Unmanned aerial vehicle forest fire extinguishing bomb automatic throwing system based on visual identification

Abstract

The invention relates to the technical field of unmanned aerial vehicle fire extinguishing, and discloses an unmanned aerial vehicle forest fire extinguishing bomb automatic throwing system based on visual identification, and the system comprises a plurality of unmanned aerial vehicles, each unmanned aerial vehicle is provided with a plurality of sensing devices, a fire extinguishing bomb mounting device and an external hanging groove; the plurality of unmanned aerial vehicles comprise a mother vehicle, an alternative vehicle and a son vehicle; a master machine control chip is arranged in the master machine core cabin; the plurality of sensing devices comprise thermal imaging sensors, cameras and satellite positioning sensors and are used for acquiring fire behavior data in real time; the fire extinguishing bomb mounting device is used for mounting and throwing fire extinguishing bombs; the external hanging groove is used for loading a standby battery or a fire extinguishing material; the master control chip generates fire information, formulates a first fire extinguishing scheme and generates a first fire extinguishing result based on the thermal imaging information, the image information and the positioning information; and if not, generating a second fire extinguishing scheme, and scheduling the sub-machine to execute fire extinguishing operation. Therefore, intelligent collaborative fire extinguishing of the unmanned aerial vehicle group is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle (UAV) fire extinguishing technology, and in particular to an automatic delivery system for UAV forest fire extinguishing bombs based on visual recognition. Background Art

[0002] Forest fires, caused by the burning of combustible materials, typically occur in densely vegetated areas such as forests and grasslands. Forest firefighting primarily relies on ground-based firefighting teams and the delivery of firefighting munitions from helicopters or fixed-wing aircraft. However, these traditional methods suffer from limited coverage, slow response times, and low precision. In recent years, the development of drone technology has provided a new solution for forest firefighting. By carrying firefighting munitions and sensing equipment, drones enable remote fire monitoring and precise delivery, improving both efficiency and safety.

[0003] Existing forest fire fighting methods, especially drone fire fighting, often have the following technical problems: First, existing drone firefighting methods typically rely solely on the deployment of pre-set firefighting ammunition, and are unable to dynamically adjust firefighting strategies based on changing fire conditions, resulting in limited firefighting effectiveness. This is especially true in complex and volatile fires, where it's difficult to find the optimal firefighting solution. Furthermore, existing drones have limited endurance, making it difficult to sustain operations at distant fires. This reduces the efficiency of firefighting operations, making it difficult to maintain stable firefighting effectiveness, especially in large-scale fires or long-duration operations. Second, existing drone forest firefighting systems often struggle to switch command in a timely manner when the mother drone malfunctions. This leads to the risk of command interruption in drone swarms in emergencies, affecting the continuity and coordination of firefighting operations and reducing overall firefighting efficiency. Third, existing drone forest firefighting systems typically rely on direct communication between the main drone and the control terminal for data synchronization during long-distance or complex terrain firefighting operations. However, in the changing environment of a fire scene, the communication distance between the main drone and the control terminal may exceed the preset distance threshold, resulting in communication interruption or delay, which in turn affects the execution and scheduling of firefighting operations. Summary of the Invention

[0004] This summary is intended to briefly introduce concepts that will be described in detail in the detailed description below. This summary is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0005] The present invention proposes an automatic delivery system for forest fire extinguishing bombs by unmanned aerial vehicles based on visual recognition to solve one or more of the technical problems mentioned in the above background technology section.

[0006] The present invention provides a UAV forest fire extinguishing bomb automatic delivery system based on visual recognition, comprising: Multiple drones, each of the multiple drones has a corresponding drone number and a fire extinguishing ammunition information table, and is equipped with multiple sensing devices, a fire extinguishing bomb mounting device, and an external mounting slot; the multiple drones include a mother drone, at least one backup drone, and at least one daughter drone; the external mounting slot, the fire extinguishing bomb mounting device, and the multiple sensing devices are fixedly connected to the drones; Multiple perception devices include thermal imaging sensors, cameras, and satellite positioning sensors; The fire extinguishing bomb mounting device is used to mount and release fire extinguishing bombs; The external hanging slots have corresponding loading types. The external hanging slots are used to load spare batteries or fire extinguishing materials. The mother machine control chip installed in the core cabin of the mother machine is used to generate fire information and a first fire extinguishing plan based on the thermal imaging information corresponding to the thermal imaging sensor, the image information corresponding to the camera, and the positioning information corresponding to the satellite positioning sensor, and execute the first fire extinguishing plan; generate a first fire extinguishing result through multiple sensing devices; if the first fire extinguishing result indicates that the fire has not been extinguished, generate a second fire extinguishing plan based on the fire information, loading type and drone number.

[0007] Optionally, the alternative machine is the first alternative machine or the second alternative machine, the external hanging slot includes a first external hanging slot and a second external hanging slot, and the loading type is one of the following: the first loading type, the second loading type or the third loading type; a ejection device is also provided in the external hanging slot; if the first external hanging slot is loaded with a spare battery, and the second external hanging slot is loaded with a spare battery, the corresponding UAV is determined as the first alternative machine; if the first external hanging slot is loaded with spare fire extinguishing materials, and the second external hanging slot is loaded with spare fire extinguishing materials, the corresponding UAV is determined as a sub-machine; if the first external hanging slot is loaded with a spare battery, and the second external hanging slot is loaded with spare fire extinguishing materials, or the first external hanging slot is loaded with spare fire extinguishing materials, and the second external hanging slot is loaded with a spare battery, the corresponding UAV is determined as the second alternative machine.

[0008] Optionally, the fire information includes the drone number, the drone's real-time coordinates, the combustion state, and the combustion type. The first fire extinguishing plan is determined by the following steps: Based on the thermal imaging information, determine the temperature abnormality area, select the area with the highest temperature from the temperature abnormality area, and determine it as the delivery area; determine the real-time coordinates of the drone based on the positioning information; determine the vegetation type based on the real-time coordinates of the drone and a preset regional vegetation information table; determine the combustion type and the type of the burning material based on the vegetation type and image information; determine the delivery ammunition number based on the combustion type, the type of the burning material, and the fire extinguishing ammunition information table; generate a first fire extinguishing plan based on the delivery ammunition number and the delivery area; The first fire extinguishing result is determined by the following steps: Through multiple sensing devices, first thermal imaging information is obtained, and it is determined whether the highest temperature in the temperature abnormality area in the first thermal imaging information is lower than a preset safety temperature threshold; if not, the first fire extinguishing result is determined as not extinguished; if so, the first image information and the second image information are respectively obtained according to the preset image acquisition interval; based on the first image information, a first smoke ratio is determined; based on the second image information, a second smoke ratio is determined; based on the first smoke ratio and the second smoke ratio, a smoke change ratio is determined; if the smoke change ratio is greater than the preset smoke change ratio threshold, the first fire extinguishing result is determined as extinguished.

[0009] Optionally, the second firefighting plan includes one of the following: a drone support plan, a ground firefighting equipment support plan, or a collaborative support plan; the second firefighting plan is determined by the following steps: Through multiple sensing devices, second thermal imaging information and third image information are obtained; based on the second thermal imaging information and the third image information, the fire status is determined; if the fire status is characterized as small, a drone support plan is generated; if the fire status is characterized as large, the distance between the drone's real-time coordinates and the forest boundary is obtained based on the drone's real-time coordinates to obtain the forest boundary distance; the forest boundary distance is compared with the boundary distance threshold to obtain a distance comparison result; if the distance comparison result is characterized as relatively close, a ground fire extinguishing equipment support plan is generated; if the distance comparison result is characterized as relatively far, a collaborative support plan is generated.

[0010] Optionally, the control chip is deployed with multiple strategies, the multiple strategies including at least one of the following: a body monitoring strategy, an inspection strategy, a fire extinguishing bomb delivery strategy, and a sub-machine scheduling strategy; The aircraft monitoring strategy is as follows: During the operation of the drone, the aircraft is monitored according to the preset monitoring frequency, including: power information, sensor operation information, hanging slot information and fire bomb information, and recorded in the aircraft information table; if the real-time power in the power information is less than or equal to the selected power threshold, it is judged as low power; The inspection strategy is: carry out inspections according to the preset inspection plan. If no abnormalities are found, the aircraft will return to the destination after the inspection. The fire extinguishing bomb delivery strategy is as follows: based on the type of combustion and the type of burning material, the fire extinguishing bomb delivery order is determined from the fire extinguishing ammunition information table, and the fire extinguishing bombs are released through the fire extinguishing bomb mounting device; it is determined whether the external hanging slot carries fire extinguishing materials that match the fire information. If so, the fire extinguishing materials are released through the ejection device; The deployment result evaluation strategy is as follows: based on multiple sensors and a preset evaluation frequency, multiple evaluation results are generated; based on the multiple evaluation results, multiple indicator change trend graphs are generated; if any indicator change trend graph in the multiple indicator change trend graphs indicates an indicator increase, the deployment result is determined to be unextinguished; The sub-machine scheduling strategy is as follows: if the deployment result indicates that the fire has not been extinguished, obtain the real-time coordinates of the drone, and match at least one drone number from the preset drone fire extinguishing configuration table based on the combustion type and the type of burning material; generate sub-machine scheduling information based on at least one drone number and the real-time coordinates of the drone.

[0011] Optionally, the multiple strategies deployed in the control chips of the master machine, the first candidate machine, and the second candidate machine further include a master machine election strategy; The master machine election strategy includes active election strategy and passive election strategy. The master machine control chip activates the active election strategy, and the control chips of the first and second backup machines activate the passive election strategy. Active election involves obtaining the aircraft information table and the aircraft monitoring information. If any abnormality is found in the aircraft monitoring information, an active election is triggered, and the drone number corresponding to the target candidate aircraft is determined from the candidate aircraft information table. The candidate aircraft information table is determined through the following steps: Obtain the drone numbers, power information, mission status, and location information corresponding to the first candidate drone and the second candidate drone to generate a candidate drone information table; the candidate drone information table is synchronized to the candidate drones at a preset synchronization frequency; Passive election is as follows: if any candidate among the candidate machines does not receive a response signal from the master machine after the response time threshold, the number of candidate machines that have not received a response signal from the master machine among the remaining candidate machines is counted; if the number is greater than or equal to the non-response threshold, the candidate machine information table is used to determine whether it is qualified as the master machine. If so, it is bootstrapped as the target candidate machine and broadcast.

[0012] Optionally, the UAV forest fire extinguishing bomb automatic delivery system based on visual recognition of the present invention further includes: The control terminal is used to receive and process feedback information from multiple drones and send call information to ground fire-fighting equipment.

[0013] Optionally, the UAV forest fire extinguishing bomb automatic delivery system based on visual recognition of the present invention further includes: Ground fire extinguishing equipment, ground fire extinguishing equipment is used to receive call information sent by the control terminal to carry out fire extinguishing support operations.

[0014] Optionally, multiple strategies also include return-to-home strategies; return-to-home strategies have different logics for different types of drones; For the slave aircraft, the fire extinguishing material consumption information is extracted from the fire extinguishing ammunition information table. If the fire extinguishing material consumption information indicates that all fire extinguishing materials are consumed or the remaining fire extinguishing materials do not match the fire extinguishing operation, the slave aircraft returns to the preset return point. For the first backup aircraft, extract fire extinguishing material consumption information from the fire extinguishing ammunition information table. If the fire extinguishing material consumption information indicates that all fire extinguishing materials have been consumed or the remaining fire extinguishing materials do not match the fire extinguishing operation, obtain the real-time power of the first backup aircraft. If the real-time power of the first backup aircraft is greater than the first backup power threshold, obtain the first actual return distance. If the first actual return distance is greater than the first return distance threshold, and the real-time power of the mother aircraft is greater than the continuous inspection power threshold, return to the preset return point. For the second candidate aircraft, extract the fire extinguishing material consumption information from the fire extinguishing ammunition information table. If the fire extinguishing material consumption information indicates that all fire extinguishing materials have been consumed or the remaining fire extinguishing materials do not match the fire extinguishing operation, obtain the real-time power of the second candidate aircraft. If the real-time power of the second candidate aircraft is greater than the second candidate power threshold, obtain the second actual return distance. If the second actual return distance is less than the second return distance threshold and the real-time power of the mother aircraft is greater than the continuous inspection power threshold, return to the preset return point. For the mother aircraft, obtain the real-time coordinates of the mother aircraft's drone, and determine the minimum power for return based on the preset return point; determine the return power threshold based on the minimum power for return and the floating power; compare the return power threshold with the election power threshold. If the return power threshold is greater than the election power threshold, update the election power threshold to the return power threshold.

[0015] Optionally, the float charge is determined by the following steps: Obtain a temporary home point between the preset home point and the operating area, and generate a temporary home point information table; obtain the real-time coordinates of the drone of the mother aircraft, and select the temporary home point closest to the real-time coordinates of the drone of the mother aircraft from the temporary home point information table, and determine it as the actual home point; obtain the power consumption between the real-time coordinates of the drone of the mother aircraft and the actual home point, and determine it as the first floating power; Obtain the real-time coordinates of the target candidate drone, and determine the second floating power according to the real-time coordinates of the drone of the mother aircraft; A floating power is generated according to the first floating power and the second floating power. The present invention has the following beneficial effects: 1. Improved the intelligence and execution efficiency of UAV forest firefighting operations. Specifically, the mother aircraft uses multiple sensors to sense the fire situation in real time and dynamically adjust the firefighting plan, which improves strategic adaptability compared to fixed ammunition delivery methods. External hanging slots can be loaded with spare batteries or firefighting materials, extending flight time and enhancing long-distance operation capabilities. The mother aircraft, sub-aircraft, and backup aircraft work together to accurately deliver firefighting bombs, improving firefighting efficiency and stability. Through these optimizations, the UAV firefighting system is more efficient and flexible, reducing dependence on manpower and enhancing autonomous decision-making capabilities. 2. Improved the command stability and operational continuity of the drone firefighting system. Specifically, both the main and backup drones deploy a main drone election strategy. When a main drone fails, an active or passive election mechanism can be used to quickly determine a new command drone to avoid command interruption. The drone monitoring strategy can monitor the drone status in real time, trigger elections in advance, and reduce the impact of sudden failures. The backup drone information table is updated at a synchronous frequency to ensure accurate and efficient election decisions. Through these optimizations, the drone cluster can still maintain coordinated operations in emergency situations, improving firefighting efficiency and stability. 3. Improved data synchronization stability for drone firefighting systems operating in complex terrain and over long distances. Specifically, the master drone prioritizes direct synchronization for data exchange with the control terminal. When the communication distance exceeds a threshold or no response is received, the relay synchronization strategy is automatically triggered, selecting appropriate slave drones as relay nodes and forwarding synchronized data step by step to ensure uninterrupted communication. This improves communication reliability for drone swarms in complex environments, ensuring smooth execution and precise scheduling of firefighting operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The above and other features, advantages, and aspects of the various embodiments of the present invention will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that the elements are not necessarily drawn to scale.

[0017] Figure 1 1 is a schematic diagram of an exemplary system structure of the automatic delivery system of forest fire extinguishing bombs by a UAV based on visual recognition according to the present invention; Figure 2 It is a stereoscopic diagram of a UAV in the automatic delivery system of UAV forest fire extinguishing bombs based on visual recognition of the present invention; Figure 3 This is a right view of the UAV in the automatic delivery system of UAV forest fire extinguishing bombs based on visual recognition of the present invention; Figure 4 This is a front view of a UAV in the automatic delivery system of UAV forest fire extinguishing bombs based on visual recognition of the present invention; Figure 5 This is a top view of a drone in the automatic delivery system of a drone forest fire extinguishing bomb based on visual recognition according to the present invention; Figure 6 The present invention provides a bottom view of a UAV in the automatic delivery system for UAV forest fire extinguishing bombs based on visual recognition. DETAILED DESCRIPTION

[0018] The present invention will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0019] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features of the embodiments of the present invention may be combined with each other.

[0020] It should be noted that the concepts of "first" and "second" mentioned in the present invention are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0021] It should be noted that the modifications of "one" and "multiple" mentioned in the present invention are illustrative rather than restrictive. Those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0022] The names of the messages or information exchanged between multiple devices of the present invention are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0023] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0024] like Figure 1 FIG. 1 is a schematic diagram of an exemplary system structure of the automatic delivery system of a UAV forest fire extinguishing bomb based on visual recognition according to the present invention, which specifically includes the following modules: Multiple drones 101, each of the multiple drones 101 has a corresponding drone number and a fire extinguishing ammunition information table, and is equipped with multiple sensing devices, fire extinguishing bomb mounting devices and external hanging slots; the multiple drones include a mother machine, at least one backup machine and at least one sub-machine; the external hanging slots, the fire extinguishing bomb mounting device and the multiple sensing devices are fixedly connected to the drones.

[0025] In some embodiments, each of the multiple drones 101 has a unique drone number for identification and management. Furthermore, each drone stores a corresponding fire extinguishing ammunition information table, which records the type of fire extinguishing ammunition currently carried by the drone and its corresponding fire extinguishing ammunition number, allowing for appropriate deployment of ammunition for different fire situations during firefighting operations.

[0026] In addition, if Figure 2As shown, each drone is equipped with multiple sensing devices, a fire extinguisher grenade mount 204, and an external mounting slot 202. The multiple sensing devices include a thermal imaging sensor 206, a camera 209, and a satellite positioning sensor 207. The thermal imaging sensor 206 is used to collect real-time temperature data from the fire scene, the camera 209 is used to collect real-time image data from the fire scene, and the satellite positioning sensor 207 is used to obtain real-time drone location information. The fire extinguisher grenade mount 204 is used to carry and deploy fire extinguisher grenade 208, enabling precise firefighting operations. The external mounting slot 202 can be loaded with backup batteries or fire extinguishing materials according to mission requirements, giving the drone greater operational flexibility and endurance when performing missions.

[0027] In terms of the composition of a drone swarm, multiple drones can be divided into a master drone, backup drones, or slave drones. The master drone is responsible for overall control, including collecting fire information, formulating firefighting strategies, and scheduling and managing slave drones. The backup drones are used to take over command tasks in the event of a master drone failure or loss of connection, ensuring the system's self-recovery capabilities. The slave drones perform specific firefighting operations and receive dispatch instructions from the master drone or backup drones. The master drone and backup drones can also extinguish fires using fire extinguishing bombs 208 mounted on a fire extinguishing bomb mounting device 204. The drone's external mounting slot 202, fire extinguishing bomb mounting device 204, and multiple sensing devices are all fixedly connected to the drone body to ensure stable operation during flight and meet the execution requirements of firefighting operations.

[0028] The plurality of sensing devices include a thermal imaging sensor 206 , a camera 209 and a satellite positioning sensor 207 .

[0029] In some embodiments, as Figure 3 As shown, the multiple sensing devices include a thermal imaging sensor 206, a camera 209, and a satellite positioning sensor 207. The thermal imaging sensor 206 is used to collect real-time fire scene temperature data, the camera 209 is used to collect real-time fire scene image data, and the satellite positioning sensor 207 is used to obtain real-time drone location information. The drone location information includes the drone's real-time coordinates.

[0030] In practice, the thermal imaging sensor 206 may be a Lepton 3.5, the camera 209 may be an IMX477, and the satellite positioning sensor 207 may be a ZED-F9P.

[0031] The Lepton 3.5 is a miniature thermal imaging module designed to capture and measure temperature changes in objects or environments. The IMX477 is a high-resolution imaging sensor designed to capture high-quality still and dynamic images. The ZED-F9P is a high-precision satellite positioning system designed to obtain precise location information.

[0032] The external hanging slot 202 has a corresponding loading type, and the external hanging slot 202 is used to load spare batteries or fire extinguishing materials.

[0033] In some embodiments, as Figure 4 As shown, the external hanging slot 202 is a modular load device installed on the UAV, and its loading content can be flexibly adjusted according to mission requirements, for example, loading spare batteries or spare fire extinguishing materials to enhance the UAV's endurance or fire extinguishing capability.

[0034] Specifically, if Figure 5 As shown, the external hanging slot 202 has different loading modes, and suitable materials such as spare batteries or spare fire extinguishing materials can be selected according to the needs of the mission to ensure that the drone has the best operating capabilities in different fire environments. When the drone performs long-term or long-distance firefighting operations, endurance becomes a key factor. The external hanging slot 202 can be used to carry spare batteries and provide additional power to extend the drone's operating time and ensure the continuity of firefighting operations. Figure 6 As shown, when the fire situation is complex, relying solely on fire extinguishing bombs 208 may not be sufficient for firefighting. External hanging slots 202 can be used to load different types of spare firefighting materials to improve firefighting effectiveness. A modular payload system is a flexible design that can be disassembled, replaced, and configured according to mission requirements.

[0035] The mother machine control chip 205 provided in the core cabin 201 of the mother machine is used to generate fire information and a first fire extinguishing plan based on the thermal imaging information corresponding to the thermal imaging sensor 206, the image information corresponding to the camera 209 and the positioning information corresponding to the satellite positioning sensor 207, and execute the first fire extinguishing plan; generate a first fire extinguishing result through multiple sensing devices; if the first fire extinguishing result indicates that the fire has not been extinguished, generate a second fire extinguishing plan based on the fire information, loading type and drone number.

[0036] In some embodiments, a motherboard control chip 205 is integrated within the core module 201 of the motherboard. This chip is responsible for receiving and processing data from multiple sensing devices and formulating a fire extinguishing plan. Specifically, the motherboard control chip 205 obtains corresponding positioning information via a satellite positioning sensor 207, corresponding thermal imaging information via a thermal imaging sensor 206, and corresponding image information via a camera 209. Subsequently, based on the thermal imaging, positioning, and image information, fire status information and a first fire extinguishing plan are generated. After the motherboard completes the first fire extinguishing plan, it again collects corresponding data via the thermal imaging sensor 206 and camera 209 to analyze and generate a first fire extinguishing result. Next, the first fire extinguishing result is evaluated. If the first fire extinguishing result indicates that the fire was not extinguished, a second fire extinguishing plan is generated based on the fire status information after executing the first fire extinguishing plan, the load type corresponding to the external mounting slot 202 of the drone, and the drone number.

[0037] These implementations enhance the intelligence and efficiency of drone-based forest firefighting operations. Specifically, the mother drone utilizes multiple sensors to detect fire conditions in real time and dynamically adjust firefighting strategies, improving strategic adaptability compared to fixed ammunition delivery methods. External mounting slots accommodate backup batteries or firefighting materials, extending flight time and enhancing long-range operation capabilities. The mother drone, slave drones, and backup drones collaborate to precisely deliver firefighting munitions, improving firefighting efficiency and stability. These optimizations make drone firefighting systems more efficient and flexible, reducing reliance on human resources and enhancing autonomous decision-making capabilities.

[0038] In some embodiments, in order to further solve the second technical problem described in the background technology section, that is, "the existing UAV forest fire extinguishing system is often difficult to complete the switching of command authority in time when the mother machine fails, resulting in the risk of command interruption of the UAV cluster in an emergency, affecting the continuity and coordination of the fire extinguishing operation and reducing the overall fire extinguishing efficiency", in some embodiments of the present invention, the alternative machine is the first alternative machine or the second alternative machine, the external hanging slot includes the first external hanging slot and the second external hanging slot, and the loading type is one of the following: the first loading type, the second loading type Or the third loading type; the external hanging slot is also provided with an ejection device; if the first external hanging slot is loaded with a spare battery, and the second external hanging slot is loaded with a spare battery, the corresponding UAV is determined as the first alternative machine; if the first external hanging slot is loaded with spare fire extinguishing materials, and the second external hanging slot is loaded with spare fire extinguishing materials, the corresponding UAV is determined as a sub-machine; if the first external hanging slot is loaded with a spare battery, and the second external hanging slot is loaded with spare fire extinguishing materials, or the first external hanging slot is loaded with spare fire extinguishing materials, and the second external hanging slot is loaded with spare batteries, the corresponding UAV is determined as the second alternative machine.

[0039] In some embodiments, the alternative machine is a first alternative machine or a second alternative machine. The external hanging slot 202 includes a first external hanging slot and a second external hanging slot. The loading type is a first loading type, a second loading type, or a third loading type. If the external hanging slot 202 is loaded with spare fire extinguishing materials, an ejection device is provided on the inner side of the external hanging slot 202 for providing thrust to release the spare fire extinguishing materials when the outer shell of the external hanging slot 202 is opened. Among them, the first loading type is that both the first external hanging slot and the second external hanging slot are loaded with spare batteries, the second loading type is that both the first external hanging slot and the second external hanging slot are loaded with spare fire extinguishing materials, and the third loading type is that the first external hanging slot is loaded with spare batteries and the second external hanging slot is loaded with spare fire extinguishing materials, or the first external hanging slot is loaded with spare fire extinguishing materials and the second external hanging slot is loaded with spare batteries.

[0040] Specifically, if both the first and second external slots of a drone are loaded with backup batteries, the corresponding drone is identified as the first candidate drone. If both the first and second external slots of a drone are loaded with backup fire extinguishing materials, the corresponding drone is identified as a slave drone. If the first external slot of a drone is loaded with backup batteries and the second external slot is loaded with backup fire extinguishing materials, or if the first external slot is loaded with backup fire extinguishing materials and the second external slot is loaded with backup batteries, the corresponding drone is identified as the second candidate drone. The external slot loaded with backup fire extinguishing materials is also equipped with an ejection device to provide thrust to assist in the release of the backup fire extinguishing materials.

[0041] The fire information includes the drone number, real-time coordinates, combustion status, and combustion type. The first fire extinguishing plan is determined through the following steps: Based on the thermal imaging information, determine the temperature abnormality area, select the area with the highest temperature from the temperature abnormality area, and determine it as the delivery area; determine the real-time coordinates of the drone based on the positioning information; determine the vegetation type based on the real-time coordinates of the drone and a preset regional vegetation information table; determine the combustion type and the type of the burning material based on the vegetation type and image information; determine the delivery ammunition number based on the combustion type, the type of the burning material, and the fire extinguishing ammunition information table; generate a first fire extinguishing plan based on the delivery ammunition number and the delivery area; The first fire extinguishing result is determined by the following steps: Through multiple sensing devices, first thermal imaging information is obtained, and it is determined whether the highest temperature in the temperature abnormality area in the first thermal imaging information is lower than a preset safety temperature threshold; if not, the first fire extinguishing result is determined as not extinguished; if so, the first image information and the second image information are respectively obtained according to the preset image acquisition interval; based on the first image information, a first smoke ratio is determined; based on the second image information, a second smoke ratio is determined; based on the first smoke ratio and the second smoke ratio, a smoke change ratio is determined; if the smoke change ratio is greater than the preset smoke change ratio threshold, the first fire extinguishing result is determined as extinguished.

[0042] In some embodiments, the fire information includes the drone number, the drone's real-time coordinates, the burning state, and the burning type. The burning state can be burning, and the burning type can be crown fire or surface fire. The first fire extinguishing plan is determined by the following steps: The motherboard control chip 205 can screen for areas of abnormal temperature from the thermal imaging information from the thermal imaging sensor 206 and select the area with the highest temperature to determine the delivery area. Subsequently, the satellite positioning sensor 207 obtains the corresponding drone's real-time coordinates and, based on the drone's real-time coordinates and a pre-set regional vegetation information table, determines the vegetation type in the current area. Next, the camera 209 uses the image information and vegetation type to determine the type of burning material and combustion. The fire extinguishing ammunition information table then matches the delivery ammunition number appropriate for the current combustion type and combustion type, and, based on the delivery area, generates a first fire extinguishing plan.

[0043] A crown fire is a fire that spreads to the crown or top of a tree, primarily in the upper layers of trees. A surface fire is a fire that occurs on the ground or in the lower vegetation layer. It usually spreads more slowly and affects a smaller area. A temperature anomaly refers to a region where the temperature is significantly higher than the standard temperature of the surrounding environment. In practice, if the normal daytime temperature range for a forested area is 25°C to 30°C, the standard temperature can be set at 29°C, and the abnormal temperature can be 40°C. The burning material can be evergreen forest, deciduous forest, shrubland, or grassland.

[0044] The first fire extinguishing result is determined by the following steps: The motherboard control chip 205 can filter out areas of abnormal temperature from the thermal imaging information corresponding to the thermal imaging sensor 206, select the highest temperature, and determine whether it is below a preset safety temperature threshold. If not, the first fire extinguishing result is determined as unextinguished. If so, the corresponding first and second image information are acquired through the camera 209 according to a preset image acquisition interval. Subsequently, the first and second smoke ratios corresponding to the first and second image information are determined, respectively. Next, the first and second smoke ratios are subtracted to obtain a smoke change ratio. A determination is made as to whether the smoke change ratio exceeds a preset smoke change ratio threshold. If so, the first fire extinguishing result is determined as extinguished. In practice, the preset safety temperature threshold can be 60°C. The preset image acquisition interval can be 20 seconds, and the preset smoke change ratio threshold can be 30%. For example, if the first smoke ratio is 50% and the second smoke ratio is 30%, the fire can be determined to be extinguished.

[0045] In practice, the motherboard control chip 205 can convert the thermal imaging information into a two-dimensional temperature matrix and perform image binarization on the two-dimensional temperature matrix based on a preset temperature threshold to extract areas with abnormal temperatures. The coordinates of the points corresponding to the highest temperature values ​​within the abnormal temperature areas are then obtained to identify the area corresponding to the coordinate points as the target delivery area. The preset temperature threshold can be 60°C. The two-dimensional temperature matrix can be obtained by performing temperature mapping on the grayscale values ​​of each pixel in the thermal imaging information. A two-dimensional temperature matrix is ​​a process that divides the thermal imaging information collected by the thermal imaging sensor into a two-dimensional array by pixel, with each element corresponding to the temperature value of a pixel. This is used to construct the temperature distribution of the target area in the spatial dimension. Image binarization compares each pixel value in the image with a preset temperature threshold and converts the pixel value into one of two discrete values ​​(e.g., 0 or 1) to distinguish between high-temperature and non-high-temperature areas. A grayscale value refers to the brightness value corresponding to a single pixel in a thermal image. A higher grayscale value indicates a greater radiation intensity in the pixel area and a corresponding higher temperature value. Grayscale values ​​are the basic unit for digitizing thermal images. Temperature mapping is the process of converting the grayscale value of each pixel in a thermal image into its corresponding actual temperature value based on the sensor's temperature response characteristics. The sensor's temperature response characteristics refer to how the thermal imaging sensor reacts to different temperature values ​​and its accuracy.

[0046] In addition, the motherboard control chip 205 can also perform preprocessing on the image information to enhance the recognizability of smoke areas. Specifically, the image is first converted from the RGB color space to the HSV color space to highlight the low saturation and low brightness characteristics of the smoke area. Subsequently, a median filter is used to denoise the image, preserving the smoke characteristics. Next, a texture feature analysis method, such as a local binary pattern (LBP), is used to analyze the image and extract the smoke area. Finally, the proportion of the smoke area is calculated to facilitate subsequent analysis and decision-making. The RGB color space is a color model based on the three primary colors of red, green, and blue. The HSV color space represents color as hue, saturation, and brightness. Median filtering is a commonly used image denoising method that replaces the current pixel value with the median value of the pixels in its neighborhood. Texture feature analysis is a technique for analyzing and classifying images by extracting texture information from the image. A LBP is a texture descriptor that compares the grayscale values ​​of each pixel with those of its neighbors to generate a binary number representing the local texture of that pixel.

[0047] The second firefighting plan includes one of the following: drone support plan, ground firefighting equipment support plan, or collaborative support plan. The second firefighting plan is determined by the following steps: Through multiple sensing devices, second thermal imaging information and third image information are obtained; based on the second thermal imaging information and the third image information, the fire status is determined; if the fire status is characterized as small, a drone support plan is generated; if the fire status is characterized as large, the distance between the drone's real-time coordinates and the forest boundary is obtained based on the drone's real-time coordinates to obtain the forest boundary distance; the forest boundary distance is compared with the boundary distance threshold to obtain a distance comparison result; if the distance comparison result is characterized as relatively close, a ground fire extinguishing equipment support plan is generated; if the distance comparison result is characterized as relatively far, a collaborative support plan is generated.

[0048] In some embodiments, the second fire extinguishing solution is a drone support solution, a ground fire extinguishing equipment support solution, or a collaborative support solution. The second fire extinguishing solution is determined by the following steps: The master control chip 205 first acquires the corresponding second thermal image information and third image information via the thermal imaging sensor 206 and camera 209. It then determines the fire severity based on the temperature anomaly areas in the second thermal image information and the smoke ratio in the third image information. Specifically, if the temperature anomaly areas account for greater than 60%, or the smoke ratio accounts for greater than 40%, the fire severity is determined to be severe. If the temperature anomaly areas account for less than or equal to 60%, or the smoke ratio accounts for less than or equal to 40%, the fire severity is determined to be mild. When the fire severity is mild, a drone support plan is generated using the slave control chip scheduling strategy. When the fire severity is severe, the satellite positioning sensor 207 acquires the corresponding drone's real-time coordinates. Using the drone's real-time coordinates and built-in preset map information, the distance to the forest boundary is determined and then compared with a threshold distance. If the comparison indicates a close proximity, the master control chip 205 sends feedback to the control terminal 102, which instructs the control terminal 102 to send a call to the ground fire extinguishing equipment 103. If the comparison result indicates a long distance, a drone support plan is first generated through the slave machine scheduling strategy, and then the ground fire extinguishing equipment 103 is called through the control terminal 102.

[0049] Map information represents the geographic spatial structure of the corresponding area, including vector boundary line data or rasterized layer information of the forest boundary. In practice, this map information can be pre-generated by a geographic information system and stored as an electronic map on the host aircraft for subsequent distance calculation and positioning analysis. The boundary distance threshold determines the relative position between the drone and the forest boundary and can be 100 meters in practice.

[0050] The second firefighting plan is generated based on real-time data provided by multiple sensing devices, including secondary thermal imaging and tertiary image information. The system first analyzes this information to determine the severity of the fire. If the analysis indicates the fire is small, the system generates a drone support plan to facilitate further firefighting operations using drones. If the fire is larger, the system calculates the distance between the drone and the forest boundary based on the drone's real-time coordinates and compares this distance to a preset boundary distance threshold. If the distance between the drone and the forest boundary is close, the system generates a ground-based firefighting equipment support plan to effectively mobilize ground-based firefighting equipment. If the distance is greater, the system generates a coordinated support plan to coordinate drones with other support resources for joint firefighting operations.

[0051] The control chip is deployed with multiple strategies, including at least one of the following: a body monitoring strategy, an inspection strategy, a fire extinguishing bomb delivery strategy, and a sub-machine scheduling strategy; The aircraft monitoring strategy is as follows: During the operation of the drone, the aircraft is monitored according to the preset monitoring frequency, including: power information, sensor operation information, hanging slot information and fire bomb information, and recorded in the aircraft information table; if the real-time power in the power information is less than or equal to the selected power threshold, it is judged as low power; The inspection strategy is: carry out inspections according to the preset inspection plan. If no abnormalities are found, the aircraft will return to the destination after the inspection. The fire extinguishing bomb delivery strategy is as follows: based on the type of combustion and the type of burning material, the fire extinguishing bomb delivery order is determined from the fire extinguishing ammunition information table, and the fire extinguishing bombs are released through the fire extinguishing bomb mounting device; it is determined whether the external hanging slot carries fire extinguishing materials that match the fire information. If so, the fire extinguishing materials are released through the ejection device; The deployment result evaluation strategy is as follows: based on multiple sensors and a preset evaluation frequency, multiple evaluation results are generated; based on the multiple evaluation results, multiple indicator change trend graphs are generated; if any indicator change trend graph in the multiple indicator change trend graphs indicates an indicator increase, the deployment result is determined to be unextinguished; The sub-machine scheduling strategy is as follows: if the deployment result indicates that the fire has not been extinguished, obtain the real-time coordinates of the drone, and match at least one drone number from the preset drone fire extinguishing configuration table based on the combustion type and the type of burning material; generate sub-machine scheduling information based on at least one drone number and the real-time coordinates of the drone.

[0052] In some embodiments, multiple strategies are deployed in the control chip, including at least one of the following: a machine body monitoring strategy, an inspection strategy, a fire extinguishing bomb delivery strategy, and a sub-machine scheduling strategy.

[0053] Among them, the body monitoring strategy is: during the operation of the drone, it will monitor multiple body information according to the preset monitoring frequency, including power information, sensor operation information, hanging slot information and fire extinguisher information, and write the monitoring results into the body information table.

[0054] For example, if the real-time power level in the power information is less than or equal to the selected power threshold, if so, the low power level is recorded in the body information table. In practice, the selected power threshold can be 60%, and the preset monitoring frequency can be once every 5 minutes.

[0055] The inspection strategy is: the drone inspects the target area according to the preset inspection plan and returns normally if no abnormalities are found. The preset inspection plan may be to inspect the target area for 40 minutes.

[0056] The fire extinguishing bomb release strategy is as follows: The drone can determine the order in which to release fire extinguishing bombs 208 based on the type of fire and the type of burning material, using a fire extinguishing ammunition information table. For crown fires, water-based fire extinguishing bombs 208 can be released; for evergreen forests, dry powder fire extinguishing bombs 208 can be released. In practice, after determining the order in which to release fire extinguishing bombs 208, the drone's corresponding control chip sends a release command to the release controller 203 mounted on the fire extinguishing bomb mount 204. The release controller 203 then releases the corresponding fire extinguishing bomb 208. The fire extinguishing bomb mount 204 can be equipped with a first and second release controllers, respectively. The first and second release controllers are redundant, preventing fire extinguishing bomb 208 release failures due to release controller failures. Simultaneously, the drone's corresponding control chip also determines whether the external mounting slot 202 contains fire extinguishing materials appropriate for the current fire situation. If so, a release command is sent to the external mounting slot 202. Upon receiving the release command, the external hanger 202 will first open its top cover, then release the backup fire extinguishing material through the power provided by the ejection device installed inside. For example, if the external hanger 202 carries a backup fire extinguishing material made of dry powder, it will be released to assist in extinguishing the fire when the burning type is evergreen forest.

[0057] Among them, water-based fire extinguishing agent is a common fire extinguishing material, which is mainly based on water and can quickly absorb the heat of the fire source and reduce the fire. Dry powder fire extinguishing agent is a commonly used fire extinguishing material, which uses dry powdered fire extinguishing chemicals to cover the fire source, thereby isolating oxygen and absorbing heat to achieve the effect of fire extinguishing. The top cover refers to the component on the external hanging slot 202 that is used to cover and seal the fire extinguishing material inside the hanging slot. The release controller is a key component installed on the fire extinguishing bomb mounting device 204, which is responsible for receiving the release command sent by the drone control chip and performing the corresponding action. The release controller controls the release process of the fire extinguishing bomb 208 to ensure that the fire extinguishing bomb 208 is released at the correct time and location. In order to improve reliability, two release controllers are usually installed for mutual redundancy to prevent the failure of a single controller from causing a release failure.

[0058] The strategy for evaluating deployment results is as follows: the master control chip 205 evaluates the thermal imaging data from the thermal imaging sensor 206 and the image data from the camera 209 at a preset evaluation frequency, generating multiple evaluation results. Each evaluation result includes multiple evaluation indicators, such as temperature and smoke concentration. Next, multiple indicator trend graphs are generated to determine whether any indicator shows an increasing trend. If so, the deployment result is determined as unextinguished. The preset evaluation frequency can be once every minute.

[0059] The slave dispatch strategy is as follows: If the deployment result corresponding to the deployment result evaluation strategy indicates that the fire is not extinguished, the real-time coordinates of the corresponding drone are first obtained through the satellite positioning sensor 207. Based on the combustion type and the type of burning material, the drone numbers suitable for the current fire situation are matched from the preset drone fire extinguishing configuration table. The real-time coordinates of the drones corresponding to these drone numbers are then sent to dispatch the slaves to perform fire extinguishing operations. The preset drone fire extinguishing configuration table includes drone numbers and corresponding fire extinguishing ammunition information. For example, the fire extinguishing ammunition information corresponding to drone number TX-001 may be m-1-a, m-2-a, m-3-b, m-4-b, s-1-a, and s-2-a. m-3-b may be the water-based fire extinguishing bomb 208 with sequence number 3, and s-1-a may be the backup fire extinguishing material with sequence number 1 and made of dry powder.

[0060] In practice, the master can activate the airframe monitoring strategy, inspection strategy, fire bomb delivery strategy, delivery result evaluation strategy, and the slave scheduling strategy. The slave can activate the airframe monitoring strategy and fire bomb delivery strategy. The backup can activate the airframe monitoring strategy, inspection strategy, and fire bomb delivery strategy. When the backup becomes the target backup through master election, it can also activate the delivery result evaluation strategy and the slave scheduling strategy.

[0061] The multiple strategies deployed in the control chips of the master machine, the first candidate machine, and the second candidate machine also include a master machine election strategy; The master machine election strategy includes active election strategy and passive election strategy. The master machine control chip activates the active election strategy, and the control chips of the first and second backup machines activate the passive election strategy. Active election involves obtaining the aircraft information table and the aircraft monitoring information. If any abnormality is found in the aircraft monitoring information, an active election is triggered, and the drone number corresponding to the target candidate aircraft is determined from the candidate aircraft information table. The candidate aircraft information table is determined through the following steps: Obtain the drone numbers, power information, mission status, and location information corresponding to the first candidate drone and the second candidate drone to generate a candidate drone information table; the candidate drone information table is synchronized to the candidate drones at a preset synchronization frequency; Passive election is as follows: if any candidate among the candidate machines does not receive a response signal from the master machine after the response time threshold, the number of candidate machines that have not received a response signal from the master machine among the remaining candidate machines is counted; if the number is greater than or equal to the non-response threshold, the candidate machine information table is used to determine whether it is qualified as the master machine. If so, it is bootstrapped as the target candidate machine and broadcast.

[0062] In some embodiments, the multiple strategies deployed in the control chips corresponding to the master machine, the first candidate machine, and the second candidate machine further include a master machine election strategy, wherein the master machine election strategy includes an active election strategy and a passive election strategy.

[0063] The motherboard control chip 205 can activate the active election strategy. Specifically, the motherboard control chip 205 will traverse the aircraft information table according to the preset aircraft monitoring frequency and determine whether there are any abnormalities. If any abnormality is found, an active election will be performed and the drone numbers corresponding to the qualified candidate aircraft will be screened from the candidate aircraft information table. The candidate aircraft information table can be determined by the following steps: Each time the master control chip 205 communicates with the first or second backup drone, it stores the corresponding drone number, battery level information, mission status, and location information, and generates a backup drone information table. This table is also synchronized to the backup drone based on a preset synchronization frequency.

[0064] When actively electing, the mother machine will filter from the candidate machine information table. First, the candidate machine information table is sorted in descending order according to the real-time power in the power information to obtain a descending table. Then, the candidate machines with idle mission status are filtered from the descending table to obtain a descending idle table. Finally, the candidate machine closest to the mother machine that meets the preset mission conditions is filtered from the descending idle table as the target candidate machine, and the drone number corresponding to the target candidate machine is broadcast to the drone group. In practice, the preset mission condition can be endurance priority or fire fighting priority. For example, when the preset mission condition is endurance priority, the target candidate machine will be filtered from the first candidate machine, and when the preset mission condition is fire fighting priority, the target candidate machine will be filtered from the second candidate machine. Among them, the preset body monitoring frequency can be once per minute, and the preset synchronization frequency can be once every 5 minutes.

[0065] The control chips corresponding to the first and second standby machines can activate a passive election strategy. Specifically, if any of the standby machines fails to receive a response signal from the master machine during a mission, the corresponding standby machine is identified as a failed response machine. The machine then determines whether the unresponsive time between the standby machine and the master machine is greater than or equal to a response time threshold. If so, the failed response machine communicates with the standby machines in the standby machine information table, counting the number of standby machines that have not received a response signal from the master machine to obtain the number of unresponsive standby machines. The machine then determines whether the number of unresponsive standby machines is greater than or equal to a unresponsive threshold. If so, the failed response machine communicates with the standby machines in the standby machine information table, prompting them to bootstrap themselves based on the standby machine information table. In practice, the standby machine with the highest real-time battery life, idle mission status, and closest proximity to the master machine successfully bootstraps. The successfully bootstrapped standby machine then broadcasts its corresponding drone number to the drone swarm. The response time threshold can be 30 seconds, and the unresponsive threshold can be 3 drones.

[0066] Among them, the present invention's automatic delivery system for forest fire extinguishing bombs using a drone based on visual recognition also includes: The control terminal 102 is used to receive and process feedback information from multiple drones 101 and send call information to the ground fire extinguishing equipment 103.

[0067] In some embodiments, the control terminal 102 may be a fixed terminal installed at a control station, or a portable terminal carried by staff. The fixed terminal may be a desktop computer, and the portable terminal may be a tablet computer. In practice, the control terminal 102 receives and processes feedback information from multiple drones 101. Specifically, when multiple drones 101 are within the communication range of the control terminal 102, the control terminal 102 can obtain fire information through feedback information, intervene in firefighting operations, and send call information to ground firefighting equipment 103 to perform firefighting support operations. When multiple drones 101 exceed the communication range of the control terminal 102, the mother machines corresponding to the multiple drones 101 can also autonomously control the fleet to perform firefighting operations. The feedback information may be fire information or synchronization data.

[0068] Among them, the present invention's automatic delivery system for forest fire extinguishing bombs using a drone based on visual recognition also includes: The ground fire extinguishing equipment 103 is used to receive the call information sent by the control terminal 102 to perform fire extinguishing support operations.

[0069] In some embodiments, the ground fire extinguishing equipment 103 may perform fire extinguishing support operations after receiving the call information sent by the control terminal 102. In practice, the ground fire extinguishing equipment 103 may be a fire truck.

[0070] Among them, multiple strategies also include return-to-home strategies; return-to-home strategies have different logics for different types of drones; For the slave aircraft, the fire extinguishing material consumption information is extracted from the fire extinguishing ammunition information table. If the fire extinguishing material consumption information indicates that all fire extinguishing materials are consumed or the remaining fire extinguishing materials do not match the fire extinguishing operation, the slave aircraft returns to the preset return point. For the first backup aircraft, extract fire extinguishing material consumption information from the fire extinguishing ammunition information table. If the fire extinguishing material consumption information indicates that all fire extinguishing materials have been consumed or the remaining fire extinguishing materials do not match the fire extinguishing operation, obtain the real-time power of the first backup aircraft. If the real-time power of the first backup aircraft is greater than the first backup power threshold, obtain the first actual return distance. If the first actual return distance is greater than the first return distance threshold, and the real-time power of the mother aircraft is greater than the continuous inspection power threshold, return to the preset return point. For the second candidate aircraft, extract the fire extinguishing material consumption information from the fire extinguishing ammunition information table. If the fire extinguishing material consumption information indicates that all fire extinguishing materials have been consumed or the remaining fire extinguishing materials do not match the fire extinguishing operation, obtain the real-time power of the second candidate aircraft. If the real-time power of the second candidate aircraft is greater than the second candidate power threshold, obtain the second actual return distance. If the second actual return distance is less than the second return distance threshold and the real-time power of the mother aircraft is greater than the continuous inspection power threshold, return to the preset return point. For the mother aircraft, obtain the real-time coordinates of the mother aircraft's drone, and determine the minimum power for return based on the preset return point; determine the return power threshold based on the minimum power for return and the floating power; compare the return power threshold with the election power threshold. If the return power threshold is greater than the election power threshold, update the election power threshold to the return power threshold.

[0071] In some embodiments, the plurality of strategies further includes a return strategy, and the return strategy has different logics in the slave machine, the first alternative machine, the second alternative machine, and the mother machine.

[0072] For the slave unit, its control chip retrieves information about fire extinguishing material consumption from the ammunition information table. When all the fire extinguishing materials are consumed, or the remaining fire extinguishing materials do not match the current firefighting operation, the unit returns directly to the preset home point. This preset home point can be a safe landing area set at the control station or at the forest boundary.

[0073] For the first alternative machine, the control chip of the first alternative machine extracts the fire extinguishing material consumption information from the fire extinguishing ammunition information table. When the fire extinguishing material is completely consumed, or the remaining fire extinguishing material does not match the current fire extinguishing operation, the real-time power of the first alternative machine is obtained, and the real-time power of the first alternative machine is compared with the first alternative power threshold. If the real-time power of the first alternative machine is greater than the first alternative power threshold, the corresponding drone real-time coordinates are obtained through the satellite positioning sensor 207. Subsequently, according to the real-time coordinates of the drone and the return point coordinates corresponding to the preset return point, the first actual return distance is calculated based on the Euclidean distance, and it is determined whether the first actual return distance is greater than the first return distance threshold. If so, further communication is performed with the mother machine to obtain the real-time power of the mother machine and determine whether the real-time power of the mother machine is greater than the continuous inspection power threshold. If so, it can return to the preset return point. Among them, the first alternative power threshold can be 50%, the first return distance threshold can be 10km, and the continuous inspection power threshold can be 70%.

[0074] For the second alternative machine, the control chip of the second alternative machine extracts the fire extinguishing material consumption information from the fire extinguishing ammunition information table. When the fire extinguishing material is completely consumed, or the remaining fire extinguishing material does not match the current fire extinguishing operation, the real-time power of the second alternative machine is obtained, and the real-time power of the second alternative machine is compared with the second alternative power threshold. If the real-time power of the second alternative machine is greater than the second alternative power threshold, the corresponding drone real-time coordinates are obtained through the satellite positioning sensor 207. Subsequently, based on the real-time coordinates of the drone and the return point coordinates corresponding to the preset return point, the second actual return distance is calculated based on the Euclidean distance, and it is determined whether the second actual return distance is less than the second return distance threshold. If so, further communication is performed with the mother machine to obtain the real-time power of the mother machine, and it is determined whether the real-time power of the mother machine is greater than the continuous inspection power threshold. If so, it can return to the preset return point. Among them, the second alternative power threshold can be 65%, the second return distance threshold can be 10km, and the continuous inspection power threshold can be 70%.

[0075] For the mothership, the mothership control chip 205 first obtains the real-time coordinates of the corresponding drone through the satellite positioning sensor 207. It then calculates the actual return distance based on the Euclidean distance and the coordinates of the return point corresponding to the preset return point. It then determines the minimum return power requirement based on the actual return distance and the drone's preset speed. Next, it sums the minimum return power requirement and the floating power requirement to obtain the return power threshold. This is then compared with the election power threshold. If the return power threshold is greater than the election power threshold, the election power threshold is updated to the return power threshold, thereby preventing the mothership from failing to return due to insufficient power.

[0076] The preset speed of the drone can be 20m / s. The election power threshold can be 60%. The minimum return power can be 60%, and the floating power can be 3%, so the return power threshold can be 63%. Furthermore, when the master drone's power drops to 63%, the master drone election strategy is triggered in advance, and the drone returns home.

[0077] The floating capacity is determined by the following steps: Obtain a temporary home point between the preset home point and the operating area, and generate a temporary home point information table; obtain the real-time coordinates of the drone of the mother aircraft, and select the temporary home point closest to the real-time coordinates of the drone of the mother aircraft from the temporary home point information table, and determine it as the actual home point; obtain the power consumption between the real-time coordinates of the drone of the mother aircraft and the actual home point, and determine it as the first floating power; Obtain the real-time coordinates of the target candidate drone, and determine the second floating power according to the real-time coordinates of the drone of the mother aircraft; A floating power is generated according to the first floating power and the second floating power.

[0078] In some embodiments, the float charge may be determined by the following steps: When the mother machine is on patrol, the mother machine control chip 205 will broadcast signals according to the preset broadcast frequency, and generate a temporary return point information table for the received response information, which includes the temporary return point coordinates. At the same time, it will also communicate with the control terminal 102. If the temporary return point information is received, it will be summarized in the temporary return point information table. Subsequently, the corresponding drone real-time coordinates are obtained through the satellite positioning sensor 207, and the temporary return point closest to the mother machine is filtered out from the temporary return point information table and determined as the actual return point. Then, based on the real-time coordinates of the drone and the temporary return point coordinates corresponding to the actual return point, the actual return distance is calculated through the Euclidean distance, and then the power consumption between the mother machine and the actual return point is calculated through the preset movement speed of the drone, and determined as the first floating power.

[0079] Then, the mother machine selects the real-time coordinates of the drones corresponding to the qualified alternative machines from the alternative machine information table, and calculates the actual distance of the alternative machine through Euclidean distance based on the drone coordinates of the mother machine and the alternative machine. Then, the second floating power is calculated by combining the actual distance of the alternative machine and the preset movement speed of the drone.

[0080] Finally, the first floating electric quantity and the second floating electric quantity are summed to obtain the floating electric quantity.

[0081] The preset broadcast frequency can be every 10 minutes. Signal broadcasting refers to the drone sending signals at a predetermined frequency to devices within a specific range so that the devices can receive relevant information or respond to the signal. A temporary return point is a temporary return point set by the drone based on mission requirements or environmental changes.

[0082] These embodiments enhance the command stability and operational continuity of the drone firefighting system. Specifically, both the master and backup drones implement a master election strategy. In the event of a master drone failure, an active or passive election mechanism quickly determines a new command drone, preventing command interruption. The drone monitoring strategy monitors drone status in real time, triggering elections in advance to mitigate the impact of sudden failures. The backup drone information table is updated at a synchronized frequency, ensuring accurate and efficient election decisions. These optimizations enable drone swarms to maintain coordinated operations even in emergencies, improving firefighting efficiency and stability.

[0083] In some embodiments, in order to further solve the third technical problem described in the background technology section, namely, "existing drone forest fire extinguishing systems usually rely on direct communication between the mother machine and the control terminal for data synchronization in fire extinguishing operations over long distances or in complex terrain. However, under the changing environment of the fire scene, the communication distance between the mother machine and the control terminal may exceed the preset distance threshold, resulting in communication interruption or delay, which in turn affects the execution and scheduling of fire extinguishing operations", in some embodiments of the present invention, the multiple strategies of the mother machine also include information synchronization strategies.

[0084] Information synchronization strategies include direct synchronization strategy and relay synchronization strategy; direct synchronization strategy is: the master machine synchronizes communication with the control terminal according to the preset synchronization frequency; Obtain the real-time coordinates of the drone on the motherboard and the real-time coordinates of the terminal on the control terminal, and determine the real-time distance based on the real-time coordinates of the drone on the motherboard and the terminal on the control terminal; determine whether the real-time distance is less than a preset communication distance threshold; if not, the motherboard sends a second relay sub-machine call message to the control terminal and obtains the drone number of the second relay sub-machine; If the mother machine does not receive the call response signal from the control terminal within the call response time threshold, the relay synchronization strategy is triggered; the relay synchronization strategy is: obtain the real-time coordinates of the control terminal; from the sub-machine information table, filter the sub-machine that is farthest from the mother machine and closest to the control terminal, and determine it as the first relay sub-machine; the mother machine transmits the synchronization data to the first relay sub-machine, so that the first relay sub-machine broadcasts, and determines whether it can receive the synchronization success signal returned by the control terminal through the first relay sub-machine; if not, transmit the drone number of the second relay sub-machine to the first relay sub-machine, and synchronize information with the control terminal through the second relay sub-machine.

[0085] In some embodiments, the multiple strategies of the master machine also include information synchronization strategies, including direct synchronization strategies and relay synchronization strategies. The direct synchronization strategy involves the master machine synchronously communicating with the control terminal 102 via a preset synchronization frequency to transmit fire information to the control terminal 102. When no fire occurs, the preset synchronization frequency may be once every five minutes; when a fire occurs, the preset synchronization frequency may be once every minute.

[0086] In practice, the master terminal can obtain the real-time coordinates of the control terminal 102 from historical communication records between the master terminal and the control terminal 102. Furthermore, based on a preset ranging frequency, it can obtain the real-time coordinates of the drone from the satellite positioning sensor 207. It then performs a difference calculation between the real-time coordinates of the drone and the real-time coordinates of the terminal, and calculates the real-time distance between the two using Euclidean distance based on this difference. It then determines whether the real-time distance is less than a preset communication distance threshold. If the real-time distance is greater than or equal to the preset communication distance threshold, it sends a second relay sub-unit call message to the control terminal 102, obtains the drone number corresponding to the called second relay sub-unit from the control terminal 102, and determines it as the second relay sub-unit number. The maximum communication distance of the drone can be 20 km, and the preset communication distance threshold can be 16 km. Euclidean distance refers to the straight-line distance between two points in space, calculated as the square root of the sum of the squares of the difference between the coordinates of the two points.

[0087] In practice, if the master terminal still hasn't received a call response signal from the control terminal 102 within the call response time threshold, the relay synchronization strategy is triggered. Specifically, the master terminal can obtain the real-time coordinates of the control terminal 102 from the historical communication records between the master terminal and the control terminal 102. It then selects the slave terminal farthest from the master terminal and closest to the control terminal 102 from a pre-built slave terminal information table and identifies it as the first relay slave terminal. The master terminal then transmits synchronization data to the first relay slave terminal, enabling the first relay slave terminal to broadcast the signal. The master terminal then determines whether it has received a synchronization success signal from the control terminal 102. If not, it further transmits the second relay slave terminal number to the first relay slave terminal, establishing a communication connection between the first and second relay slave terminals. The first relay slave terminal then transmits synchronization data to the second relay slave terminal, which then synchronizes information with the control terminal 102. When establishing a communication connection with a backup or slave, the master unit records the corresponding drone number and real-time coordinates to generate a slave information table. During subsequent communication, the master unit updates the real-time coordinates of the backup or slave unit. The call response time threshold can be 60 seconds. The synchronized data can be the aircraft information table.

[0088] These embodiments enhance the data synchronization stability of drone firefighting systems in complex terrain and long-distance environments. Specifically, the master unit prioritizes direct synchronization for data exchange with the control terminal. When the communication distance exceeds a threshold or no response is received, a relay synchronization strategy is automatically triggered, selecting appropriate slave units as relay nodes and forwarding synchronized data step by step to ensure uninterrupted communication. This improves the communication reliability of drone swarms in complex environments, ensuring smooth execution and precise scheduling of firefighting operations.

[0089] The above descriptions are merely some preferred embodiments of the present invention and illustrate the underlying technical principles. Those skilled in the art should understand that the scope of the present invention is not limited to technical solutions formed by specific combinations of the aforementioned technical features. It also encompasses other technical solutions formed by any combination of the aforementioned technical features or their equivalents, without departing from the aforementioned inventive concept. For example, a technical solution formed by replacing the aforementioned features with (but not limited to) technical features with similar functions disclosed in this invention.

Claims

1. A UAV forest fire extinguishing bomb automatic delivery system based on visual recognition, characterized in that: include: Multiple drones, each of the multiple drones having a corresponding drone number and a fire extinguishing ammunition information table, and being equipped with multiple sensing devices, a fire extinguishing ammunition mounting device, and an external mounting slot; the multiple drones include a mother drone, at least one backup drone, and at least one daughter drone; the external mounting slot, the fire extinguishing ammunition mounting device, and the multiple sensing devices are fixedly connected to the drones; The plurality of sensing devices include thermal imaging sensors, cameras, and satellite positioning sensors; The fire extinguishing bomb mounting device is used to mount and release the fire extinguishing bomb; The external hanging slot has a corresponding loading type, and the external hanging slot is used to load spare batteries or fire extinguishing materials; The motherboard control chip provided in the core cabin of the motherboard is used to generate fire information and a first fire extinguishing plan based on the thermal imaging information corresponding to the thermal imaging sensor, the image information corresponding to the camera, and the positioning information corresponding to the satellite positioning sensor, and execute the first fire extinguishing plan; generating a first fire extinguishing result through the plurality of sensing devices; If the first fire extinguishing result indicates that the fire has not been extinguished, a second fire extinguishing plan is generated based on the fire information, the loading type and the drone number.

2. The automatic delivery system of UAV forest fire extinguishing bombs based on visual recognition according to claim 1 is characterized in that: The alternative machine is the first alternative machine or the second alternative machine, the external hanging slot includes a first external hanging slot and a second external hanging slot, and the loading type is one of the following: the first loading type, the second loading type or the third loading type; the external hanging slot is also provided with a ejection device; if the first external hanging slot is loaded with a spare battery and the second external hanging slot is loaded with a spare battery, the corresponding UAV is determined to be the first alternative machine; if the first external hanging slot is loaded with spare fire extinguishing materials and the second external hanging slot is loaded with spare fire extinguishing materials, the corresponding UAV is determined to be a sub-machine; if the first external hanging slot is loaded with a spare battery and the second external hanging slot is loaded with spare fire extinguishing materials, or the first external hanging slot is loaded with spare fire extinguishing materials and the second external hanging slot is loaded with a spare battery, the corresponding UAV is determined to be the second alternative machine.

3. The automatic delivery system of forest fire extinguishing bombs by UAV based on visual recognition according to claim 2 is characterized in that: The fire information includes the drone number, the drone's real-time coordinates, the combustion state, and the combustion type. The first fire extinguishing plan is determined by the following steps: Determine an abnormal temperature area based on the thermal imaging information, and select an area with the highest temperature from the abnormal temperature area to determine it as a delivery area; determining the real-time coordinates of the drone based on the positioning information; determining the vegetation type based on the real-time coordinates of the drone and a preset regional vegetation information table; determining the combustion type and the combustion material type based on the vegetation type and the image information; determining the release ammunition number based on the combustion type, the combustion material type, and the fire extinguishing ammunition information table; and generating a first fire extinguishing plan based on the release ammunition number and the release area; and The first fire extinguishing result is determined by the following steps: Acquiring first thermal imaging information through the multiple sensing devices, and determining whether the highest temperature in the temperature abnormality area in the first thermal imaging information is lower than a preset safety temperature threshold; if not, determining the first fire extinguishing result as unextinguished; If yes, acquiring first image information and second image information respectively according to a preset image acquisition interval; determining a first smoke ratio according to the first image information; determining a second smoke ratio according to the second image information; determining a smoke change ratio according to the first smoke ratio and the second smoke ratio; If the smoke change ratio is greater than a preset smoke change ratio threshold, the first fire extinguishing result is determined to be extinguished.

4. The automatic delivery system of forest fire extinguishing bombs by UAV based on visual recognition according to claim 3 is characterized in that: The second fire extinguishing plan includes one of the following: a drone support plan, a ground fire extinguishing equipment support plan, or a collaborative support plan; the second fire extinguishing plan is determined by the following steps: Acquire second thermal imaging information and third image information through the multiple sensing devices; determine a fire status based on the second thermal imaging information and the third image information; if the fire status is characterized as small, generate a drone support plan; if the fire status is characterized as large, obtain a distance between the drone's real-time coordinates and a forest boundary based on the drone's real-time coordinates to obtain a forest boundary distance; compare the forest boundary distance with a boundary distance threshold to obtain a distance comparison result; If the distance comparison result indicates that the distances are relatively close, a ground fire extinguishing equipment support plan is generated; if the distance comparison result indicates that the distances are relatively far, a collaborative support plan is generated.

5. The automatic delivery system of forest fire extinguishing bombs by UAV based on visual recognition according to claim 2 is characterized in that: The control chip is deployed with multiple strategies, and the multiple strategies include at least one of the following: a body monitoring strategy, an inspection strategy, a fire extinguishing bomb delivery strategy, and a sub-machine scheduling strategy; The aircraft monitoring strategy is as follows: during the operation of the UAV, the aircraft is monitored according to a preset monitoring frequency, including power information, sensor operation information, hanging slot information, and fire bomb information, and is recorded in the aircraft information table; if the real-time power in the power information is less than or equal to the selected power threshold, it is determined that the power is too low; The inspection strategy is: carry out inspections according to the preset inspection plan. If no abnormalities are found, the aircraft will return to the destination after the inspection. The fire extinguishing bomb delivery strategy is as follows: according to the combustion type and the type of the burning material, the fire extinguishing bomb delivery order is determined from the fire extinguishing ammunition information table, and the fire extinguishing bombs are released through the fire extinguishing bomb mounting device; it is determined whether the external hanging slot carries fire extinguishing materials that match the fire information, and if so, the fire extinguishing materials are released through the ejection device; The deployment result evaluation strategy is as follows: generating multiple evaluation results based on the multiple sensors and a preset evaluation frequency; generating multiple indicator change trend graphs based on the multiple evaluation results; if any indicator change trend graph in the multiple indicator change trend graphs indicates an indicator increase, determining the deployment result as not extinguished; The sub-machine scheduling strategy is as follows: if the delivery result indicates that the fire has not been extinguished, obtain the real-time coordinates of the drone, and match at least one drone number from the preset drone fire extinguishing configuration table according to the combustion type and the type of burning material; generate sub-machine scheduling information based on the at least one drone number and the real-time coordinates of the drone.

6. The automatic delivery system of forest fire extinguishing bombs by UAV based on visual recognition according to claim 5 is characterized in that: The multiple strategies deployed in the control chips of the master machine, the first candidate machine, and the second candidate machine further include a master machine election strategy; The master machine election strategy includes an active election strategy and a passive election strategy; The control chip of the master machine activates the active election strategy, and the control chips of the first candidate machine and the second candidate machine activate the passive election strategy; The active election is as follows: obtaining an aircraft information table and obtaining aircraft monitoring information; if any of the aircraft monitoring information is abnormal, the active election is triggered, and the drone number corresponding to the target candidate aircraft is determined from the candidate aircraft information table; the candidate aircraft information table is determined by the following steps: Obtaining the drone numbers, power information, mission status, and location information corresponding to the first candidate drone and the second candidate drone to generate a candidate drone information table; synchronizing the candidate drone information table to the candidate drones at a preset synchronization frequency; The passive election is as follows: if any of the alternative machines has not received a response signal from the master machine after a response time threshold, the number of the remaining alternative machines that have not received a response signal from the master machine is counted; if the number is greater than or equal to the non-response threshold, whether the candidate machine qualifies as a master machine is determined based on the alternative machine information table. If so, the candidate machine is bootstrapped as the target alternative machine and broadcast.

7. The automatic delivery system of forest fire extinguishing bombs by UAV based on visual recognition according to claim 2 is characterized in that: Also includes: A control terminal is used to receive and process feedback information from the multiple drones and send call information to ground fire-fighting equipment.

8. The automatic delivery system of forest fire extinguishing bombs by UAV based on visual recognition according to claim 7 is characterized in that: Also includes: Ground fire extinguishing equipment, the ground fire extinguishing equipment is used to receive the call information sent by the control terminal to perform fire extinguishing support operations.

9. The automatic delivery system of forest fire extinguishing bombs by UAV based on visual recognition according to claim 6 is characterized in that: The multiple strategies also include a return strategy; the return strategy has different logics for different types of drones; For the slave aircraft, extract the fire extinguishing material consumption information from the fire extinguishing ammunition information table. If the fire extinguishing material consumption information indicates that all fire extinguishing materials are consumed or the remaining fire extinguishing materials do not match the fire extinguishing operation, return to the preset return point; For the first backup aircraft, extract fire extinguishing material consumption information from the fire extinguishing ammunition information table. If the fire extinguishing material consumption information indicates that all fire extinguishing materials have been consumed or the remaining fire extinguishing materials do not match the fire extinguishing operation, obtain the real-time power of the first backup aircraft. If the real-time power of the first backup aircraft is greater than the first backup power threshold, obtain the first actual return distance. If the first actual return distance is greater than the first return distance threshold, and the real-time power of the mother aircraft is greater than the continuous inspection power threshold, return to the preset return point. For the second candidate aircraft, extract fire extinguishing material consumption information from the fire extinguishing ammunition information table. If the fire extinguishing material consumption information indicates that all fire extinguishing materials have been consumed or the remaining fire extinguishing materials do not match the fire extinguishing operation, obtain the real-time power of the second candidate aircraft. If the real-time power of the second candidate aircraft is greater than the second candidate power threshold, obtain the second actual return distance. If the second actual return distance is less than the second return distance threshold and the real-time power of the mother aircraft is greater than the continuous inspection power threshold, return to the preset return point. For the mother aircraft, obtain the real-time coordinates of the mother aircraft's drone and determine the minimum power required for returning to the home point based on the preset return point; determine the return power threshold based on the minimum power required for returning to the home point and the floating power; The return power threshold is compared with the election power threshold. If the return power threshold is greater than the election power threshold, the election power threshold is updated to the return power threshold.

10. The automatic delivery system of forest fire extinguishing bombs by UAV based on visual recognition according to claim 9 is characterized in that: The floating capacity is determined by the following steps: Obtaining a temporary home point between the preset home point and the operating area, and generating a temporary home point information table; obtaining the real-time coordinates of the drone of the mother aircraft, and selecting the temporary home point closest to the real-time coordinates of the drone of the mother aircraft from the temporary home point information table, and determining it as the actual home point; obtaining the power consumption between the real-time coordinates of the drone of the mother aircraft and the actual home point, and determining it as the first floating power; Obtaining the real-time coordinates of the drone of the target candidate aircraft, and determining a second floating power according to the real-time coordinates of the drone of the mother aircraft; A floating power is generated according to the first floating power and the second floating power.

Images