Air-ground cooperative wetland air patrol method and device and electronic equipment

Abstract

The invention provides an air-ground cooperative wetland air patrol method and device and electronic equipment, and the method comprises the steps: making a patrol flight plan based on an operation region and a flight limiting region defined by an operation control management center; daily first flight confirmation is carried out, the unmanned aerial vehicle is controlled to take off under the condition that the flight control right is obtained, and automatic patrol is carried out according to the patrol flight plan. In the automatic patrol process of the unmanned aerial vehicle, image video data are collected in real time, the operation condition is monitored, and the real-time equipment state and alarm information of the unmanned aerial vehicle are obtained. Analyzing according to the image video data to obtain an inspection result, and generating an alarm signal when an abnormal condition exists in the inspection result. And judging whether the unmanned aerial vehicle has a fault based on the real-time state of the equipment and the alarm information, and controlling the unmanned aerial vehicle according to a preset fault handling strategy when the unmanned aerial vehicle has the fault. According to the method, an air-ground cooperation technology and an intelligent technology are combined, the wetland inspection efficiency is effectively improved, and intelligent management of the wetland protection area is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic information and aviation industry, and in particular to an air-ground coordinated wetland aerial patrol method, device and electronic equipment. Background Art

[0002] With the increasing awareness of ecological environment protection and the advancement of science and technology, wetland inspections are needed to protect the ecological environment of wetlands. The traditional manual inspection method has gradually failed to meet the current real-time and rapid needs.

[0003] Drone technology, with its unique aerial perspective, flexible control performance and efficient data processing capabilities, provides a new solution for wetland inspection. The high-definition cameras and advanced sensors carried by drones can monitor wetlands in real time, detect environmental changes in a timely manner, and effectively prevent and respond to ecological problems. Summary of the invention

[0004] The present invention provides an air-ground coordinated wetland aerial patrol method, device and electronic equipment, which are used to integrate AI technology with wetland protection scenarios, and achieve more efficient and accurate wetland monitoring and protection through a closed data loop. The technical solutions proposed by the present invention are as follows: In a first aspect, the present invention provides an air-ground coordinated wetland aerial patrol method, comprising: Formulate patrol flight plans based on the operation areas and restricted flight areas defined by the Operation and Control Management Center; Conduct daily first flight confirmation, control the drone to take off after obtaining flight control rights, and conduct automatic patrols according to the patrol flight plan; During the automatic patrol process of the drone, real-time image and video data are collected and the operation status is monitored to obtain the real-time status and alarm information of the drone equipment; Analyze the image and video data to obtain inspection results, and generate an alarm signal when there is an abnormality in the inspection results; Based on the real-time status of the device and the alarm information, it is determined whether the drone has a fault. If the drone has a fault, the drone is controlled according to a pre-set fault handling strategy.

[0005] Optionally, the image and video data are analyzed in the following manner: The image video data is detected by using a pre-trained first target detection model to identify whether there are any violators in the wetland protection area, and a virtual identification is generated for the violators; The image video data is detected by a pre-trained second target detection model to identify whether there is illegal use of fire or fire hazard in the wetland protection area; The image video data is detected through a pre-trained third target detection model to identify whether there are wild animals in the wetland reserve, and a tracking algorithm is used to track the wild animals in continuous images. The activity trajectory of the wild animals is analyzed in combination with the timestamp and geographic location information of the image.

[0006] Optionally, the first flight of the day can be confirmed by: The environmental monitoring system installed on the drone nest collects the environmental information of the drone before takeoff; Acquiring meteorological information, and performing meteorological environment analysis on the environmental information and meteorological information to determine whether the conditions for takeoff are met; When the conditions for takeoff are met, apply for flight control.

[0007] Based on the operation area and restricted flight area defined by the Operation and Control Management Center, formulate a patrol flight plan, including: The Operation and Control Management Center defines the operation area and restricted flight area based on the geographical location, ecological and environmental characteristics of the wetland reserve and relevant laws and regulations; Based on the definitions of the operating area and the restricted flight area, and combined with the patrol missions of the wetland conservation area, a patrol flight plan is formulated; the patrol information in the patrol flight plan includes patrol waypoint routes, patrol speed, patrol altitude, patrol frequency, maneuvering requirements of the onboard camera, and fault handling strategies.

[0008] Optionally, when a fault occurs in the drone, controlling the drone according to a preset fault handling strategy includes: Obtain weather information and airspace conditions, predict the flight risks faced by drones based on airspace conditions, patrol flight plans and weather information, and resolve conflicts according to preset flight rules when there are flight risks; Obtaining operational data on the drone’s flight systems when deconfliction is not possible; Determine whether the drone can return based on weather information and operational data; In the case that the drone cannot return, the target alternate landing field closest to the current drone position is selected from the preset alternate landing fields of different distance levels, and the drone is controlled to make an alternate landing to the target alternate landing field.

[0009] Optionally, when a fault occurs in the drone, controlling the drone according to a preset fault handling strategy further includes: If the drone cannot make an emergency landing, control the drone to send a crash warning to the ground operation and control management center, and update the flight dynamic information to the ground operation and control management center; In the absence of flight dynamics information, the crash dynamics model is used to identify the crash site and alert the integrated patrol personnel.

[0010] Optionally, resolving the conflict according to preset flight rules includes: Control one or more drones to rise or descend to a preset altitude; the preset altitude is determined according to airspace conditions and performance parameters of the drones; When two drones are flying in the same direction and the distance between them is gradually shortened, a command is sent to the drones to change their flight directions, so that one of the drones deflects while the other continues to maintain its original direction; The flight speed of the drone is adjusted to change the time it takes for the drone to reach a preset location. In a second aspect, the present invention also provides an air-ground coordinated wetland aerial patrol device, comprising: An acquisition module is used to acquire image and video data collected by the UAV when executing the patrol flight plan; The reasoning module is used to analyze the image and video data to identify whether there are illegal personnel, illegal use of fire, fire hazards or wild animals in the wetland reserve, and obtain inspection results; The feedback module is used to feed back the inspection results to the ground operation control management center and generate an alarm signal when there are abnormalities in the inspection results.

[0011] In a third aspect, the present invention further provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the air-ground collaborative wetland aerial patrol method as described in the first aspect above is implemented.

[0012] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the air-ground collaborative wetland aerial patrol method as described in the first aspect above.

[0013] In a fifth aspect, the present invention further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the air-ground collaborative wetland aerial patrol method as described in the first aspect above.

[0014] Based on the above technical solution, the beneficial effects of the present invention compared with the prior art are as follows: The air-ground collaborative wetland aerial patrol method, device and electronic equipment provided by the present invention can further reduce flight risks and ensure the safety of drones by applying for flight control rights while ensuring that the take-off conditions meet the requirements. When the flight control rights are obtained, the drone is controlled to take off and automatically patrols according to the patrol flight plan. The collected image and video data are identified and analyzed to obtain the inspection results, so as to achieve the purpose of detecting and alarming wetland wild animals, potential fires, and illegal intrusions by personnel, realize poaching inspections, fire alarms and animal monitoring to promote the integration of AI technology and wetland protection scenarios, and realize more efficient and accurate wetland monitoring and protection through data closed loops. At the same time, the ground operation and control management center can monitor the status of the drone in real time to ensure that it operates under safe conditions. The method combines air-ground collaboration and intelligent technology, effectively improves the efficiency of wetland inspections, and realizes the intelligent management of wetland protection areas. Other features and advantages of the present invention will be described in the subsequent description, and part of them will become obvious from the description, or be understood by implementing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures specifically pointed out in the description, claims and drawings.

[0015] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 It is the overall architecture diagram of the air-ground collaborative wetland aerial patrol system provided by the present invention.

[0018] Figure 2 It is a schematic flow chart of the air-ground collaborative wetland aerial patrol method provided by the present invention.

[0019] Figure 3 It is a schematic diagram of the air-ground coordinated wetland aerial patrol operation process provided by the present invention.

[0020] Figure 4 It is a structural schematic diagram of the air-ground coordinated wetland aerial patrol device provided by the present invention.

[0021] Figure 5 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] Combine the following Figure 1-Figure 5 The present invention describes the air-ground collaborative wetland aerial patrol method, device and electronic equipment.

[0024] The air-ground coordinated wetland aerial patrol method provided by the present invention can be applied to an air-ground coordinated wetland aerial patrol system. The system is an integrated wetland comprehensive drone automatic patrol system that can simultaneously perform wetland drone automatic patrols and achieve multiple functions. Figure 1 As shown in FIG, the system architecture of the air-ground collaborative wetland aerial patrol system is divided into a hardware layer, a service layer, and a user layer. The hardware layer includes drones, and the user layer is the ground operation control management center.

[0025] The system works in coordination with the existing signal system to jointly realize the role of drones and ground operation and control management centers in patrolling wetlands, and realizes multiple functions such as behavior identification and expulsion of violators, wetland fire prevention patrols, and counting of wetland wild animals (such as elk).

[0026] (1) The hardware layer consists of drones and drone nests. The drones are equipped with a shooting sensor system, a flight system, and a communication system. The drone nests are equipped with a communication system, a storage and charging system, an automatic take-off and landing system, and an environmental monitoring system.

[0027] ① UAVs can automatically execute flight plans and video collection tasks, quickly and comprehensively obtain various information about wetlands, and provide rich data support for wetland monitoring and research. At the same time, they can also monitor the real-time status of wetlands, detect abnormal situations such as fires in a timely manner, and issue early warnings. The hardware layer connects to the service layer through the public network and transmits information to the service layer.

[0028] ② The above-mentioned drone nest is set up outside the wetland. The drone nest can realize the functions of automatic operation, take-off and landing environmental monitoring. The drone nest has the functions of automatic charging, storage and maintenance of drones. The supporting drones can realize the functions of efficient data collection, monitoring and early warning.

[0029] (2) The service layer mainly provides public services, including push and pull services, GIS map services, meteorological services, AI model services, video and image storage services, and real-time reasoning services.

[0030] The service layer can implement information and data interaction with the hardware layer and user layer through push and pull stream services: ① The service layer sends GIS map information and weather information to the hardware layer through streaming, maintains real-time communication with the drone, sends control instructions to the drone, and ensures the flight of the drone. The hardware layer receives information through streaming, including the status information of the drone, and collects image and video data taken by the drone, which is stored in the video and image storage service of the service layer and analyzed and identified through the AI ​​model service.

[0031] ② The service layer can pass information to the user layer through push streaming, and the inspection results after analysis are promptly fed back to the management personnel of the ground operation control management center, and the instructions of the user layer are received through pull streaming.

[0032] (3) The user layer is the ground operation control management center. The applications of the ground operation control management center include mobile work orders and WEB management terminals, including equipment management, flight planning, task scheduling, drone control, abnormal alarm, work order management, AI perception and other functions.

[0033] Wetland managers can publish inspection tasks or adjust the drone inspection time or route through the user-level WEB management terminal according to actual needs. During the inspection, users can view the progress and status of the inspection tasks in real time, and understand the drone's inspection trajectory, its own status, and data collection. When the system detects abnormal situations such as fire hazards in wetlands and illegal entry of personnel, the user level will promptly issue an alarm prompt so that users can quickly take alarm measures to remind managers and enhance wetland management.

[0034] Reference Figure 2 and Figure 3 As shown, the air-ground coordinated wetland aerial patrol method provided by the present invention includes the following: Step S101: formulate a patrol flight plan based on the operation area and restricted flight area defined by the operation control management center.

[0035] As the command center of the entire system, the Operation and Control Management Center is responsible for defining the operation area and the restricted flight area. The operation area is a specific area where drones need to perform patrol missions, which is usually determined based on the scope of the wetland protection area, the location of important ecological areas, and the key elements that need to be monitored. The restricted flight area is an area where drones are prohibited or restricted from flying based on safety, regulations or environmental factors, such as densely populated areas and around military facilities.

[0036] Based on the definition of the operation area and the restricted flight area, the Operation Control Management Center will formulate a detailed patrol flight plan. This includes the flight route, flight altitude, speed, waypoint setting, patrol schedule, etc. The flight plan is designed to ensure that the drone can cover the entire operation area efficiently and safely while avoiding entering the restricted flight area.

[0037] Step S102: Confirm the first flight of the day. When the flight control right is obtained, control the UAV to take off and perform automatic patrol according to the patrol flight plan.

[0038] Before performing daily patrol missions, drone operators must apply for flight control rights from the Operation and Control Management Center or relevant airspace management departments to ensure that the drone takes off under legal and compliant conditions. After obtaining flight control rights, the operator controls the drone to take off through the ground control station or remote control, and automatically patrols according to the pre-established patrol flight plan. The drone will fly automatically according to the preset routes and parameters without human intervention.

[0039] Step S103: During the automatic patrol of the UAV, real-time image and video data are collected and the monitoring operation status is monitored to obtain the real-time status and alarm information of the UAV equipment.

[0040] During the automatic patrol process, the drone collects images and video data of the wetland reserve in real time through the sensors such as the cameras carried by it. These data are used for subsequent analysis and evaluation. At the same time, the drone will also transmit its equipment status information in real time, such as battery power, flight speed, altitude, temperature, etc., as well as any possible warning information, such as low power, sensor failure, etc. Specifically, the present invention uses a drone that adapts to the complex terrain of the wetland. This type of drone has functions such as intelligent dynamic tracking, high-definition zoom, thermal imaging, etc., and can capture the wetland ecological environment in all directions. According to the specific conditions of the wetland reserve, the flight route and altitude of the drone are formulated to ensure coverage of all key areas, such as water areas, densely vegetated areas, wildlife habitats, etc. During the flight, the drone collects image and video data of the wetland in real time through the sensors such as the high-definition camera and multi-spectral analyzer carried by it, including the ecological status of the wetland, changes in water quality, wildlife activities, etc.

[0041] The data collected by the drone is sent to the control service layer in real time through wireless transmission technology. The service layer receives the image and video data sent by the hardware layer through streaming, and stores it efficiently and securely through the video and picture storage service, which is convenient for subsequent analysis and call.

[0042] The GIS map service of the service layer uses GIS technology to generate detailed maps of the wetland reserve, including information such as terrain, water system, and vegetation distribution, to provide accurate navigation for drone patrols. The meteorological service obtains real-time meteorological information of the wetland reserve, such as wind direction, wind speed, temperature, humidity, etc., to provide safety and decision support for drone flights. GIS map information and meteorological information are sent to the ground operation control management center and the hardware layer through streaming to ensure that all parties can keep abreast of the latest situation. The service layer sends the GIS map information generated by the GIS map service and the meteorological information generated by the meteorological service to the ground operation control management center and the hardware layer through streaming.

[0043] Step S104: Analyze the image and video data to obtain inspection results, and generate an alarm signal when there is an abnormality in the inspection results.

[0044] The ground control station or remote data center receives the image and video data transmitted by the drone and analyzes and processes it. This includes identifying the health of wetland vegetation, water quality, wildlife activities, etc. If the analysis results show abnormal conditions, such as vegetation destruction, water pollution, illegal activities, etc., the ground control station will automatically generate an alarm signal and send relevant information to relevant departments or personnel so that timely measures can be taken. The above-mentioned AI model service uses advanced AI algorithms and models to analyze and process the image and video data collected by drones. It mainly identifies whether there are illegal activities of people (such as illegal fishing, poaching, etc.), illegal use of fire, hidden dangers of fire use, and the activities of wild animals in the wetland protection area. Through the analysis of the AI ​​model, the inspection results are obtained and abnormal areas or targets are marked.

[0045] The inspection results are fed back to the ground operation control management center in real time through streaming so that management personnel can take corresponding measures in a timely manner. The ground operation control management center formulates targeted protection and management plans based on the inspection results to ensure the stability and health of the wetland ecological environment. If the above inspection results contain any illegal activities of personnel (such as illegal fishing, poaching, etc.), illegal use of fire, fire hazards, and any of the wild animals, a corresponding alarm signal will be generated according to the event type. The alarm signal includes at least the following information: Event type: clearly indicate the type of event that triggers the alarm, such as "illegal personnel activity alarm", "fire hazard alarm", "illegal use of fire alarm", "wild animal activity alarm", etc.

[0046] Occurrence time: record the specific time when the event occurred, including date and time.

[0047] Location of occurrence: Provide precise location information of the incident, including latitude and longitude coordinates, map link, specific address description or area identification (such as protected area number, monitoring point name, etc.).

[0048] Description of on-site situation: Briefly describe the situation observed on-site, such as "two persons were found using illegal fishing gear in the prohibited fishing zone", "signs of open fire burning were detected in a forest area", or "infrared cameras captured the appearance of rare wild animals".

[0049] Image or video evidence: Attach images or video screenshots related to the incident so that the recipient can intuitively understand the on-site situation.

[0050] Step S105: determine whether the drone has a fault based on the real-time status of the device and the alarm information, and if the drone has a fault, control the drone according to a preset fault handling strategy.

[0051] Based on the real-time status and warning information of the device, the ground control station can determine whether the drone has a fault. For example, if the battery is too low or the sensor is faulty, the drone may send a warning message. Once the drone is confirmed to have a fault, the ground control station will control the drone according to the pre-set fault handling strategy. This includes emergency landing, returning to a safe place, adjusting flight parameters to mitigate the impact of the fault, etc.

[0052] The air-ground collaborative wetland aerial patrol method provided by the present invention uses the service layer as the middle layer and realizes the management and control of the hardware layer drone flight by the user layer through public network information interaction. At the same time, the image and video data collected by the hardware layer are identified and analyzed through the AI ​​model service to obtain the inspection results, and the inspection results are displayed on the ground operation and control management center client interface of the user layer, thereby achieving the purpose of detecting and alarming wetland wild animals, potential fires, and illegal intrusions by people, realizing poaching inspections, fire alarms and animal monitoring, promoting the integration of AI technology and wetland protection scenarios, and realizing more efficient and accurate wetland monitoring and protection through a data closed loop.

[0053] Through the high-definition cameras, multi-spectral analyzers and other sensors carried by drones, changes in the wetland ecological environment can be monitored in real time, improving monitoring capabilities. Using AI models to analyze and process image and video data can accurately identify violations and potential risks in wetland protection areas, and take timely measures to intervene and protect them, thus realizing intelligent and precise management of the wetland ecological environment and providing strong support for wetland protection and management. Moreover, drone patrols can quickly cover large areas of wetlands, greatly improving patrol efficiency compared to traditional human patrols. It also reduces the input of manpower and material resources and reduces patrol costs.

[0054] The present invention can reduce the dependence on human resources and reduce patrol costs through automatic patrol by unmanned aerial vehicles. This is particularly important for wetland conservation areas with limited resources. When unmanned aerial vehicles conduct aerial patrols in wetland conservation areas, they can prevent personnel from entering dangerous areas and reduce the risk of casualties. At the same time, the ground control station can monitor the status of the unmanned aerial vehicle in real time to ensure that it operates under safe conditions. The method combines air-ground collaboration and intelligent technology to achieve intelligent management of wetland conservation areas. Through functions such as data analysis and alarm signal generation, dynamic monitoring and early warning of wetland conservation areas can be achieved, improving management levels and efficiency. Real-time collected image video data and analysis results can also provide valuable data support for wetland scientific research, which helps to gain a deeper understanding of the changing laws and influencing factors of wetland ecosystems.

[0055] The application scenario design of the air-ground coordinated wetland aerial patrol system of the present invention is mainly divided into two parts: normal patrol and fault handling and utilization. Figure 3 As shown in the figure, comprehensive patrol includes making patrol plans, obtaining daily first flight confirmation, drone takeoff, drone patrol, and drone return. Fault handling includes drone interval deployment risks, drone alternate landing, and drone crash.

[0056] (1) Comprehensive patrol plan design

[0057] 1) Develop a patrol plan

[0058] Users set the patrol waypoint route, patrol speed, patrol altitude, patrol frequency, maneuvering requirements of the onboard camera, and out-of-control or emergency response procedures for each patrol drone through the ground operation and control management center to form a daily patrol flight plan.

[0059] 2) Daily first flight confirmation

[0060] Manually export flight missions, manually apply for / report flight missions on the UOM platform, and wait for flight plan approval.

[0061] 3) Drone takes off

[0062] Before and during the drone takes off, the control environment monitoring system continuously collects environmental information. This information includes but is not limited to wind speed, wind direction, temperature, humidity, air pressure, visibility, and possible obstacle locations. For example, before takeoff (such as 5 minutes before takeoff), the drone nest uses integrated rain gauges, anemometers, thermometers and other sensors to collect environmental information such as ambient temperature and humidity, wind speed, and rainfall, and combines it with weather forecasts to monitor the meteorological environment before and during the drone takes off to determine whether the conditions for takeoff are met.

[0063] In an optional embodiment, the step S101 described above, formulating a patrol flight plan based on the operation area and the restricted flight area defined by the operation control management center, includes: S201. The Operation and Control Management Center defines the operation area and restricted flight area based on the geographical location, ecological and environmental characteristics of the wetland reserve and relevant laws and regulations.

[0064] The Operation and Control Management Center first determines the areas that need to be monitored and protected as the operation areas based on the geographical location and ecological and environmental characteristics of the wetland reserve. These areas usually include key habitats in the wetland ecosystem, rare species activity areas, and ecologically sensitive areas.

[0065] According to relevant laws and regulations and the special management requirements of wetland conservation areas, the Operation Control Management Center will define some restricted flight areas, namely restricted flight areas. These areas may include no-fly zones within wetland conservation areas (such as core conservation areas, breeding grounds for rare species, etc.), height-restricted areas (such as bird migration corridors at the edge of wetlands), warning areas (such as areas with frequent human activities), etc. The purpose of defining restricted flight areas is to protect the wetland ecological environment and avoid interference or harm to wetland organisms caused by flight activities.

[0066] S202. Based on the definitions of the operation area and the restricted flight area, and in combination with the patrol missions of the wetland conservation area, a patrol flight plan is formulated; the patrol information in the patrol flight plan includes patrol waypoint routes, patrol speed, patrol altitude, patrol frequency, maneuvering requirements of the onboard camera, and fault handling strategies.

[0067] The Operation and Control Management Center will set detailed patrol information for the drone based on the specific patrol mission. This information includes but is not limited to patrol waypoint routes (i.e. key locations and paths that the drone needs to fly over), patrol speed (ensuring that the drone flies at a suitable speed to obtain clear images and videos), patrol altitude (determined according to the wetland terrain and obstacles), patrol frequency (determines how long the drone needs to complete a full patrol cycle), maneuver requirements of the onboard camera (such as zoom, rotation, etc.), and out-of-control or emergency handling procedures (ensuring that countermeasures can be taken quickly when the drone encounters a problem). Based on the above patrol information, a detailed patrol flight plan is generated and sent to the drone for execution. The patrol information is set specifically in the following way: According to the definition of the operation area and the restricted flight area, a reasonable patrol waypoint route is planned. The route should cover the key monitoring points in the operation area as much as possible, while avoiding the restricted flight area to ensure the legality and safety of the flight activities.

[0068] The appropriate patrol speed should be determined based on the performance of the aircraft and the specific conditions of the wetland reserve. The speed should take into account both monitoring effect and flight efficiency to ensure that the aircraft can complete the patrol mission within the specified time.

[0069] Under the premise of complying with the regulations of the restricted flying area, the appropriate patrol altitude should be determined according to the flight performance of the aircraft and the monitoring needs. The altitude should be low enough to obtain clear monitoring images, but high enough to avoid disturbing wetland organisms.

[0070] According to the ecological and environmental characteristics of the wetland reserve and the monitoring needs, a reasonable patrol frequency should be determined. The frequency should be selected to be able to timely detect changes and problems in the wetland ecological environment and provide strong support for wetland protection.

[0071] To ensure the monitoring effect, the airborne camera should have high mobility and stability. The camera should be able to flexibly adjust the shooting angle and focal length to meet the needs of different monitoring points. At the same time, the camera should have high image quality and stability to ensure the accuracy and reliability of the monitoring data.

[0072] Develop detailed fault handling strategies, including aircraft failure, camera failure, data transmission failure, etc. In the event of a failure, the emergency handling procedure should be initiated immediately to ensure the safety of the aircraft and onboard equipment, and to resume monitoring work as soon as possible.

[0073] Through reasonable patrol waypoint routes and patrol frequencies, it is possible to ensure that key monitoring points in the wetland protection area are fully covered and improve monitoring efficiency. Strictly abide by the regulations of the restricted flight zone to avoid interference or damage to wetland organisms caused by flight activities, which will help protect the wetland ecological environment. The high maneuverability and stability of the onboard camera can ensure the accuracy and reliability of the monitoring data, providing strong data support for wetland protection. Detailed fault handling strategies can enhance emergency response capabilities, ensure that monitoring work can be quickly restored and losses reduced when a fault occurs. The present invention formulates patrol flight plans based on the operating areas and restricted flight zones defined by the operation and control management center, which can ensure that monitoring work in the wetland protection area is carried out efficiently, safely and accurately, and provide strong support for wetland protection.

[0074] The present invention uses an environmental monitoring system and real-time reasoning services to accurately analyze the meteorological environment, ensuring that the UAV takes off and flies under safe conditions, greatly reducing the risk of flight accidents. According to the specific requirements of the meteorological environment and the patrol mission, a patrol flight plan is tailored for the UAV to ensure that the UAV can efficiently complete the patrol mission and reduce unnecessary flight time and energy consumption. The patrol information can be flexibly adjusted according to actual needs, such as adjusting the patrol speed, height and frequency, etc., to adapt to different wetland environments and protection needs. By formulating out-of-control or emergency response procedures, it is ensured that countermeasures can be taken quickly when the UAV encounters problems, reducing losses and ensuring the safety of personnel and equipment. By integrating the environmental monitoring system, real-time reasoning services and the operation and control management center, the intelligent management of UAV patrols is realized, and the scientific and technological level and management efficiency of wetland protection are improved.

[0075] In an optional embodiment, the daily first flight confirmation in step S102 is performed in the following manner: S301, collecting environmental information of the drone before take-off through the environmental monitoring system installed on the drone nest.

[0076] The environmental monitoring system installed on the drone nest usually includes temperature sensors, humidity sensors, wind speed and direction sensors, air pressure sensors, etc., which are used to comprehensively monitor the environmental conditions before the drone takes off. Before the drone is ready to take off, the environmental monitoring system starts working and collects real-time environmental information of the drone's take-off location, including key parameters such as temperature, humidity, wind speed, wind direction, and air pressure.

[0077] S302: Acquire meteorological information, perform meteorological environment analysis on the environmental information and meteorological information, and determine whether take-off conditions are met.

[0078] Obtain weather information about the take-off location and nearby areas of the drone through the above-mentioned weather services, local weather stations or specialized weather monitoring equipment, including weather conditions (such as sunny, cloudy, rainy, etc.), temperature, humidity, wind speed and direction, etc. Ensure that the weather information obtained is up-to-date and accurate for subsequent weather environment analysis.

[0079] The environmental information collected by the environmental monitoring system is comprehensively analyzed with the meteorological information obtained to evaluate whether the meteorological environment is suitable for the drone to take off. According to the performance parameters and flight requirements of the drone, a series of take-off conditions are set, such as wind speed not exceeding a certain value, temperature within a certain range, no heavy rainfall, etc. Then, the actual meteorological environment parameters are compared with these conditions to determine whether the take-off conditions are met.

[0080] Specifically, the control service layer receives environmental information from the environmental monitoring system by streaming. The real-time reasoning service is used to conduct a comprehensive analysis of the received environmental information and the meteorological information provided by the meteorological service. The real-time reasoning service is a rule engine-based system that can pre-set reasoning rules. According to the analysis results, it is determined whether the current meteorological environment meets the take-off conditions of the drone. This includes evaluating whether the wind speed is within a safe range, whether the visibility is sufficient, and whether there are potential obstacles. For example, if the ambient wind speed is less than 8m / s, the temperature is higher than -20℃, and there is no water immersion inside the nest, the drone meets the take-off conditions. At the same time, the drone takes off as planned for self-inspection. If the above conditions are met, the drone starts to take off and patrol as planned, otherwise it will alarm the comprehensive patrol personnel through the ground operation control management center. After the drone takes off at the patrol height, the flight environment is checked through the onboard visible light camera.

[0081] S303. Apply for flight control rights when conditions for take-off are met.

[0082] After confirming that the conditions for takeoff are met, the drone operator or ground control station applies to the operation and control management center or relevant airspace management department through designated communication channels for flight control rights. After receiving the application, the operation and control management center or relevant airspace management department decides whether to approve the application for flight control rights based on factors such as the current airspace usage and flight plan. After obtaining flight control rights, the drone operator or ground control station makes final preparations for takeoff, including checking the drone status, setting flight parameters, etc., and then controls the drone to take off.

[0083] The present invention can accurately judge whether the meteorological environment is suitable when the UAV takes off through comprehensive analysis of the environmental monitoring system and meteorological information, and avoid flight accidents caused by bad weather. Applying for flight control rights while ensuring that the take-off conditions meet the requirements can further reduce flight risks and ensure the safety of the UAV. The automated environmental monitoring and meteorological information acquisition process reduces the time for manual intervention and judgment, and improves the efficiency of daily first flight confirmation. Rapidly applying for flight control rights when the take-off conditions are met can shorten the take-off preparation time of the UAV and enable it to be put into patrol missions more quickly. On the premise of ensuring safety, flight plans can be arranged more flexibly and the use efficiency of the UAV can be optimized. Moreover, through accurate meteorological environment analysis, it is possible to avoid taking off the UAV under unsuitable weather conditions, thereby saving energy and reducing the wear and tear of the UAV. The above-mentioned daily first flight confirmation process integrates intelligent technologies such as environmental monitoring, meteorological information acquisition, and data analysis, which improves the intelligence level of the entire UAV patrol system.

[0084] In an optional embodiment, the ground operation control management center collects and stores all drone nests, drone operation data, and control instructions. By analyzing the data of the drone nests, the operation index data of the drone nests is obtained, and the data can be used as flight verification data of experimental drones.

[0085] The drones fly along the perimeter of the wetland and key patrol areas as planned. The ground operation and control management center analyzes the information sent back by the camera in real time. The system uses AI algorithms to identify the videos of wild animals sent back and collects and stores the photo data. Through data analysis, the number of wild animals in the park is counted. The system uses AI algorithms to identify illegal entrants. If illegal entrants are found, the integrated operation control manually takes over the drone to lower the altitude to collect evidence and take photos of the illegal entrants, and alerts the integrated patrol personnel for confirmation. The system uses AI algorithms to identify whether there are illegal uses of fire or fire hazards in the protected area. If illegal use of fire is found, the ground operation and control management center automatically lowers the altitude of the drone in the air to collect evidence and take photos of the illegal use of fire, and alerts the integrated patrol personnel for confirmation. If the situation is confirmed to be true, the fire information (location, photos, fire intensity) will be automatically pushed to the fire department.

[0086] The comprehensive patrol personnel monitor the patrol process of the drone and can take over the control of the drone through the ground operation and control management center. The comprehensive patrol personnel manually control the drone to complete the required patrol operation or evidence collection operation. After the comprehensive patrol personnel manually remotely control the drone to complete the required patrol or evidence collection operation, they will return the control to the ground operation and control management center. After the drone returns to the manual takeover point, it will continue to fly on the automatic route.

[0087] The above step S104 of analyzing the image and video data to obtain the inspection result specifically includes: S401, detecting the image video data by using a pre-trained first target detection model to identify whether there are any violators in the wetland protection area, and generating a virtual identification for the violators.

[0088] First, the input image and video data is analyzed frame by frame using a pre-trained first target detection model (such as FasterR-CNN and YOLO based on convolutional neural network CNN). After being trained with a large amount of labeled data, the model can identify violators in the image (such as tourists who enter the protected area without permission, illegal hunters, etc.). In the image taken by the drone, the violators are framed or marked with special symbols and transmitted to the monitoring screen of the ground operation control management center in real time to realize virtual identification for subsequent processing or manual verification. The information of the detected violators, including location, timestamp, etc., is output as the first result. If an illegal entrant is found, the integrated operation control system automatically lowers the altitude of the aerial drone to collect evidence and take photos of the illegal entrant, drives the person away through the onboard loudspeaker, and alerts the ground comprehensive patrol personnel for confirmation. If necessary, law enforcement personnel will go to the scene to deal with it.

[0089] The first target detection model is obtained in the following way: By collecting a large amount of image data of wetland violations, such as images of people illegally hunting, fishing, and destroying vegetation in wetlands, these images are trained using deep learning algorithms, and the target detection algorithm deployed by the data processing unit is the deep learning target detection model YOLO (You Only Look Once) to learn the characteristics of violators, including clothing color, body shape, and tools carried. At the same time, it combines multiple sensor data, such as image data from high-definition cameras and thermal imaging sensor data. Thermal imaging can detect the presence of people at night or in low visibility conditions, making up for the shortcomings of visible light cameras.

[0090] The specific training process is as follows: 1. Image data collection: Regularly collect image data from high-definition camera systems within the wetland reserve, ensuring that the data covers different time periods (including daytime, dusk, and night) and weather conditions (sunny, cloudy, rainy, etc.). Pay special attention to capturing scenes of illegal hunting, fishing, and vegetation destruction. In addition to image data, data from thermal imaging sensors must also be integrated. Thermal imaging sensors can capture infrared radiation generated by human or animal body temperature, thereby detecting targets at night or in low visibility conditions.

[0091] 2. Data Preprocessing

[0092] Image annotation: Manually annotate the collected images to mark the specific location, category (such as illegal hunters and fishermen) and characteristics (such as clothing color, body shape, tools carried, etc.) of the offenders. Remove blurry, repeated or irrelevant images to ensure the quality and diversity of training data. Enhance image data by rotating, scaling, cropping, adding noise, etc. to increase the generalization ability of the model.

[0093] 3. Model selection and training

[0094] Select the YOLO model as the target detection model and use the labeled image data to train the YOLO model. During the training process, the model will learn how to identify the offending persons and features in the image. When training, you also need to consider combining thermal imaging data. You may need to fuse thermal imaging images with visible light images, or train the model to process both types of data at the same time. By adjusting the model parameters and optimizing the algorithm, you can improve the detection accuracy and efficiency of the model.

[0095] 4. Model deployment and testing

[0096] Deploy the trained YOLO model to the data processing unit of the wetland reserve and connect it to the high-definition camera and thermal imaging sensor. The system receives image data from the camera in real time and uses the deployed YOLO model to detect targets. When a violation is detected, the system should immediately trigger an alarm. Perform performance tests on the deployed system to evaluate its detection accuracy, false alarm rate, and missed alarm rate in the actual environment. Collect feedback from wetland managers to understand the performance of the system in actual applications, including false alarms, missed alarms, and suggestions for improvement. Based on feedback and new image data, continuously iterate and optimize the model to improve its detection performance and stability.

[0097] Through the above steps, the system can effectively monitor and identify violations in wetlands.

[0098] S402: Detect the image video data using a pre-trained second target detection model to identify whether there is any illegal use of fire or fire hazard in the wetland conservation area.

[0099] Use the second object detection model to detect image and video data. This model is trained to recognize fire features such as flames and smoke, as well as possible illegal use of fire (such as outdoor barbecues, bonfires, etc.). The model outputs the location, size, timestamp, and other information of the detected illegal use of fire points or fire hazard areas as the second result.

[0100] The above second target detection model is obtained in the following way: By collecting a large amount of image data of smoke and flames, such as images of smoke and fire in forests and wetlands, the detection model is trained using a deep learning algorithm, and the target detection algorithm deployed by the data processing unit is a deep learning target detection model YOLO (You Only Look Once) that learns the characteristics of smoke and flames, including smoke color, flame color, temperature data, etc. If information about smoke or flames is found, the ground operation control management center automatically lowers the altitude of the aerial drone to collect evidence and take photos of illegal fire use, and alerts the ground comprehensive patrol personnel for confirmation, notifying law enforcement personnel to deal with it on the spot.

[0101] The specific training process is as follows: The process involves a highly integrated monitoring system to identify illegal fire use or fire hazards in the wetland reserve and take appropriate countermeasures. The following are the detailed steps of the process: 1. Data collection and annotation

[0102] Image data collection: HD cameras and infrared cameras (for detection at night or in smoky environments) installed in the wetland reserve are used to continuously collect image data containing fire-related features such as smoke and flames. The collected images are manually annotated to mark key features such as the location, size, color, and temperature of smoke and flames. This annotated information will be used to train the deep learning model.

[0103] 2. Deep Learning Model Training

[0104] Train the YOLO model using labeled image data. During the training process, the model will learn how to identify fire features such as smoke and flames in the image. To improve the accuracy of the model, temperature data can be introduced as an additional input feature. The performance of the model can be optimized by adjusting the model parameters, optimizer, learning rate, etc., and using data augmentation techniques such as random cropping, rotation, scaling, etc.

[0105] 3. System deployment and real-time monitoring

[0106] The trained YOLO model is deployed to the ground operation and control management center of the wetland reserve and connected to hardware devices such as cameras and drones. The system receives image data from the camera in real time and uses the deployed YOLO model for target detection. Once smoke or flames are detected, the system will immediately trigger an alarm. Feedback from inspectors and law enforcement officers is collected to understand the performance of the system in actual applications, including false positives, missed positives, and suggestions for improvement. Based on feedback and new image data, the deep learning model is iterated and optimized to improve its detection performance and stability.

[0107] Through the above steps, the system can effectively identify illegal use of fire or fire hazards in wetland protection areas, and automatically trigger response mechanisms to ensure the safety of wetland resources.

[0108] S403, the image video data is detected by a pre-trained third target detection model to identify whether there are wild animals in the wetland reserve. In order to avoid repeated counting, a tracking algorithm is used to track the wild animals in the continuous images. When a wild animal is detected multiple times in the continuous images, it is counted only once. At the same time, the activity trajectory of the wild animal is analyzed in combination with the timestamp and geographic location information of the image.

[0109] Wildlife recognition is performed on image video data through the third target detection model. This model can identify the types and numbers of various wild animals in the protected area. Tracking algorithms (such as SORT, DeepSORT, etc.) are used to track wild animals in continuous images to ensure that their activities can be continuously monitored. Combined with the timestamp and geographic location information of the image (such as GPS data), the activity trajectories of wild animals are comprehensively analyzed to evaluate their activity range, migration patterns, etc. The types, numbers, activity trajectories and other information of wild animals are output as the third result. The above inspection results are obtained by combining the above first, second and third results.

[0110] The system uses AI algorithms to identify wild animals in wetland reserves and count their numbers. In the future, it can formulate reasonable inspection routes and flight altitudes based on the living habits of wild animals and the areas where they may appear. Consider factors such as the range of activities of wild animals, habitat characteristics, and topography to ensure coverage of key areas. Set a suitable flight speed and shooting frequency, conduct flight inspections along the planned routes, and continuously capture ground images and videos through the camera. During the shooting process, pay attention to adjusting the angle and focal length to ensure that wild animals at different distances and angles can be captured.

[0111] The third target detection model is trained in the following way: After collecting a large amount of image and video data containing wild animals, the collected data is labeled. After that, the labeled data is divided into training set, validation set and test set. The model is trained using the training set. The target detection algorithm is the deep learning target detection model YOLO (You Only Look Once). By continuously adjusting the parameters of the model, the model can accurately identify wild animals. The trained AI model analyzes the pre-processed image and video data to identify the wild animals in it. The model extracts and matches the features in the image to determine whether there are wild animals and determine their location.

[0112] The specific training process is as follows: This process describes in detail how to use the deep learning object detection model YOLO (You Only Look Once) to train a system that can accurately identify wild animals. The following are the detailed steps of the process: 1. Data Collection

[0113] First, collect a large amount of image and video data containing wildlife from various sources (such as nature reserve cameras, wildlife photographers, public datasets, etc.). This data should cover different types of wildlife, different ecological environments, different lighting conditions and weather conditions to ensure the generalization ability of the model. Carefully annotate each collected image and video frame, draw a bounding box for each wildlife instance, and annotate its species. The annotation information includes the coordinates and size of the bounding box and the species label of the wildlife.

[0114] 2. Dataset Division

[0115] The labeled data is divided into three independent sets: training set, validation set and test set. Usually, the training set occupies the majority (e.g. 70%) and is used to train the model; the validation set occupies a portion (e.g. 15%) and is used to adjust the model parameters during the training process to prevent overfitting; the test set occupies the remaining portion (e.g. 15%) and is used to evaluate the performance of the final model.

[0116] 3. Model Training

[0117] The YOLO model is trained using the training set data, the loss is calculated through forward propagation, and then the model parameters are updated through back propagation. During the training process, the performance of the model is regularly evaluated using the validation set, and the model parameters (such as learning rate, optimizer, regularization term, etc.) are adjusted as needed. The training process is repeated until the performance of the model on the validation set reaches a satisfactory level or the preset training rounds are reached.

[0118] 4. Model Evaluation

[0119] After training, the model is evaluated using the test set to understand the model's performance on unseen data. Evaluation metrics include accuracy, recall, F1 score, and average precision (AP). Analyze the model's performance on different types of wild animals and different environmental conditions to identify possible errors and limitations.

[0120] 6. Model deployment and application

[0121] Deploy the trained model to the actual application scenario, use the deployed model to analyze the pre-processed image and video data, identify wild animals in real time, and determine their location. Display the detection results to the user in a graphical way, such as displaying the location and species of wild animals on the monitoring screen. At the same time, collect user feedback for subsequent model optimization and improvement.

[0122] Through the precise identification of AI models, the present invention can promptly detect and report illegal personnel, illegal use of fire and other behaviors, and effectively curb behaviors that damage the ecological environment of the protected area. When a fire hazard is discovered or a fire occurs, it can be quickly located and reported, providing valuable time for emergency response, reducing disaster losses, and enhancing protection effects. Continuous monitoring and analysis of the activity trajectories of wild animals can help scientific researchers understand the ecological habits and migration patterns of wild animals, providing a scientific basis for ecological protection. This process realizes intelligent and efficient monitoring and management of wetland reserves through the application of AI technology, which is of great significance to protecting the ecological environment and promoting biodiversity. Moreover, the automated inspection process of the present invention greatly reduces the need for manual inspections, improves the efficiency and coverage of monitoring, and ensures that various abnormal situations in the protected area can be discovered in a timely manner. It reduces the reliance on a large number of human resources and reduces the labor cost of protected area management.

[0123] In an optional embodiment, in the above step S105, when the drone has a fault, controlling the drone according to a preset fault handling strategy includes: S501. Obtain meteorological information and airspace conditions, predict the flight risks faced by the UAV based on the airspace conditions, patrol flight plan and meteorological information, and resolve conflicts according to preset flight rules when there are flight risks.

[0124] Through professional airspace management systems, radar systems, meteorological observation stations and other multi-source data, the current airspace situation is obtained in real time. The airspace situation includes but is not limited to airspace occupancy, flight trajectories of other aircraft, no-fly zones, restricted zones and other information. The acquired airspace data is processed and analyzed to extract key information related to drone flights, such as whether the airspace is crowded and whether there is potential conflict.

[0125] The ground control and management center conducts a comprehensive assessment of the flight risks that the drone may face during the flight based on the acquired airspace conditions, patrol flight plans and meteorological information. The assessment includes but is not limited to the risk of collision with other aircraft, the risk of crossing no-fly zones or restricted areas, and the flight risks caused by bad weather. Specifically, the communication monitoring system can be used to analyze the communication signals between the drone and the ground control station in real time, and the approximate distance between the drones can be inferred through parameters such as signal strength and time difference. If the adjacent drones do not meet the preset straight-line spacing (such as 500m), and the drone and ground obstacles do not meet the preset spacing distance (such as 20m) requirements, the automatic warning system will make a comprehensive judgment at this time. Once it is determined that the possibility of a flight conflict is high, the conflict will be resolved immediately according to the flight rules preset by the system. Specific measures include: Adjust flight plan: Replan the flight route of the drone to avoid potential risk areas or conflict points with other aircraft.

[0126] Notify relevant parties: If the flight plan adjustment affects the normal flight of other aircraft, relevant parties need to be notified in a timely manner for coordination and communication.

[0127] Initiate emergency procedures: In extreme cases, such as severe weather or emergencies, emergency procedures need to be initiated immediately to ensure the safety of the drone.

[0128] Real-time monitoring and adjustment: During the flight of the drone, its status and airspace conditions are continuously monitored, and the flight plan is adjusted and optimized in real time according to the actual situation.

[0129] S502: When conflict resolution cannot be performed, obtain meteorological information and operational data of the UAV's flight system.

[0130] If conflict resolution is not possible (for example, due to a serious fault or airspace restrictions), the drone will obtain the operating data of its flight system to evaluate the current status of the drone. The ground operation control management center is responsible for monitoring and controlling the overall operating status of the drone. The hardware layer collects the operating data of the drone's flight system in real time, including but not limited to battery power, flight speed, altitude, attitude, GPS signal strength, engine status, etc.

[0131] S503: Determine whether the UAV can return based on the meteorological information and operation data.

[0132] Based on the meteorological information and the operating data of the flight system, the drone is evaluated to see whether it has the conditions to return. If the weather conditions are good and the flight system can support return, the drone will attempt to return. Specifically, the service layer receives the operating data transmitted by the hardware layer through streaming. The service layer combines the current meteorological information (such as wind speed, wind direction, rainfall, etc.) and the operating data of the drone to make a comprehensive judgment to determine whether the drone meets the conditions to return. When the drone completes the planned patrol mission, the return procedure is automatically triggered.

[0133] The drone will return to the base in the following situations and alert the integrated patrol personnel through the ground operation and control management center when returning: ①Complete the planned patrol mission

[0134] ② The performance of the drone cannot support the completion of the patrol mission

[0135] ③Low power

[0136] ④ UAV C2 link failure

[0137] ⑤ The drone returns to the drone nest, and the drone nest automatically charges the drone

[0138] If the drone has not completed the planned patrol mission when it returns, it will re-enter the "drone take-off" phase after completing charging, and fly to the mission interruption location to continue the patrol mission. If the drone has completed the patrol mission when it returns and there are other patrol missions on the same day, the drone will continue to complete the "drone take-off", "drone patrol" and "drone return" missions according to the plan.

[0139] If the performance of the drone (such as battery power, engine status) cannot support the mission, it will return immediately. If the battery power is lower than the preset threshold, the return procedure will be initiated. After the drone returns to the drone nest, the nest automatically charges the drone through the storage and charging system. When the communication link (C2 link) between the drone and the control center fails, it must return immediately to ensure the safety of the drone. During the return process, an alarm message is sent to the comprehensive patrol personnel through the ground operation and control management center to ensure that they understand the status of the drone.

[0140] If the planned mission is not completed when returning, the drone will re-enter the "UAV take-off" phase after charging is completed and fly to the interrupted mission location to continue patrolling. If the patrol mission has been completed but there are other missions on the same day, the drone will continue to perform subsequent missions as planned.

[0141] S504: When the drone is unable to return, select a target alternate landing field closest to the current drone location from the preset alternate landing fields at different distance levels, and control the drone to land at the target alternate landing field.

[0142] If the drone cannot return (for example, due to insufficient power, serious malfunction, etc.), when returning to the drone nest, you can select an alternate landing field for alternate landing according to the pre-study settings. Select the target alternate landing field closest to the current drone location from the pre-set alternate landing fields of different distance levels. Then, the drone will adjust its flight route and control it to alternate landing at the target alternate landing field. At the same time, the ground operation control management center will alert the integrated patrol personnel to assist the integrated patrol personnel in manually controlling the drone to return or go to the alternate landing field.

[0143] The design principles of the alternate landing strategy are as follows: ① Give priority to selecting a suitable site that is close to the current position of the drone as an alternate landing site. Generally speaking, within the normal flight range, multiple alternate landing sites at different distance levels should be determined, such as suitable sites at 5 kilometers, 10 kilometers, and 20 kilometers from the current position. For long-distance flight missions, alternate landing sites at certain intervals can be planned in advance according to the flight route to ensure that the alternate landing site can be reached within an acceptable time if a problem occurs at any location.

[0144] ② Topography

[0145] Avoid choosing areas with complex terrain, mountains, canyons, forests and other obstacles as alternate landing sites. Give priority to flat and open sites, such as large grasslands, empty farmland, abandoned runways, etc., so that the drone can land safely. Consider whether there are large buildings, high-voltage lines and other facilities that may affect landing in the surrounding area to ensure the safety of the environment around the alternate landing site.

[0146] The conditions for triggering a diversion are as follows: When the drone encounters heavy rain, strong wind, communication interruption, positioning loss, or a single engine failure on its way back, the alternate landing conditions will be triggered and it will land on the spot.

[0147] ① Failure of key system components of drones

[0148] When encountering the following situations, the drone will immediately trigger the alternate landing procedure: Abnormal data appears in the flight control system; The power output of the power system is unstable; A single engine failure of the drone; Communication interruption exceeds 1 minute; The drone loses its positioning for more than 1 minute; Abnormal sensor data of the drone was detected, such as large deviations in the altimeter, gyroscope, accelerometer, etc.

[0149] ② Weather conditions deteriorate

[0150] Real-time monitoring of weather changes in the flight area. When the following situations occur, the drone will immediately trigger the alternate landing procedure: Wind speed is greater than 10m / s; The rainfall in 24 hours is 25mm-50mm; The execution steps of the alternate landing procedure are designed as follows: ①Fault detection and judgment

[0151] The automatic detection system of the drone continuously monitors various parameters. Once an abnormality is found, it immediately performs fault diagnosis to determine the type and severity of the fault. Then, based on the fault diagnosis results, it determines whether an alternate landing is required. If the fault can be repaired during flight or does not affect flight safety, the mission can continue; if the fault cannot be repaired and may endanger flight safety, the alternate landing procedure is immediately initiated.

[0152] ②Choose an alternate airport

[0153] Automatically select the most suitable alternate landing site based on the preset alternate landing site selection criteria and the current UAV location, weather conditions and other factors. If there are multiple alternate landing sites to choose from, give priority to sites with short distances, good weather conditions and safe terrain. After selecting an alternate landing site, report the selected alternate landing site information to the ground command center so that ground personnel can prepare.

[0154] ③ Navigate to the alternate landing field

[0155] The drone's navigation system automatically plans the best route to the alternate landing site, taking into account factors such as flight altitude, wind direction, and obstacles to ensure safe and rapid arrival at the alternate landing site. During the flight, the flight parameters and weather conditions are continuously monitored, and if necessary, the navigation route can be adjusted according to actual conditions.

[0156] ④ Landing preparation

[0157] When approaching the alternate landing site, the drone lowers its flight altitude, adjusts its flight attitude, and prepares for landing. At the same time, the drone establishes communication with the ground commander at the alternate landing site to obtain the latest site information and landing instructions.

[0158] ⑤Safe landing

[0159] Follow the instructions of the ground commander and land the drone smoothly at the right time. During the landing process, pay close attention to the status of the drone to ensure landing safety. After landing, immediately report the landing location and drone status to the ground command center and wait for further instructions.

[0160] The above data are only examples, and those skilled in the art can set them according to actual needs.

[0161] If the drone cannot return for some reason, the system will select the target alternate landing field closest to the current drone location based on the pre-set alternate landing fields of different distance levels. The drone will be controlled to fly to the target alternate landing field for emergency landing to ensure the safety of the drone and mission data.

[0162] The present invention can significantly reduce the risks of UAVs colliding with other aircraft, mistakenly entering no-fly zones or restricted areas, etc. during flight by acquiring meteorological information and airspace conditions, predicting flight risks, and performing conflict resolution, thereby improving the safety of patrol missions. Reasonable flight plan adjustments and conflict resolution measures can reduce unnecessary waiting and detours for UAVs during flight, thereby improving the execution efficiency of patrol missions. The entire process relies on the support of multi-source data and intelligent algorithms such as professional airspace management systems, radar systems, meteorological observation stations, etc., to achieve intelligent management and control of UAV patrol missions. Through accurate flight risk prediction and conflict resolution, patrol flight plans and task allocation can be better arranged, thereby optimizing resource allocation and improving overall operating efficiency. It reduces mission interruptions or delays caused by flight conflicts or safety risks, and improves user satisfaction and trust in UAV patrol services.

[0163] The present invention also monitors the operating data and meteorological information of the drone in real time, timely judges and triggers the return program, and effectively avoids safety accidents caused by insufficient power, performance degradation or bad weather. After charging, the drone can automatically continue the unfinished task to ensure the continuity and integrity of the patrol mission. In the case that the drone cannot return, it can quickly screen out the nearest alternate landing site and make an emergency landing, reducing losses and ensuring the safety of the drone. The whole process realizes the intelligent management and control of drone patrol tasks, and improves the efficiency and level of wetland protection. By sending alarm information to comprehensive patrol personnel, they can understand the status of the drone and the progress of the task in a timely manner, which enhances the user experience and satisfaction.

[0164] In an optional embodiment, the conflict resolution according to the preset flight rules in step S501 above includes: S601, controlling one or more drones to ascend or descend to a preset height.

[0165] When the ground operation control management center detects that two or more drones are at risk of crossing paths or approaching each other in the vertical direction, it will control one or more drones to rise or descend to a preset altitude. This preset altitude is determined based on the current airspace conditions (such as the flight altitude of other aircraft, terrain altitude, etc.) and the performance parameters of the drone (such as maximum climb / descent rate, safe flight altitude range, etc.). By sending an altitude adjustment command to the drone, the drone will adjust its flight control system to change the flight altitude after receiving the command, thereby avoiding vertical conflicts with other drones.

[0166] S602: When two UAVs are flying in the same direction and the distance between them is gradually shortened, send commands to the UAVs to change their flight directions, so that one of the UAVs deflects while the other maintains its original direction, thereby achieving safe separation.

[0167] When two drones are flying in the same direction and the distance between them is gradually shortened, in order to avoid them arriving at the same location at the same time and colliding, the ground operation control management center will send a command to one of the drones to change its flight direction, making it deflect, while the other drone continues to maintain its original course. Based on the relative position and speed of the two drones, the optimal deflection angle and timing are calculated, and a heading adjustment command is sent to the drone that needs to be deflected. After receiving the command, the drone will adjust its flight control system to change the flight direction.

[0168] S603, adjust the flight speed of the drone to change the time it takes for the drone to arrive at the preset location, thereby avoiding conflict. If two drones are expected to arrive at the same location at the same time, one of them can be instructed to speed up or slow down so that it is staggered with the other drone in time to reduce the risk of conflict.

[0169] In order to change the time it takes for the drone to reach the preset location and avoid conflicts with other drones while keeping the drone's flight direction unchanged, the ground operation control center will adjust the drone's flight speed. Based on the drone's current speed, distance from the preset location, and relative position and time difference with other drones, the speed value that needs to be adjusted is calculated, and a speed adjustment command is sent to the drone. After receiving the command, the drone will adjust its power output to change the flight speed.

[0170] If the conventional strategies such as S601-S603 above cannot effectively resolve the flight conflict, emergency measures need to be initiated to notify ground personnel through the background to perform manual control.

[0171] The present invention can effectively avoid flight conflicts and collision risks between UAVs and improve the safety of patrol missions by combining and applying multiple strategies such as rising / descending altitude, changing heading and adjusting speed. The preset flight rules provide a variety of conflict resolution strategies, and the ground operation control management center can flexibly select and apply these strategies according to actual conditions to deal with different types of flight conflicts. By adjusting the flight speed and heading of the UAV, the flight path and schedule of the UAV can be optimized, reducing unnecessary waiting and circling time, and improving the execution efficiency of patrol missions. Through timely and effective conflict resolution measures, it can be ensured that the UAV can complete the patrol mission continuously and stably, avoiding mission interruptions or delays caused by flight conflicts.

[0172] Optionally, in the above step S105, when the drone has a fault, controlling the drone according to a preset fault handling strategy further includes: S701. When the UAV is unable to make an emergency landing, control the UAV to send a crash warning to the ground operation and control management center, and update the flight dynamic information to the ground operation and control management center.

[0173] If a drone encounters a bird strike during flight, it may cause damage to the drone, sensor failure, or power system failure. Extreme weather conditions such as strong winds, thunderstorms, and haze may affect the drone's flight stability and navigation capabilities. The drone may experience equipment failures during flight, including but not limited to internal equipment problems such as battery exhaustion, motor failure, and navigation system failure.

[0174] Once the drone detects any one or more of the above abnormal conditions and determines that it cannot return safely or continue the mission, it will immediately send a crash warning signal to the ground operation control management center through a preset communication link. The signal contains key data such as the current status, location information, and fault type of the drone so that the control system can respond quickly. At the same time, the drone continues to send the latest flight dynamic information to the ground operation control management center, such as altitude, speed, direction, etc. This information is of great significance for subsequent analysis of the cause of the crash, optimization of flight strategies, and improvement of drone safety. After receiving the crash warning, the ground operation control management center immediately initiates the emergency response procedure. The system attempts to continue monitoring the flight dynamics of the drone to obtain more information about the pre-crash status.

[0175] S702: When it is impossible to obtain flight dynamic information, use the crash dynamics model to identify the crash site and alert the integrated patrol personnel.

[0176] If the ground operation control management center cannot continue to monitor the flight dynamics of the drone due to a drone failure, the crash dynamics model is used to perform simulation calculations in combination with the last sent location information, speed, altitude, wind direction and speed and other environmental factors, as well as the physical characteristics and fault type of the drone. Through model calculations, the possible crash site range of the drone is predicted, and detailed geographic location information is generated to identify the crash site. The system sends the predicted crash site information to the comprehensive patrol personnel, issues an emergency alert, and requires immediate search in the area. Auxiliary tools such as map positioning and path planning are provided to assist comprehensive patrol personnel in finding the crashed drone on site.

[0177] The air-ground coordinated wetland aerial patrol device provided by the present invention is described below. The air-ground coordinated wetland aerial patrol device described below and the air-ground coordinated wetland aerial patrol method described above can be referenced to each other.

[0178] The above-mentioned air-ground coordinated wetland aerial patrol device refers to Figure 4 As shown, including: The plan making module 801 is used to make a patrol flight plan based on the operation area and restricted flight area defined by the operation control management center; The first flight confirmation module 802 is used to confirm the first flight every day. When the flight control right is obtained, the UAV is controlled to take off and automatically patrol according to the patrol flight plan. The real-time monitoring module 803 is used to collect image and video data and monitor the operation status in real time during the automatic patrol of the UAV, and obtain the real-time status and alarm information of the UAV equipment; The data analysis module 804 is used to analyze the image and video data to obtain the inspection results, and generate an alarm signal when there is an abnormality in the inspection results; The fault judgment module 805 is used to judge whether the drone has a fault based on the real-time status of the device and the alarm information. If the drone has a fault, the drone is controlled according to a preset fault handling strategy.

[0179] Figure 5 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 5 As shown, the electronic device may include: a processor 910, a communication interface 920, a memory 930, and a communication bus 940, wherein the processor 910, the communication interface 920, and the memory 930 communicate with each other through the communication bus 940. The processor 910 may call the logic instructions in the memory 930 to execute the air-ground collaborative wetland aerial patrol method.

[0180] In addition, the logic instructions in the above-mentioned memory 930 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0181] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the air-to-ground collaborative wetland aerial patrol method provided by the above-mentioned methods.

[0182] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the air-ground collaborative wetland aerial patrol method provided by the above-mentioned methods.

[0183] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0184] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0185] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for aerial patrol of wetlands in air-ground collaboration, characterized in that: include: Formulate patrol flight plans based on the operation areas and restricted flight areas defined by the Operation and Control Management Center; Conduct daily first flight confirmation, control the drone to take off after obtaining flight control rights, and conduct automatic patrols according to the patrol flight plan; During the automatic patrol process of the drone, real-time image and video data are collected and the operation status is monitored to obtain the real-time status and alarm information of the drone equipment; Analyze the image and video data to obtain inspection results, and generate an alarm signal when there is an abnormality in the inspection results; Based on the real-time status of the device and the alarm information, it is determined whether the drone has a fault. If the drone has a fault, the drone is controlled according to a pre-set fault handling strategy.

2. The air-ground coordinated wetland aerial patrol method according to claim 1 is characterized in that: The image and video data are analyzed in the following ways: The image video data is detected by using a pre-trained first target detection model to identify whether there are any violators in the wetland protection area, and a virtual identification is generated for the violators; The image video data is detected by a pre-trained second target detection model to identify whether there is illegal use of fire or fire hazard in the wetland protection area; The image video data is detected through a pre-trained third target detection model to identify whether there are wild animals in the wetland reserve, and a tracking algorithm is used to track the wild animals in continuous images. The activity trajectory of the wild animals is analyzed in combination with the timestamp and geographic location information of the image.

3. The air-ground coordinated wetland aerial patrol method according to claim 1 is characterized in that: The first flight of each day is confirmed by: The environmental monitoring system installed on the drone nest collects the environmental information of the drone before takeoff; Acquiring meteorological information, and performing meteorological environment analysis on the environmental information and meteorological information to determine whether the conditions for takeoff are met; When the conditions for takeoff are met, apply for flight control.

4. The air-ground coordinated wetland aerial patrol method according to claim 1 is characterized in that: When the drone fails, the drone is controlled according to a preset fault handling strategy, including: Obtain weather information and airspace conditions, predict the flight risks faced by drones based on airspace conditions, patrol flight plans and weather information, and resolve conflicts according to preset flight rules when there are flight risks; Obtaining operational data on the drone’s flight systems when deconfliction is not possible; Determine whether the drone can return based on weather information and operational data; In the case that the drone cannot return, the target alternate landing field closest to the current drone position is selected from the preset alternate landing fields of different distance levels, and the drone is controlled to make an alternate landing to the target alternate landing field.

5. The air-ground coordinated wetland aerial patrol method according to claim 4 is characterized in that: The conflict resolution according to the preset flight rules includes: Control one or more drones to rise or descend to a preset altitude; the preset altitude is determined according to airspace conditions and performance parameters of the drones; When two drones are flying in the same direction and the distance between them is gradually shortened, a command is sent to the drones to change their flight directions, so that one of the drones deflects while the other continues to maintain its original direction; Adjust the drone's flight speed to change the time it takes for the drone to reach a preset location.

6. The air-ground coordinated wetland aerial patrol method according to claim 4 is characterized in that: When the drone fails, controlling the drone according to a preset fault handling strategy also includes: If the drone cannot make an emergency landing, control the drone to send a crash warning to the ground operation and control management center, and update the flight dynamic information to the ground operation and control management center; In the absence of flight dynamics information, the crash dynamics model is used to identify the crash site and alert the integrated patrol personnel.

7. An air-ground coordinated wetland aerial patrol device, characterized in that: include: The planning module is used to formulate patrol flight plans based on the operation areas and restricted flight areas defined by the operation and control management center; The first flight confirmation module is used to confirm the first flight every day. When the flight control right is obtained, the UAV is controlled to take off and automatically patrol according to the patrol flight plan. Real-time monitoring module, used to collect image and video data and monitor operation status in real time during the automatic patrol process of the drone, and obtain the real-time status and alarm information of the drone's equipment; A data analysis module, used to analyze the image and video data to obtain inspection results, and generate an alarm signal when there is an abnormality in the inspection results; The fault judgment module is used to judge whether the drone has a fault based on the real-time status of the device and the alarm information. If the drone has a fault, the drone is controlled according to a preset fault handling strategy.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, it implements the air-ground collaborative wetland aerial patrol method as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the air-ground collaborative wetland aerial patrol method as described in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the air-ground collaborative wetland aerial patrol method as described in any one of claims 1 to 6 is implemented.

Images