Forest fire extinguishing system based on multiple groups of unmanned aerial vehicles

Abstract

The invention relates to a forest fire extinguishing system based on multi-group unmanned aerial vehicles, and the system comprises a plurality of unmanned aerial vehicle groups which comprise a searching unmanned aerial vehicle group, a fire extinguishing unmanned aerial vehicle group and carrying temperature sensing equipment; the central server is in communication connection with the unmanned aerial vehicle group, and the central server is configured to determine the quantitative proportion of the searching unmanned aerial vehicles and the fire extinguishing unmanned aerial vehicles by taking the minimum total response time as a distribution optimization function based on the estimated fire area and the estimated response time; dividing a target search area, driving the search unmanned aerial vehicle group to perform a search task of the target search area, when the search unmanned aerial vehicle detects a fire point, predicting a fire area, determining the number of required fire extinguishing unmanned aerial vehicles based on the predicted fire area, and driving the fire extinguishing unmanned aerial vehicle group to fly to the target fire point area to perform fire extinguishing operation. According to the invention, the scheduling efficiency of fire extinguishing resources and the automation degree of system operation are effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of emergency rescue, in particular to a forest fire extinguishing system based on multiple groups of unmanned aerial vehicles. BACKGROUND

[0002] The forest fire extinguishing scene refers to a series of extinguishing actions taken in a forest, grassland or other vegetation dense area in response to an existing fire. The core goal is to control the spread of fire, extinguish open fires and prevent rekindling, while ensuring personnel safety.

[0003] Traditional forest fire extinguishing mainly relies on manual patrol, satellite remote sensing, ground fire fighting forces and a small amount of manned aircraft support. Traditional monitoring relies on satellites (long revisit period, greatly affected by clouds and fog) or manual lookout towers (limited coverage, easily blocked by terrain), which is difficult to discover hidden fire sources (such as smoldering fires) or small initial fire points in real time. Since forest fires often occur in complex terrain such as mountains and valleys, ground fire vehicles are difficult to reach, manual extinguishing is low in efficiency, and firefighters have to face high temperature, thick smoke (containing CO, PM2.5 and other toxic gases), flying fire and other threats, with a high risk of injury and death. If large helicopters are used for extinguishing, large helicopters are limited by load (about 12 tons of water per single spraying), flight height (need to avoid tree crowns) and fuel restrictions (short single operation time), and cannot be covered accurately and frequently. SUMMARY

[0004] The embodiment of the present application provides a forest fire extinguishing system based on multiple groups of unmanned aerial vehicles, comprising: A plurality of unmanned aerial vehicle groups carrying temperature sensing devices; A central server in communication connection with the unmanned aerial vehicle groups, the central server being configured to include: A task allocation module, configured to include a search unmanned aerial vehicle group and a fire extinguishing unmanned aerial vehicle group, based on an estimated fire area , an estimated response time , a distribution optimization function is determined to minimize the total response time to determine the number ratio of search unmanned aerial vehicles and fire extinguishing unmanned aerial vehicles; A task execution module, which divides a target search area and drives the search unmanned aerial vehicle group to perform a search task in the target search area When the search unmanned aerial vehicle detects a fire point , predicts the fire area and determines the number of fire extinguishing unmanned aerial vehicles required based on the predicted fire area, and drives the determined number of fire extinguishing unmanned aerial vehicle groups to fly to the target fire point area to perform fire extinguishing operation, wherein the target search area is divided based on the target forest area and the number of search unmanned aerial vehicle groups ​The task monitoring module detects the fire area change rate in real time, and adjusts the number ratio of the search unmanned aerial vehicle and the fire extinguishing unmanned aerial vehicle according to the ratio of the fire area change rate to an area change threshold.

[0005] In some embodiments, the task monitoring module is configured to: monitor the perceived temperature of the target fire point in real time based on the temperature sensing device If the perceived temperature is lower than a set temperature threshold , it is determined that the fire extinguishing task of the target fire point is completed, and the unmanned aerial vehicle performing the fire extinguishing task of the target fire point is released back to the resource pool.

[0006] In this way, the scheduling efficiency of fire extinguishing resources and the automation degree of system operation are effectively improved. By introducing the temperature threshold as the judgment standard of fire extinguishing completion, the delay and misjudgment caused by manual confirmation can be avoided, and the energy waste caused by redundant retention of unmanned aerial vehicles is reduced. After the unmanned aerial vehicle is automatically released back to the resource pool, it can be re-assigned to a newly discovered fire point area, significantly improving the overall response frequency and resource utilization of the system.

[0007] In some embodiments, the driving the search unmanned aerial vehicle group to perform the target search area further includes: generating a flight waypoint of the search unmanned aerial vehicle based on a Levy flight strategy, performing path search according to an A* algorithm to obtain a flight path with the lowest total cost.

[0008] If the next waypoint position of the search unmanned aerial vehicle group is , and the current position of the maximum information unmanned aerial vehicle in the search unmanned aerial vehicle group is , , then:

[0009] wherein, is a search step scaling factor, is a random step length from a Levy distribution, is a random direction angle, , is a uniform distribution sampling, is a direction angle of the maximum information unmanned aerial vehicle , and is a direction deviation range. The above , is a direction scaling constant, is an exponential sensitivity constant. To perceive temperature.

[0010] In the above embodiments, the path total cost is calculated based on the following calculation model: , is the actual path cost from the starting point to the node , is the heuristic estimated cost from the node to the target node, for example but not limited to, calculated by Euclidean distance.

[0011] Based on the above configuration, the application is based on the above path strategy to make the search unmanned aerial vehicle have the ability of self-adaptive adjustment of direction and task focus under the conditions of complex geography and heat distribution such as dynamic fire scene and high temperature dense area, so as to realize more effective target positioning, fire source identification and pre-scouting deployment of fire extinguishing task.

[0012] In some embodiments, the number ratio of search unmanned aerial vehicles and fire extinguishing unmanned aerial vehicles is calculated based on the following calculation model:

[0013] wherein, is the number of search unmanned aerial vehicles, is the total number of unmanned aerial vehicles in the unmanned aerial vehicle group, , is the allocation optimization function, whose input is the estimated fire area , , , is a preset threshold value, can be iterated by genetic algorithm to minimize the total response time , and a scheduling priority coefficient is output by balancing the normalized indexes of area and time to adjust the ratio of search unmanned aerial vehicles and fire extinguishing unmanned aerial vehicles, and the total response time is expressed as: , is the fire discovery time, is the fire extinguishing time, that is, the maximum time delay between fire triggering and effective disposal.

[0014] Based on the above configuration, the system is designed to balance resources and improve the response efficiency of multiple point fires; the maximum value of the fire discovery time and the fire extinguishing time is taken as the total operation time of the system to ensure the parallel and synchronous of search task and fire extinguishing task.

[0015] In some embodiments, the predicted fire area specifically includes: a time field defining a fire boundary, predicting an evolution of the fire boundary based on an Eikonal equation simulating the time field ; wherein the evolution of the fire boundary is represented by is a time field function representing a time of arrival, is a gradient modulus of the time field function at a location .

[0016] a predicted fire area is calculated based on the time field of the fire boundary and a fire spread rate:

[0017] wherein, is an integral along the current fire spread fire boundary, is a boundary infinitesimal length, is a fire spread rate.

[0018] In some embodiments, the fire spread rate

[0019] wherein, is an initial spread rate; is a fuel correction factor; is a terrain correction factor; is a wind correction factor; is a wind direction angle,

[0020] In some embodiments, the determination of the number of required fire-extinguishing drones based on the predicted fire area can be obtained based on the following calculation model:

[0021] wherein, is a fire-extinguishing efficiency,

[0022] In some embodiments, after determining the number of required fire-extinguishing drones based on the predicted fire area, the drones are assigned to perform the fire-extinguishing task: the sector angle is divided equally based on the number of drones of the fire-extinguishing drone group , the coverage area of a single fire-extinguishing bomb is calculated​​​​​​​​​ , estimate the number of fire extinguishing bombs required for each sector based on the coverage area of a single fire extinguishing bomb , for reflecting the number of fire extinguishing bombs required for each unmanned aerial vehicle sector task wherein, , is the explosion radius of the fire extinguishing bomb, is the efficiency of the fire extinguishing bomb ; According to the number of fire extinguishing bombs carried by a single fire extinguishing unmanned aerial vehicle Calculate the total fire extinguishing bomb capacity of the fire extinguishing unmanned aerial vehicle group , based on the ratio of the total fire extinguishing bomb capacity and the number of fire extinguishing bombs required for each sector ; If , then expand the number of fire extinguishing unmanned aerial vehicles, otherwise, the current ammunition resources meet the task deployment requirements, wherein, , .

[0023] In some embodiments, the fire area change rate is represented as: , wherein, is the perimeter of the fire point , is the current area of the fire point , the shape of the fire point is approximately circular. In some embodiments, the number ratio of search unmanned aerial vehicles and fire extinguishing unmanned aerial vehicles is adjusted, specifically:

[0024] , , is the adjustment step size, is the area change threshold value.

[0025] Based on the above embodiments, the system adjusts the number ratio of search unmanned aerial vehicles and fire extinguishing unmanned aerial vehicles dynamically by monitoring the change rate of the fire area over time. If the fire area change rate is greater than 0, it indicates that the fire is expanding, and the system increases the number of search unmanned aerial vehicles to expand the reconnaissance area and discover more new fire points. Conversely, when the area change rate tends to be stable or negative, the system reduces the number of search unmanned aerial vehicles and increases the number of fire extinguishing unmanned aerial vehicles to strengthen the suppression capability.

[0026] The above implementation realizes a task resource adaptive control strategy based on dynamic changes of the fire, and combines a real-time monitoring and feedback control mechanism for fire points, so that the system has stronger scheduling flexibility and response efficiency when facing complex, multi-point, and multi-stage fire evolution.​

[0027] The details of one or more embodiments of the application are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the application will be apparent from the description and drawings, and from the claims. BRIEF DESCRIPTION OF DRAWINGS

[0028] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings: Figure 1 is a structural diagram of a forest fire extinguishing system according to an embodiment of the application; Figure 2 is another structural diagram of a forest fire extinguishing system according to an embodiment of the application; Figure 3 is another structural diagram of a forest fire extinguishing system according to an embodiment of the application; Figure 4 is a flowchart of a forest fire extinguishing system according to an embodiment of the application.

[0029] In the drawings: 1, central server; 2, unmanned aerial vehicle group; 101, task allocation module; 102, task execution module; 103, task monitoring module. DETAILED DESCRIPTION

[0030] In order to make the objects, technical solutions and advantages of the application clearer, the application is described and explained below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the application and should not be used to limit the application. Based on the embodiments provided in the application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of the application.

[0031] Obviously, the drawings in the following description are only some examples or embodiments of the application, and for those of ordinary skill in the art, the application can be applied to other similar scenarios without creative labor on the basis of these drawings. In addition, it can be understood that although the efforts made in this development process can be complex and lengthy, for those of ordinary skill in the art related to the content disclosed in the application, some design, manufacture or production changes based on the technical content disclosed in the application are only routine technical means and should not be understood as insufficient disclosure of the content disclosed in the application.

[0032] References to "embodiments" in this application mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it refer to independent or alternative embodiments that are mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described in this application may be combined with other embodiments unless there is a conflict.

[0033] Unless otherwise defined, technical or scientific terms used herein shall have the ordinary meaning as understood by persons of ordinary skill in the art to which this application belongs. The terms "a," "an," "an," "the," and similar expressions used herein do not denote quantitative limitations and may refer to either the singular or the plural. The terms "comprise," "include," "have," and any variations thereof, used herein, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or modules (units) is not limited to the listed steps or units but may also include steps or units not listed, or may include other steps or units inherent to the process, method, product, or apparatus. The terms "connected," "connected," "coupled," and similar expressions used herein are not limited to physical or mechanical connections but may include electrical connections, whether direct or indirect. As used herein, "plurality" means two or more. "And / or" describes an association between associated objects, indicating that three possible relationships exist. For example, "A and / or B" may mean: A exists alone; A and B exist simultaneously; or B exists alone. The character " / " generally indicates that the objects before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific order for the objects.

[0034] Multi-swarm UAVs refers to a distributed system composed of multiple drone swarms (Swarm). Within each swarm, a large number of small drones complete local tasks through autonomous collaboration, while different swarms achieve global goals through higher-level collaboration.

[0035] This embodiment provides a forest fire extinguishing system based on a swarm of drones. The system is used to implement the above-mentioned embodiments and preferred embodiments. Details that have already been described will not be repeated. As used below, the terms "module," "unit," "subunit," etc. may refer to a combination of software and / or hardware that implements a predetermined function. Although the systems described in the following embodiments are preferably implemented in software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0036] Figures 1-3 is a structural diagram of a forest fire extinguishing system according to an embodiment of the present application, Figure 4 is a flowchart of a forest fire extinguishing system according to an embodiment of the present application, as shown in the figure, the system comprises: Figures 1 to 4 a plurality of unmanned aerial vehicle groups 2, equipped with temperature sensing devices, such as infrared thermal imaging devices; a central server 1, which is communicatively connected to the unmanned aerial vehicle groups, the central server 1 is configured to include: a task allocation module 101, which configures the unmanned aerial vehicle groups 2 to include search unmanned aerial vehicle groups and fire extinguishing unmanned aerial vehicle groups, determines the number of search unmanned aerial vehicles and fire extinguishing unmanned aerial vehicles based on the estimated fire area , the estimated response time , and a distribution optimization function that minimizes the total response time; a task execution module 102, which divides the target search area and drives the search unmanned aerial vehicle groups to perform the search task (step S101) in the target search area , when the search unmanned aerial vehicle detects a fire point , predicts the fire area and determines the number of fire extinguishing unmanned aerial vehicles based on the predicted fire area, and drives the determined number of fire extinguishing unmanned aerial vehicle groups to fly to the target fire point area to perform fire extinguishing operations (step 102), wherein the target search area is divided based on the target forest area and the number of search unmanned aerial vehicle groups , i.e. , to avoid search overlap and improve coverage efficiency; a task monitoring module 103, which detects the fire area change rate in real time and adjusts the number ratio of the search unmanned aerial vehicles and the fire extinguishing unmanned aerial vehicles according to the ratio of the fire area change rate to an area change threshold value.

[0037] In the above embodiment, the search unmanned aerial vehicle can be replaced by a small unmanned aerial vehicle with infrared imaging function or an unmanned aerial vehicle equipped with LIDAR sensor to adapt to night or poor visibility environment; the fire extinguishing unmanned aerial vehicle can adopt a multi-rotor high-load unmanned aerial vehicle equipped with water bag, dry powder extinguisher or aerosol spraying device and other fire extinguishing devices to meet the needs of different fire conditions, preferably, the fire extinguishing unmanned aerial vehicle of the present application is equipped with fire extinguishing bomb; the central server can be deployed on a ground command vehicle or integrated into a mobile communication base station for remote deployment; the predicted fire area can be realized by a deep learning fire condition prediction model; the target search area can be divided by an adaptive grid division algorithm, such as a dynamic search division method based on Voronoi diagram; the dynamic adjustment parameters of the search and fire extinguishing ratio can be personalized set according to historical fire data, terrain complexity or meteorological factors to enhance the robustness and universality of the system.

[0038] ​Based on the above embodiments, the present application significantly improves the flexibility and response efficiency of unmanned aerial vehicle scheduling. By adjusting the division of search and fire extinguishing tasks on demand, resource waste is reduced; the fire area change rate feedback mechanism further enhances the response ability of the system to sudden fire, and has strong environmental adaptability and emergency control ability.

[0039] In some embodiments, the task monitoring module 103 is configured to further include: monitoring the perceived temperature of the target fire point in real time based on the temperature perception device, and determining whether the perceived temperature is lower than a set temperature threshold (step S103), if yes, determining that the fire extinguishing task of the target fire point is completed, and releasing the unmanned aerial vehicle performing the fire extinguishing task of the target fire point back to the resource pool, and optionally setting the temperature threshold may be 50°C. The resource pool is a set of state flags and parameters maintained in the central server, used to record the task state, position, circuit, etc. of each unmanned aerial vehicle. When the unmanned aerial vehicle completes the fire extinguishing or search task, it is determined whether it can be reused (such as sufficient power and not damaged), and it is released back to the resource pool. When a new fire point is identified or an existing task needs reinforcement, an unmanned aerial vehicle in the resource pool is selected for task allocation until all fire point tasks are completed.

[0040] This embodiment monitors the fire extinguishing process through a temperature perception mechanism, and uses the infrared thermal imaging device carried by the fire extinguishing unmanned aerial vehicle or the search unmanned aerial vehicle to obtain the real-time perceived temperature of the target fire point. The task monitoring module 103 continuously compares the current fire point temperature with the set temperature threshold. Once the detection value is lower than the threshold, it is considered that the fire point has been effectively extinguished. Subsequently, the system automatically updates the state of the fire point, and releases all unmanned aerial vehicles participating in the fire extinguishing task of the target fire point back to the resource pool for subsequent new fire extinguishing task scheduling. This monitoring mechanism can be embedded into the task execution module 102 to realize closed-loop integrated control with functions such as fire point prediction and fire extinguishing unmanned aerial vehicle allocation.

[0041] In the above manner, the scheduling efficiency of fire extinguishing resources and the automation level of system operation are effectively improved. By introducing the temperature threshold as the judgment standard for fire extinguishing completion, the delay and misjudgment caused by manual confirmation can be avoided, and the energy waste caused by the redundant retention of unmanned aerial vehicles is reduced. After the unmanned aerial vehicle is automatically released back to the resource pool, it can be re-assigned to a newly discovered fire point area, significantly improving the overall response frequency and resource utilization of the system.

[0042] In the above embodiments, the monitoring method of the perceived temperature can be replaced according to the environment and requirements, such as using a thermistor array, a non-cooled infrared detector, or an AI vision algorithm based on image temperature fitting; the temperature threshold can also be dynamically set or adaptively adjusted according to the season, terrain, vegetation type, etc.

[0043] In the above embodiments, in addition to directly returning the resource pool, the release strategy can also set a buffer mechanism to reactivate the corresponding unmanned aerial vehicle to perform a secondary fire extinguishing task when the system detects a fire point rekindling trend; Specifically, when the system determines that a certain target fire point reaches a set temperature threshold, the unmanned aerial vehicle performing the task is not immediately released back to the schedulable resource pool, but is transferred to a temporary buffer state.

[0044] The unmanned aerial vehicle in this state remains in standby mode, such as hovering in the nearby airspace or returning to the nearest residence point, waiting for the system's determination result of the fire point rekindling risk within a certain time window. If the system determines that the rekindling probability exceeds the set threshold (such as 80%), the corresponding unmanned aerial vehicle in the buffer is reactivated by the task scheduling module, and immediately returns to the fire point area to implement a secondary fire extinguishing operation. Otherwise, after the buffer time window ends (such as 5 minutes), it is automatically released to the resource pool.

[0045] In addition, this mechanism can also be extended to the application scenario of monitoring multiple fire points concurrently, realizing coordinated control of multi-point collaborative judgment and resource backflow scheduling. The system can set corresponding buffer task queues for multiple fire points at the same time, and dynamically prioritize and reallocate control of unmanned aerial vehicles in the buffer area according to fire point priority, resource tightness, and regional heat risk distribution, thereby realizing collaborative judgment and resource backflow scheduling optimization among multiple fire points.

[0046] In some embodiments, the driving the search unmanned aerial vehicle group to perform a target search area searching task further includes: generating a flight waypoint of the search unmanned aerial vehicle based on a Levy flight strategy, performing path search according to an A* algorithm to obtain a flight path with the lowest total cost .

[0047] Suppose the next waypoint position of the search unmanned aerial vehicle group is , and the current position of the maximum information unmanned aerial vehicle in the search unmanned aerial vehicle group is , , then:

[0048] wherein is a search step scaling factor, is a random step length from a Levy distribution, is a random direction angle, , is a uniform distribution sampling, is a direction angle of the maximum information unmanned aerial vehicle . a direction deviation range; The above , a direction scaling constant, an exponential sensitivity constant, a perceived temperature, used to calculate a deviation range based on the perceived temperature using a Sigmoid function, to determine a search direction of the search drone by constraining the random direction angle distribution, to narrow the deviation at high temperatures, to focus on hot spots, to improve search accuracy and responsiveness in high-temperature regions; Based on this, the path estimation algorithm adopts an A* heuristic search strategy to plan a minimum total cost path to the target point in the search area and dynamically adjust the direction deviation angle.

[0049] In the above embodiments, the maximum information drone is the drone that detects the maximum temperature change rate in the search drone group , that is , so as to filter the maximum information drone based on the temperature value change rate perceived by the drone, to update the path in real time, to improve the fire point detection speed and cooperation efficiency.

[0050] By determining the high-temperature target area through the perceived temperature range, a higher direction bias weight is given to the area near the high-temperature point, so as to drive the search drone to gather in the high-temperature area, to realize focused search towards the target heat source.

[0051] In the above embodiments, the path total cost is calculated based on the following calculation model: , is the actual path cost from the starting point to the node , and is the heuristic estimated cost from the node to the target node, for example but not limited to, calculated by Euclidean distance.

[0052] In the above embodiments, the path generation strategy can select a path planning method based on deep reinforcement learning, such as DDPG, PPO algorithm, etc.

[0053] Based on the above configuration, based on the above path strategy, the search drone has the ability to adaptively adjust the direction and focus on the task under complex geographical and thermal distribution conditions such as dynamic fire field and high-temperature dense area, to realize more effective target positioning, fire source identification and pre-scouting deployment of fire extinguishing tasks.

[0054] In some embodiments, the number ratio of search drones and fire extinguishing drones is obtained based on the following calculation model:

[0055] wherein, is the number of search drones, is the total number of drones in the drone swarm, , is the allocation optimization function, with inputs of estimated fire area , estimated response time , , is a preset threshold, can be iterated by genetic algorithm to minimize the total response time , and output a scheduling priority coefficient by balancing the normalized indicators of area and time factors, to adjust the ratio of search drones and fire extinguishing drones, the total response time is expressed as: , is the fire discovery time, is the fire extinguishing time, i.e. the maximum delay between fire triggering and effective disposal, which is obtained by fusing the fire discovery time and the system response time, to ensure that the scheduling strategy can adapt to different urgency of sudden fire, and help the system to automatically adjust the number of search and fire extinguishing drones according to the change of fire, so that more search drones are preferentially configured to improve the efficiency of fire point discovery in the early stage of fire or when the fire point density is low; in the case of fire expansion or large response delay, the system tends to increase the fire extinguishing drone resources to strengthen the fire suppression capacity, thereby realizing dynamic optimization of scheduling resources.

[0056] In the above embodiment, the estimated fire area and the estimated response time are calculated by the central server using satellite scanning data, i.e. through hot spot recognition and boundary extraction of remote sensing images of the fire area, combined with historical terrain data and image sequence change to calculate the estimated fire area , and according to the average response period between fire discovery and completion of drone preparation to estimate the estimated response time , based on the allocation optimization function to increase the search proportion in large area fire to cover more potential fire points and improve the overall fire extinguishing efficiency and the adaptability of the system.

[0057] In addition, the system can also make regionalized setting of and based on different vegetation types, terrain difficulty or wind speed conditions, to enhance the geographical adaptability and practicality of the model.

[0058] In the above embodiment, the total response time By taking the maximum value of the fire discovery time and the fire extinguishing time as the evaluation index of the total operation time, the synchronization constraint of the parallel tasks is ensured, the overall timeliness of the fire extinguishing task and the parallel efficiency of the multi-point fire handling are improved, and the delay of the sequential process is avoided.

[0059] Based on the above configuration, the system balance resources are ensured, and the response efficiency of the multi-point fire is improved; the maximum value of the fire discovery time and the fire extinguishing time is taken as the total operation time of the system to ensure the parallel synchronization of the search task and the fire extinguishing task.

[0060] In the task execution module 102, the UAV routing scheme is generated by using a path generation method, such as the A* algorithm, to plan the task route for each fire extinguishing UAV, and the feasibility of the UAV routing scheme is evaluated based on a preset fitness function through a genetic algorithm to ensure that the response time of each task point meets or is as early as possible as the deadline.

[0061] The fitness function is expressed as: , wherein, is the deadline of the fire point , which is used to represent a time threshold at which the fire extinguishing task must be started to control the fire range and ensure that the fire extinguishing task is performed in time as the fire spreads, is the task start time of the UAV to the fire point .

[0062] By minimizing the sum of the difference between the deadline of all fire points and the start time of the UAV, if the fitness function is positive, it means that all task points can be responded in time, and the system executes the current path scheme for scheduling to ensure that each task is started before the deadline; otherwise, the system iterates the path combination, selects a better scheduling sequence through a genetic optimization algorithm to improve the task completion rate.

[0063] In the above fitness function, the task start time is calculated based on the following constraint conditions:

[0064] , wherein, is the distance from the UAV to the fire point , is the flight speed of the UAV , is the completion time of the UAV to the previous fire point , is the distance between the fire point and the previous fire point , is the task sequence number of the UAV .

[0065] In the above embodiment, at the initial stage of system startup, the central server calculates the fire point position and priority based on the existing satellite data or initial fire point report, and uses a centralized task allocation algorithm (such as the nearest neighbor greedy algorithm or heuristic planning) to assign an initial task list to each fire-extinguishing unmanned aerial vehicle and mark the initial sequence number. The system maintains a task execution queue and a corresponding sequence number list for each unmanned aerial vehicle.

[0066] During the execution of the fire-extinguishing task, the search unmanned aerial vehicle continuously uploads the coordinates of newly identified fire points to the central server in real time. After receiving the new fire point, the central server performs the following operations according to the current state of the fire-extinguishing unmanned aerial vehicle (including its position, power, load, and idle degree): Calculate the flight distance or shortest path time from the new fire point to the current position of each fire-extinguishing unmanned aerial vehicle, evaluate the incremental cost (such as the change in delay response function value) of inserting the new task into the existing task queue, use an insertion reordering mechanism to insert the new fire point as a new task node into the task queue of the most suitable unmanned aerial vehicle, and reassign sequence numbers to all affected task points; if the current scheduling model supports concurrent insertion and local optimization, only update the task sequence of the affected unmanned aerial vehicle to maintain the overall stability of the system.

[0067] In the above fitness function, the deadline Based on the following calculation model:

[0068] wherein, is the critical area threshold of the fire point, and exceeding this threshold indicates that the fire is uncontrollable, is the initial area of the fire point , which is calculated when the search unmanned aerial vehicle first identifies the fire point, is the fire spread rate, , is the fire-extinguishing rate, which represents the speed at which the fire-extinguishing agent in the fire-extinguishing bomb suppresses the fire.

[0069] By setting the deadline for processing each fire point, it is beneficial to prioritize high-risk points, prevent fire spread, and support timely intervention. By setting the critical area to judge the controllability of the fire, the unmanned aerial vehicle allocation is triggered for adjustment.

[0070] In the above embodiments, in the specific path planning process, the central server combines the current positions of the unmanned aerial vehicles, the positions of the task points, the flight speeds, the numbers of the carried fire extinguishing bombs and other parameters to comprehensively schedule, generate a multi-task allocation table, and encode the task order as a chromosome structure. Through mechanisms such as crossover, mutation, fitness evaluation and the like, the optimal path combination is repeatedly evolved and converged. When multiple unmanned aerial vehicles exist path conflicts or task coverage overlaps, the system can introduce a conflict avoidance mechanism or a guided random disturbance to maintain the feasibility of scheduling.

[0071] Based on the above configuration, the feasibility of the unmanned aerial vehicle routing scheme is evaluated by using a fitness function, the task scheduling is ensured to be completed within the deadline, and the reliability of the fire extinguishing business is improved.

[0072] In some of the embodiments, the predicted fire area specifically includes: defining a time field of the fire boundary, simulating evolution of the fire boundary based on an Eikonal equation to predict the time field of the fire boundary ; wherein the evolution of the fire boundary is represented by , , is a time field function, representing the arrival time, is a gradient modulus of the time field function, and the value of the time field at a position is the current time , represented as .

[0073] Based on the time field of the fire boundary and the fire propagation rate, the predicted fire area is calculated as:

[0074] wherein, is an integral along the current fire propagation boundary, is a boundary element length, and the element length can be flexibly configured, is a fire propagation rate, which is used to calculate the growth of the fire area by boundary integration, and is combined with the CA processing to predict the future area for handling the fire propagation boundary under irregular shapes or complex terrains. In particular, in areas where the wind direction suddenly changes or the terrain changes significantly, the CA rule is used to dynamically adjust the propagation direction and rate, thereby improving the ability to depict the boundary shape.

[0075] In some of the embodiments, the fire propagation rate is calculated based on the following calculation model:

[0076] wherein, is an initial propagation rate, which is the natural expansion rate of the fire under the conditions of no wind, no slope and uniform combustible materials; is a fuel correction factor, used to reflect the influence of fuel type, moisture content and density on fire spread; is a terrain correction factor, used to correct the acceleration or deceleration effect of terrain features such as slope, concave and convex on fire spread; is a wind correction factor, considering the influence of wind speed and direction on fire tongue stretching and conduction direction; is a wind direction angle, is a slope direction angle, the included angle between the two determines the resultant propagation direction of wind and terrain, if the wind and slope direction are consistent, the propagation is the fastest, if the wind and slope are opposite, the propagation is weakened; the fire spread rate is input into the Rothermel model to simulate fire spread.

[0077] Further, when processing the new fire point reported by the search unmanned aerial vehicle, the central server combines the terrain grid data of the current position, the wind field simulation diagram and the vegetation coverage diagram to calculate the propagation rate distribution diagram of the region in real time, and drives the boundary evolution simulation and response time estimation based on the diagram, so as to dynamically adjust the number of fire extinguishing unmanned aerial vehicles and the scheduling priority.

[0078] For example but not limited to, the initial propagation rate of grassland or shrub land is about 0.05-0.15 m / s, and the initial propagation rate of forest needle leaf ground fire is about 0.01-0.03 m / s. The fuel correction factor can be based on the influence factor of vegetation type, fuel particle size, moisture content and the like on the increase and decrease of propagation rate, and the numerical range thereof is about 0.3-2.5. The terrain correction factor can be defined as linear or exponential variation with slope angle. For example, the terrain correction factor of flat land is 1.0, and the terrain correction factor of uphill is greater than 1.0. The wind correction factor can be configured to be positively correlated with wind speed, and the wind speed data can be obtained based on a meteorological interface.

[0079] In some embodiments, the determination of the required number of fire extinguishing unmanned aerial vehicles based on the predicted fire area can be obtained based on the following calculation model:

[0080] wherein, is a fire extinguishing efficiency, indicating the fire area that can be effectively suppressed per unit of fire extinguishing load, is a fire extinguishing capacity of a single unmanned aerial vehicle, and the number of unmanned aerial vehicles is dynamically adjusted by dividing the area by the fire extinguishing efficiency and the fire extinguishing capacity.

[0081] In some embodiments, after determining the required number of fire extinguishing unmanned aerial vehicles based on the predicted fire area, the unmanned aerial vehicles are allocated to perform fire extinguishing tasks: The sector angle is divided by the number of unmanned aerial vehicles in the fire extinguishing unmanned aerial vehicle group on average The sector area is calculated based on the sector angle , is a fire range radius; Calculate the coverage area of ​​a single fire bomb , based on the coverage area of ​​a single fire bomb, estimate the number of fire bombs required for each sector , used to reflect the amount of fire-fighting bombs required for each UAV sector mission; in, , is the explosion radius of the fire extinguishing bomb, The fire extinguishing bomb efficiency is constrained to a value range of 0.7~0.9, taking into account the influence of terrain and wind; ; According to the number of fire-fighting bombs carried by a single fire-fighting drone Calculate the total fire-fighting bomb capacity of a fire-fighting drone swarm , to determine whether the fire-fighting resources in the current mission area are sufficient, based on the total fire-fighting bomb capacity The fire extinguishing efficiency factor is evaluated by the ratio of the number of fire extinguishing bombs required for each sector ; like , it indicates that the current number of fire-fighting bombs is insufficient and the number of fire-fighting drones needs to be expanded. Otherwise, the current ammunition resources meet the mission delivery requirements. , .

[0082] In some embodiments, in order to further evaluate the effectiveness of the fire extinguishing bomb's mission, the system calculates the coverage boundary length of the fire extinguishing bomb based on the fire extinguishing efficiency factor. , used to assist in fire line control and path planning, among which, .

[0083] In the above embodiment, the type of fire extinguishing bomb can be replaced with water bomb, aerosol bomb, dry powder bomb, etc. according to the actual load.

[0084] In some embodiments, the fire area change rate is expressed as: , in, Fire point The circumference of Fire point The current area, , fire point The shape is approximately circular.

[0085] In some embodiments, the adjustment of the ratio of the number of search drones to the number of firefighting drones is specifically as follows: , To adjust the step size, is the area change threshold.

[0086] Based on the above embodiment, the system adjusts the number ratio of the search unmanned aerial vehicle and the fire extinguishing unmanned aerial vehicle by monitoring the change rate of the fire point area over time. If the change rate of the fire area is greater than 0, it indicates that the fire is expanding, and the system increases the number of search unmanned aerial vehicles to expand the reconnaissance area and find more new fire points. Conversely, when the area change rate tends to be stable or negative, the system reduces the number of search unmanned aerial vehicles and increases the number of fire extinguishing unmanned aerial vehicles to strengthen the suppression capability.

[0087] The above embodiment realizes a task resource adaptive regulation strategy based on the dynamic change of the fire, and combines a real-time monitoring and feedback control mechanism for the fire point, so that the system has stronger scheduling flexibility and response efficiency when facing complex, multi-point and multi-stage fire evolution.

[0088] It should be noted that the above modules can be functional modules or program modules, and can be implemented by software or hardware. For modules implemented by hardware, the above modules can be located in the same processor; or the above modules can also be located in different processors in any combination.

[0089] The technical features of the above-described embodiments can be combined in any manner. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described, but it should be considered that any combination of the technical features is within the scope of the present disclosure.

[0090] The above-described embodiments only express several implementation manners of the present application, and the description is specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be noted that for those skilled in the art, without departing from the concept of the present application, some modifications and improvements can be made, which are all within the scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A forest fire extinguishing system based on multiple swarms of drones, characterized in that: include: Multiple drone swarms equipped with temperature sensing equipment; A central server, communicatively connected to the drone swarm, is configured to include: The task allocation module configures the drone group to include a search drone group and a fire-fighting drone group based on the estimated fire area. , estimated response time , the number ratio of search drones and firefighting drones is determined by minimizing the total response time as the allocation optimization function; Mission execution module, divides the target search area, drives the search drone group to the target search area Search mission, when the search drone detects a fire point When the fire area is predicted, the number of fire-fighting drones required is determined based on the predicted fire area, and the determined number of fire-fighting drones are driven to fly to the target fire point area to perform fire-fighting operations; The task monitoring module detects the fire area change rate in real time and adjusts the ratio of the number of search drones and fire-fighting drones according to the ratio of the fire area change rate to an area change threshold.

2. The forest fire extinguishing system based on multiple swarms of drones according to claim 1 is characterized in that: The task monitoring module is configured to: Real-time monitoring of the target fire point's perceived temperature based on the temperature sensing device. If the perceived temperature is lower than a set temperature threshold , then the fire extinguishing mission of the target fire point is judged to be completed, and the drone performing the fire extinguishing mission of the target fire point is released back to the resource pool.

3. The forest fire extinguishing system based on multiple swarms of drones according to claim 1 or 2, characterized in that: The driving search drone group performs a target search area The search tasks further include: Generate flight waypoints for the search drone based on the Levy flight strategy, perform path search based on the A* algorithm, and obtain the total path cost. Lowest flight path.

4. The forest fire extinguishing system based on multiple swarms of drones according to claim 3 is characterized in that: The ratio of search drones to firefighting drones is calculated based on the following calculation model: in, To search for the number of drones, is the total number of drones in the drone swarm, , To allocate an optimization function, , total response time Expressed as: , Time of fire discovery, Time to extinguish the fire.

5. The forest fire extinguishing system based on multiple swarms of drones according to claim 4 is characterized in that: The predicted fire area specifically includes: Define the time field of the fire boundary and simulate the evolution of the fire boundary based on the Eikonal equation to predict the time field of the fire boundary ; The predicted fire area is calculated based on the time field of the fire boundary and the fire spread rate: in, is the integral of the fire boundary along the current fire, is the length of the boundary element, is the fire spread rate.

6. The forest fire extinguishing system based on multiple swarms of drones according to claim 5 is characterized in that: Fire spread rate Calculated based on the following calculation model: in, is the initial transmission rate; is the fuel correction factor; is the terrain correction factor; is the wind correction factor; is the wind direction angle, is the slope angle.

7. The forest fire extinguishing system based on multiple swarms of drones according to claim 6 is characterized in that: The number of firefighting drones required based on the predicted fire area can be determined based on the following calculation model: in, For fire extinguishing efficiency, The fire extinguishing capacity of a single drone.

8. The forest fire extinguishing system based on multiple swarms of drones according to claim 7 is characterized in that: After determining the number of firefighting drones required based on the predicted fire area, assign drones to perform firefighting tasks: The sector angle is evenly divided based on the number of drones in the fire-fighting drone group , calculating the area of ​​a single sector based on the sector angle; Calculate the coverage area of ​​a single fire bomb , based on the coverage area of ​​a single fire bomb, estimate the number of fire bombs required for each sector ; According to the number of fire-fighting bombs carried by a single fire-fighting drone Calculate the total fire-fighting bomb capacity of a fire-fighting drone swarm , based on total fire extinguishing bomb capacity and the number of fire extinguishers required for each sector Fire extinguishing efficiency factor ,like , then expand the number of fire-fighting drones.

9. The forest fire extinguishing system based on multiple swarms of drones according to claim 8, characterized in that: The fire area change rate is expressed as: ， in, Fire point The circumference of Fire point The current area, , fire point The shape is approximately circular.

10. The forest fire extinguishing system based on multiple swarms of drones according to claim 9, characterized in that: The adjustment of the ratio of search drones to firefighting drones is as follows: , To adjust the step size, is the area change threshold.

Images