas with slopes greater than 15° have a cost of 50.

[0110] The prior environment map can be represented as ,in, Indicates the first in the map Each grid cell contains information such as its spatial location, terrain elevation, slope, obstacle status, and passability status, and can be represented as:

[0111]

[0112] in, Indicates the spatial location of the grid cell. This indicates information about slope or terrain undulation. Indicates the state of the obstacle. This indicates the basic accessibility of the grid.

[0113] In the third stage, the system constructs a dynamic fire scene environment map.

[0114] After obtaining the prior environment map, the system further overlays the fire line status information and infrared high-temperature areas onto the prior environment map to form a dynamic fire scene environment map. The prior environment map mainly provides information on terrain structure, obstacle distribution, and passable areas, while the dynamic fire scene environment map adds fire line location, fire line movement direction, high-temperature areas, safety distance constraints, and attributes of the areas on both sides of the fire line, which are used for subsequent fire line following path planning and fire monitor spray target generation.

[0115] Let the current fire status be:

[0116]

[0117] in, This represents the three-dimensional position information of the fire line obtained by the binocular camera. This indicates the direction of fire line movement obtained through inter-frame motion analysis. This represents the temperature distribution of the fire line and its surrounding area as obtained by an infrared thermal imaging camera.

[0118] For any grid cell in the map The system calculates its three-dimensional position relative to the current fire line. Distance between:

[0119]

[0120] when Less than the set fire line influence distance When the grid is considered to be within the fire line's influence range, it can be denoted as:

[0121]

[0122] in, This indicates that the grid belongs to the fire zone or the fire-affected zone. This indicates that the grid is not within the fire-affected area.

[0123] Simultaneously, the system identifies high-temperature areas based on temperature values ​​obtained from the infrared thermal imaging camera. Let the... The temperature value corresponding to each grid cell is The high temperature threshold is The high-temperature region can then be represented as:

[0124]

[0125] in, This indicates that the grid is a high-temperature area, which may correspond to an open flame area, a residual high-temperature area, or an area with a risk of reignition. This indicates that the grid has not met the high temperature judgment criteria.

[0126] To enable the dynamic fire scene environment map to directly serve mobile platform path planning, the system further constructs a comprehensive access cost for each grid cell. This cost consists of terrain cost, obstacle cost, fire line impact cost, and high-temperature zone cost, and can be expressed as:

[0127]

[0128] in, Indicates the first The total cost of each grid cell at the current moment; This indicates the cost of traversing terrain caused by topographical features such as slope and undulation. Indicates the cost of obstacles; Indicates the cost of the frontline area; This indicates the cost of operating in high-temperature areas; Indicates the costs associated with safe distance from the line of fire; These are the weighting coefficients for each cost.

[0129] Among them, the cost of safe distance The settings can be adjusted based on the distance between the grid unit and the fire line. For example, when the grid unit is too close to the fire line, the safety risk is high; when the distance is within the effective spray range of the fire monitor, it is more suitable as a candidate area for following the path. Therefore, the safety distance cost can be expressed as:

[0130]

[0131] in, This represents the desired safe following distance between the mobile platform and the fire line. The formula indicates that the closer the grid is to the fire line, the lower the safety distance cost; the cost increases when the grid is too close or too far from the fire line.

[0132] In actual path filtering, candidate grids should also meet the following requirements. ,in For minimum safe distance, This represents the effective spray distance of the fire monitor. The formula means that if the distance to the fire line is too far, it is safe, but the fire monitor cannot reach the target; if it is too close, it is dangerous.

[0133] Ultimately, the system generates a dynamic fire scene environment map for the current moment:

[0134]

[0135] This dynamic fire scene environment map not only includes the terrain structure, obstacles, and passable areas from the prior environment map, but also the current fire line area, high-temperature area, fire line movement direction, and the comprehensive passability cost of each grid. Using this map, the system can determine which areas are suitable for chassis passage, which areas are high-temperature hazard zones, which areas are near the fire line, and which areas are more suitable as safe fire-following paths.

[0136] In addition, the system can also combine the direction of fire movement. Based on infrared temperature distribution, the system distinguishes the areas on both sides of the fire line. Generally, the side the fire line is moving towards is more likely to be the unburned, expanding side, while the side moving in the opposite direction and exhibiting residual high temperatures is more likely to be the burned side. The system overlays this judgment onto a dynamic fire scene environment map to assist the fire suppression mission planner in determining which side of the fire line the mobile platform should travel along, thereby improving the reliability of fire-fighting while maintaining a safe distance.

[0137] In the fourth stage, the system plans and coordinates firefighting tasks.

[0138] The fire suppression mission planner uses a dynamic fire scene environment map as its primary input. This map includes 3D fire line information, fire line temperature zones, fire line movement direction, high-temperature hazard zones, terrain structure, obstacles, passable areas, real-time vehicle positions, and the cost of traversing each area. Based on this information, the fire suppression mission planner determines the fire line segments that require priority suppression and further generates discrete spray angle tasks for fire monitors, as well as travel speed and follow-motion tasks for the mobile chassis.

[0139] First, the firefighting mission planner needs to establish the coordinate transformation relationship between the fire line, the pan-tilt unit, the fire monitor, and the vehicle body. Let the map coordinate system be... The vehicle coordinate system is The coordinate system of the binocular camera or sensing gimbal is The coordinate system of the fire monitor is Dual RTK provides the pose transformation matrix of the mobile platform in the map coordinate system, denoted as:

[0140]

[0141] The mounting relationship between the binocular camera or sensing gimbal and the vehicle body is denoted as:

[0142]

[0143] The installation relationship of the fire monitor relative to the vehicle body is recorded as follows:

[0144]

[0145] If the three-dimensional coordinates of a point on the fire line in the binocular camera coordinate system are:

[0146]

[0147] The position of the fire point in the map coordinate system can then be represented as:

[0148]

[0149] This formula indicates that the task planner can uniformly convert the 3D fire line points identified by the binocular camera into a dynamic fire scene environment map. Subsequently, in order to calculate the control angle required for the fire monitor to aim at this fire line point, it is also necessary to convert the fire line points in the map coordinate system to the fire monitor coordinate system:

[0150]

[0151] set up:

[0152]

[0153] The horizontal rotation angle and elevation angle required for the fire monitor to aim at the fire line point can be expressed as follows:

[0154]

[0155]

[0156] in, Indicates the horizontal rotation angle of the fire monitor. Indicates the elevation angle of the fire monitor. This represents the pitch compensation angle after taking into account the descent of the water or foam jet, the jet pressure, and the jet distance. Through the above coordinate transformation relationship, the task planner can convert the three-dimensional information of the fire line into angle control tasks that the fire monitor can perform.

[0157] After coordinate unification, the firefighting task planner, based on the 3D fireline information and temperature zones in the dynamic fire scene environment map, divides the continuous fireline into several fireline segment sub-tasks. Let the current 3D fireline point set be:

[0158]

[0159] The task planner categorizes fire lines based on their spatial continuity, length, temperature distribution, and effective coverage area of ​​fire monitors:

[0160]

[0161] in, Indicates the first Each fire line segment task mainly contains three types of information: first, the three-dimensional point set of the fire line segment; second, the temperature region corresponding to the fire line segment; and third, the direction of movement of the fire line segment. In this way, continuous fire lines are transformed into multiple fire extinguishing sub-tasks that can be processed, tracked, and fed back sequentially.

[0162] For each fire line segment, the firefighting mission planner queries the dynamic fire scene environment map for terrain and navigability information near the fire line segment to determine which side the vehicle should follow. If a side has high-temperature areas, obstacles, areas with excessive slopes, or impassable areas, that side is unsuitable as a chassis following area; if the other side has lower navigability and the distance to the fire line meets the safety distance and effective range requirements of the fire monitor, then that side can be used as the fire line following side. Therefore, the mission planner does not simply guide the vehicle closer to the fire line, but rather selects a safe fire line following path suitable for low-speed driving and continuous spraying under the constraints of the dynamic fire scene environment map.

[0163] When generating a fire monitor mission, the mission planner selects several discrete spray target points on the fire line segment based on the current three-dimensional point set and temperature region:

[0164]

[0165] in, Indicates the first The first of the fire segments There are several spray target points. Based on the aforementioned coordinate transformation relationship, the task planner converts each spray target point into the corresponding horizontal rotation and elevation angles of the fire monitor, forming a discrete angle control task for the fire monitor:

[0166]

[0167] The discrete angle task is sent to the fire monitor's end control module. The fire monitor control module then uses these discrete angle points for trajectory interpolation, speed limiting, and smoothing control to generate the actual continuous motion trajectory of the fire monitor. The task planner is primarily responsible for telling the fire monitor which points to spray and in what order, while the fire monitor control module is responsible for ensuring the nozzle rotates smoothly.

[0168] Simultaneously, the fire extinguishing task planner needs to be coupled with the low-speed moving chassis for planning. Since the system's fire extinguishing method involves moving along the fire line and continuously spraying, the chassis speed cannot be determined solely based on path length or obstacle avoidance; it also needs to be dynamically adjusted according to the fire extinguishing effect of the current fire segment. The system binds each fire segment sub-task to a corresponding image region and infrared temperature region, continuously reading the visible light flame changes and infrared temperature changes of that fire segment during execution.

[0169] The remaining fire intensity of the current fire segment can be defined as follows:

[0170]

[0171] in, Indicates the first The area of ​​the visible flame region of each fire segment at the current moment. This indicates the area of ​​the flame zone when the fire line segment begins to be suppressed; This indicates the average infrared temperature of the current fire segment. This indicates the average infrared temperature at the start of pressing; Indicates the safe temperature threshold; and These are the weighting coefficients. The larger the value, the stronger the remaining fire in that fire segment, and the less effective the suppression. The smaller the value, the more significant the decrease in both the flame and temperature of that fire segment.

[0172] Based on the remaining fire intensity, vehicle attitude error, and following distance error, the task planner generates the chassis travel speed:

[0173]

[0174] in, Indicates the current chassis speed. This indicates the permitted higher driving speed. Indicates the minimum following speed when the jet is executed; This indicates the deviation between the vehicle's heading and the desired following direction. This indicates the deviation between the vehicle and the expected safe distance from the fire line. and This is the weighting coefficient. The formula indicates that when the fire line segment has not yet been effectively suppressed... The larger size and reduced chassis speed allow the fire monitor more time to cover the current fire line; when the fire line suppression effect is good, If the speed is relatively low, the chassis speed can be increased, and the vehicle can continue moving towards the next section of the firing line. However, if the vehicle's posture deviates significantly or the safe distance from the firing line deviates considerably, the vehicle should appropriately reduce its speed to ensure following stability and injection accuracy.

[0175] When a certain fire segment meets the suppression completion condition, for example:

[0176]

[0177] in, To suppress the fire, the task planner marks the fire segment as complete and increases the chassis speed, allowing the vehicle to quickly move to the following position for the next fire segment. If the current fire segment has not yet met the suppression completion criteria, the task planner reduces the chassis speed, maintains or redistributes the fire monitor's spray target point, and allows the fire monitor to continue covering and suppressing the fire segment. If the fire line moves, spreads, or reignites during this process, the task planner recalculates the fire segment, spray target point, and vehicle speed based on the updated dynamic fire scene map, achieving dynamic tracking of the fire line.

[0178] Therefore, in this stage, the input to the fire extinguishing task planner is a dynamic fire scene environment map containing information on the fire line, temperature, terrain, obstacles, traversable areas, and vehicle position. The output mainly includes two types of control tasks: one is the discrete spray angle control task sent to the fire monitor end control module, namely several sets of fire monitor horizontal rotation and elevation angles; the other is the vehicle speed and following path task sent to the chassis motion control module. In this way, the task planner achieves coupled planning between the fire line sub-tasks, the fire monitor spray target, and the low-speed chassis motion, enabling the fire extinguishing platform to dynamically adjust its speed based on fire line suppression feedback and continuously complete follow-up fire extinguishing along the fire line.

[0179] In the fifth stage, the system performs end control of the fire monitor, judges the suppression effect, and updates the task.

[0180] The fire monitor task output by the fire extinguishing task planner is a set of discrete spray target angle points. After receiving these discrete angle points, the fire monitor end control module does not directly control the fire monitor to jump point by point, but performs trajectory smoothing based on the relationship between adjacent angle points.

[0181] Set the current fire segment The corresponding fire monitor discrete angle task is:

[0182]

[0183] in, Indicates the first The horizontal rotation angle of the fire monitor corresponding to each spray target point This represents the corresponding pitch angle. After receiving these discrete angle points, the fire monitor's end control module does not directly control the fire monitor to jump point by point. Instead, it performs trajectory smoothing based on the relationship between adjacent angle points to generate a continuous desired control trajectory for the fire monitor.

[0184] For example, for two adjacent angle points and Intermediate transition angles can be generated using linear interpolation or spline interpolation.

[0185]

[0186]

[0187] in, , and These represent the smoothed desired horizontal angle and desired pitch angle, respectively. In actual control, the angular velocity and angular acceleration of the fire monitor also need to be limited.

[0188]

[0189]

[0190] This prevents the fire monitor from suddenly swinging between adjacent spray points, ensuring a continuous and stable spraying process.

[0191] During the operation of the fire monitor, the control module performs closed-loop control based on the deviation between the current actual angle and the desired angle. Let the current actual horizontal angle and elevation angle of the fire monitor be... , The expected angles are respectively , Then the angle error is:

[0192]

[0193]

[0194] The fire monitor control module outputs horizontal rotation and elevation control commands based on the aforementioned errors, causing the monitor muzzle to gradually track the desired angle trajectory. Since the mobile chassis remains in a low-speed driving state during firefighting, the vehicle's position and attitude continuously change. Therefore, the fire monitor's angle is constantly adjusted based on the vehicle's real-time position and the target point on the fire line to ensure the muzzle always covers the current fire segment.

[0195] While the fire monitors are spraying water, the system continuously monitors the suppression effect of the current fire line segment using a binocular camera and an infrared thermal imaging camera. Visible light images are mainly used to determine whether the flame area has shrunk, whether the flame brightness has decreased, and whether the fire line boundary has broken; infrared thermal imaging images are mainly used to determine whether the temperature of the fire line area has decreased, whether the high-temperature area has shrunk, and whether there is still residual high temperature or risk of reignition.

[0196] For the current fire line segment The system can record the visible flame area at the start of suppression. and average infrared temperature And calculate the current flame area in real time during the fire extinguishing process. and current average infrared temperature If the following conditions are met:

[0197]

[0198] and

[0199]

[0200] This indicates that the open flame area of ​​the fire segment has significantly decreased, and the temperature has dropped below the set threshold, meaning that the fire segment can be determined to have been successfully suppressed. Indicates the flame area attenuation threshold. This indicates the temperature safety threshold.

[0201] To avoid misjudgments in a single frame, the system requires that the above conditions be met for several consecutive frames (e.g., 5 consecutive frames within 3 seconds) before marking the fire segment as "suppressed." Once a fire segment is determined to be suppressed, the fire suppression task planner removes it from the current task queue or marks it as completed, while simultaneously updating the fire line status and high-temperature zone status in the dynamic fire scene environment map. Subsequently, the system selects the next fire segment as the new fire suppression target and sends a new discrete angle control task to the fire monitor terminal control module, while simultaneously sending a new driving speed and following path task to the chassis motion control module.

[0202] If the current fire segment has not met the conditions for suppression completion, such as a still large visible flame area, an infrared temperature still above the threshold, or an expansion of the high-temperature area, the system considers the fire segment not to have been sufficiently suppressed. In this case, the fire suppression task planner reduces the chassis speed, allowing the mobile platform to remain near the fire segment for a longer period; simultaneously, it reselects or adds fire monitor spray target points, ensuring the fire monitors continue to cover the high-temperature area and leading edge of the fire segment.

[0203] If, during the suppression process, the fire line is detected to have moved, changed direction, reignited, or a new high-temperature zone appears, the system updates the three-dimensional information of the fire line, the temperature zone, and the direction of fire line movement, and overlays the updated information onto the dynamic fire scene environment map. The fire suppression task planner then regenerates the fire line segment tasks, fire monitor discrete angle tasks, and chassis travel speed based on the new dynamic fire scene environment map.

[0204] Thus, the system has formed a complete closed-loop control process, from fire line detection, map building, task planning, chassis following, fire monitor spraying, to effect feedback. Through the above specific implementation methods, the mobile fire extinguishing platform can autonomously follow the fire line of forest and grassland fires and dynamically adjust its travel speed and spray targets according to the fire line suppression effect, achieving continuous and dynamic suppression of continuous fire lines.

[0205] All parts not covered in this application are the same as or can be implemented using existing technology. Although embodiments of this application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for automatically extinguishing forest and grassland fires by following the fire line and moving automatically, characterized in that, Includes the following steps: Step S1: Obtain fire scene perception information through the multi-source sensing unit mounted on the mobile fire extinguishing platform, segment and identify the fire line area based on the visible light image in the perception information, combine infrared thermal imaging information to verify authenticity and identify high temperature areas, and extract the fire line movement direction through continuous inter-frame motion analysis to form fire line status information including three-dimensional fire line information, fire line temperature area and fire line movement direction. Step S2: Construct a priori environment map using lidar and dual RTK positioning and orientation modules. The priori environment map includes terrain structure, obstacle distribution, and passable area information. Overlay the fire line status information and high-temperature area obtained in step S1 onto the priori environment map, and introduce safety distance constraints and comprehensive passage costs to generate a dynamic fire scene environment map. Step S3: The fire extinguishing task planner divides the continuous fire line into one or more fire line segments based on the dynamic fire scene environment map, determines the safe following path of the fire line and the target of the fire monitor spray, and through the coordinated control of the chassis and the fire monitor, enables the mobile fire extinguishing platform to move autonomously along the safe following path, while controlling the fire monitor to continuously spray and suppress the current fire line segment. Step S4: During the spray suppression process, the suppression effect of the current fire segment is judged in real time based on the changes in visible light flame and infrared temperature. If the suppression completion conditions are met, the process switches to the next fire segment sub-task. If the conditions are not met or the fire status changes, the fire status information, dynamic fire scene environment map, and fire extinguishing task are dynamically updated to achieve closed-loop control. In step S2, the method for constructing the dynamic fire scene environment map specifically includes: Let the prior environment map be ,in, Indicates the first in the map A grid cell is represented as: ,in, Indicates the spatial position of the grid cell. This indicates information about slope or terrain undulation. Indicates the state of the obstacle. This indicates the basic accessibility of the grid; Let the current fire status be... ,in The three-dimensional spatial location of the fire line. The direction of fire line movement. Temperature distribution; For any grid Calculate its three-dimensional spatial position relative to the fire line. distance ,definition ， in The distance affected by the fire line, among which, This indicates that the grid belongs to the fire zone or the fire-affected zone. This indicates that the grid does not belong to the fire-affected area; Let the first The temperature value corresponding to each grid cell is The high temperature threshold is The high-temperature region is represented as: ， in, This indicates that the grid is a high-temperature region. This indicates that the grid has not met the high temperature judgment condition; Build a grid Comprehensive traffic costs , ， in, Indicates the cost of traversing the terrain. Indicates the cost of obstacles. This indicates the costs associated with maintaining a safe distance from the line of fire. , These are the weighting coefficients for each cost item. This indicates the desired safe following distance between the mobile platform and the fireline; Candidate grids also need to meet the following requirements ,in For minimum safe distance, This refers to the effective spray distance of the fire monitor. Finally, a dynamic fire scene environment map is generated. , In step S4, the method for determining the suppression completion condition includes: Record the area of ​​the flame zone when the current fire line segment begins to suppress. and average infrared temperature ; During firefighting, the visible flame area is calculated in real time at the current moment. and the average infrared temperature of the fire segment at the current moment. If both conditions are met: and , Then it is determined that the fire line segment has been suppressed, where The flame area attenuation threshold. The temperature safety threshold; Once a fire line segment is determined to be suppressed, the fire suppression task planner removes the fire line segment from the current task queue or marks it as completed, while updating the fire line status and high-temperature zone status in the dynamic fire scene environment map.

2. The forest and grassland fire-fighting method with fire-following movement according to claim 1, characterized in that, Step S1 specifically includes: The YOLOv5 image segmentation model was used to segment the flame region in the visible light image acquired by the binocular camera to obtain the fire line segmentation mask; The displacement changes of the fire line segmentation mask in adjacent frames are analyzed using the frame difference method. Combined with the spatial information obtained from the binocular camera, the three-dimensional spatial position of the fire line is determined. and direction of movement ; The high-temperature area in the infrared thermal imaging image is matched with the visible light segmentation result. When the visible light flame area and the infrared high-temperature area overlap, it is determined to be a real fire line. The infrared temperature distribution is used to distinguish between the burned side and the unburned side of the fire line. Through the above steps, fire line status information including three-dimensional fire line information, fire line temperature range, and fire line movement direction is generated.

3. The forest and grassland fire-fighting method with fire-following movement according to claim 1, characterized in that, In step S2, the method for constructing the prior environment map includes: Point cloud information of the surrounding environment of the mobile platform is collected by lidar, and then filtered, segmented, and extracted to obtain basic terrain structure information. The real-time position and vehicle heading of the mobile platform are obtained through dual RTK positioning and orientation modules, and the local point cloud data collected by the lidar is converted into a unified map coordinate system. Based on changes in ground elevation, slope, and obstacle distribution, the environment is divided into passable areas, potentially passable areas, and impassable areas, generating a navigation map in the form of a cost map. Areas with obstacles, ditches, and excessive slopes are assigned higher passable costs, while areas suitable for chassis passage are assigned lower passable costs.

4. The forest and grassland fire-fighting method with fire-following movement according to claim 1, characterized in that, In step S3, the specific method for dividing the continuous fire line into fire line segment sub-tasks includes: Let the current three-dimensional point set of the fire line be... Based on the continuity of the fire line space, the length of the fire line, the temperature distribution, and the effective spray coverage of the fire monitor, Divided into Each of the following is a separate, unrelated sentence: "In each of the following sub-missions..." This includes the three-dimensional point set of the fire line segment, the temperature region corresponding to the fire line segment, and the direction of movement; The fire extinguishing task planner queries the terrain and accessibility information near each fire line segment in the dynamic fire scene environment map, and selects the side with lower access cost and a distance from the fire line that meets the requirements of safe distance and effective range of fire monitors as the fire line following side.

5. The forest and grassland fire-fighting method with fire-following movement according to claim 1, characterized in that, In step S3, the coordinated control of the chassis and fire monitor includes a dynamic adjustment method for the chassis driving speed: defining the remaining fire intensity of the current fire segment. : in, Indicates the first The area of ​​the visible flame region of each fire segment at the current moment. This indicates the area of ​​the flame zone when the fire line segment begins to be suppressed. This indicates the average infrared temperature of the current fire segment. This indicates the average infrared temperature at the start of pressing. Indicates the safe temperature threshold. and These are the weighting coefficients; Based on the remaining fire intensity and the vehicle's heading deviation and safety distance deviation Generate chassis driving speed , ， in, This indicates the permitted higher driving speed. Indicates the minimum following speed when the jet is executed. and These are the weighting coefficients; when The current fire segment suppression is completed at the time of determination. To suppress the completion threshold, the task planner marks the fire segment as completed and increases the chassis speed.

6. A forest and grassland fire-following mobile automatic fire extinguishing system, used to implement the method according to any one of claims 1 to 5, characterized in that, include: The mobile chassis is used to carry the various components of the system and drive the system to move autonomously and safely along the fire line. Fire monitors, mounted on mobile chassis, are capable of horizontal rotation and pitch adjustment, and are used to spray extinguishing media onto fire-prone areas. The multi-source sensing unit includes a binocular camera, an infrared thermal imaging camera, a lidar, and a gimbal. The binocular camera is used to acquire visible light images of the fire line, the infrared thermal imaging camera is used to acquire the temperature distribution of the fire line and the surrounding area, the lidar is used to acquire terrain structure and obstacle information, and the gimbal is used to support and adjust the observation direction of the binocular camera and the infrared thermal imaging camera so that they are continuously facing the fire line area. The positioning and orientation unit uses a dual RTK module, with two RTK antennas arranged at different positions on the top of the vehicle body to obtain high-precision position and heading information of the mobile platform. The control computing unit is electrically connected to the mobile chassis, fire monitor, multi-source sensing unit, and positioning and orientation unit, respectively. The control computing unit includes: The fire status identification and modeling module is used to execute step S1; The dynamic fire scene environment map construction module is used to execute step S2; Firefighting task planner, used to execute task planning in step S3; The chassis motion control module is used to execute the chassis motion control in step S3. The fire monitor end control module is used to execute the fire monitor spray control in step S3; The feedback update module is used to execute step S4.

Images