A method and device for unmanned aerial vehicle radar monitoring of a nuclear power plant intake sea area

Abstract

This invention relates to the field of unmanned aerial vehicle (UAV) patrol technology, and more particularly to a UAV radar monitoring method and equipment for the sea area near a nuclear power plant's water intake. The UAV radar monitoring method for the sea area near a nuclear power plant's water intake includes: acquiring the monitoring and early warning requirements for floating objects on the sea surface and the target flight area, obtaining the monitoring and early warning threshold and the location range of the UAV's patrol flight area; determining the UAV patrol path and patrol strategy based on the monitoring and early warning threshold and the location range of the UAV's patrol flight area; conducting patrols according to the patrol strategy and the UAV patrol path, and acquiring graphic data of the target patrol path; performing recognition processing on the graphic information of the target patrol path to obtain the recognition result; and determining whether the recognition result meets preset alarm conditions, and issuing an early warning if it does. This invention enables rapid, large-scale patrols, recording, and analysis of the sea area near a nuclear power plant's water intake, improving the accuracy of monitoring floating objects in the large-scale sea area near a nuclear power plant's water intake.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) patrol technology, and in particular to a UAV radar monitoring method and equipment for the sea area of ​​a nuclear power plant intake. Background Technology

[0002] The sea area at the intake of a nuclear power plant provides cooling water to the nuclear power units, which is crucial for their safe and stable operation. To prevent floating debris from clogging pump station filters and threatening the safe operation of the units, monitoring of the intake area is necessary to promptly detect floating debris that may affect the safety of the nuclear power plant's cooling source and develop cleanup plans. Existing monitoring methods mainly employ shore-based fixed radar and optical video imaging monitoring. Shore-based radar can overcome the effects of weather and lighting to detect floating debris on the sea surface, but it cannot overcome interference from sea surface clutter, making it difficult to accurately monitor oil spills and small floating objects. Furthermore, it cannot image and monitor floating debris over large areas, resulting in unintuitive monitoring results and difficulty in quickly and accurately identifying potential floating debris. Optical video imaging monitoring methods are easily affected by factors such as lighting, weather, and smoke, cannot guarantee all-weather monitoring, and have a limited monitoring range. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a method and equipment for unmanned aerial vehicle (UAV) radar monitoring of the sea area where a nuclear power plant's water intake is located.

[0004] The technical solution adopted by this invention to solve its technical problem is: a method for unmanned aerial vehicle (UAV) radar monitoring of the sea area where a nuclear power plant's water intake is located, comprising: Step S10: Obtain the requirements for monitoring and early warning of floating objects on the sea surface and the target flight area, and obtain the monitoring and early warning threshold and the location range of the UAV inspection flight area; Step S20: Determine the UAV inspection path and inspection strategy based on the monitoring and early warning threshold and the location range of the UAV inspection flight area; Step S30: Perform inspection according to the inspection strategy and the UAV inspection path, and obtain the target inspection path graphic data; Step S40: The target inspection path graphic information is processed for identification to obtain the identification result; Step S50: Determine whether the identification result meets the preset alarm conditions. If it does, issue an early warning.

[0005] Optionally, step S10 includes: Obtain the aforementioned requirements for monitoring and early warning of floating objects on the sea surface; The horizontal projected area of ​​the floating objects requiring early warning is obtained according to the aforementioned requirements for monitoring and early warning of floating objects on the sea surface; The monitoring and early warning threshold is determined based on the horizontal projected area.

[0006] Optionally, step S20 includes: The drone inspection path is determined according to the location range of the drone inspection flight area and the drone radar detection range, based on a preset overlap. The UAV inspection path is parallel or perpendicular to the baseline direction of the UAV inspection flight area.

[0007] Optionally, step S30 includes: Acquire single-shot radar monitoring data and spatial location of the UAV; The single radar monitoring data is fused with the spatial location to obtain a single radar image; Based on the UAV inspection path, the single radar image is stitched together according to the required image size and the spatial position registration to obtain the target inspection path graphic data.

[0008] Optionally, step S40 includes: The identification result is obtained by identifying the type, size, and location of floating objects on the target inspection path graphic data through feature recognition.

[0009] Optionally, step S40 includes: Establish a dataset of floating object radar images; An artificial neural network algorithm was trained using the aforementioned floating object radar image dataset to obtain a recognition model; The recognition model is used to identify and detect the graphic data of the target inspection path to obtain the recognition result.

[0010] Optionally, step S50 includes: The projected area is obtained from the recognition result; Determine whether the projected area is greater than a preset area; if so, determine that the preset alarm condition is met.

[0011] Optionally, it also includes: Step S60: Determine the abnormal area based on the warning information; Step S70: Determine whether the drone's battery life information meets the abnormal inspection requirements; if it does, then replan the drone's inspection path.

[0012] Optionally, it includes: Obtain meteorological information and remaining battery information for the abnormal area; Based on the meteorological information and the remaining flight information, determine whether the drone meets the abnormal inspection requirements; If the conditions are met, the UAV inspection path will be replanned based on the abnormal area. If the conditions are not met, the inspection will continue according to the described drone inspection path.

[0013] The present invention also provides a drone radar monitoring device, including a memory and a processor; The memory stores computer programs; The processor executes the steps of the unmanned aerial vehicle (UAV) radar monitoring method for the sea area of ​​the nuclear power plant intake as described in any of the preceding claims by calling the computer program stored in the memory.

[0014] Implementing this invention has the following beneficial effects: This invention utilizes UAV radar monitoring to conduct inspections based on inspection strategies and UAV inspection paths, thereby achieving large-scale rapid inspection, recording, and analysis of the sea area around the nuclear power plant's water intake, and improving the accuracy of monitoring floating objects in the large-scale sea area around the nuclear power plant's water intake. Attached Figure Description

[0015] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 This is a flowchart of a method for monitoring the sea area of ​​a nuclear power plant's water intake using unmanned aerial vehicle (UAV) radar, as described in one embodiment. Figure 2 This is a schematic diagram of the drone scanning path in one embodiment. Detailed Implementation

[0016] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0017] like Figure 1 As shown, the present invention provides a method for unmanned aerial vehicle (UAV) radar monitoring of the sea area near the intake of a nuclear power plant, comprising: Step S10: Obtain the requirements for monitoring and early warning of floating objects on the sea surface and the target flight area, and obtain the monitoring and early warning threshold and the location range of the UAV inspection flight area.

[0018] To effectively obtain the relevant requirements for monitoring and early warning of floating objects on the sea surface, and to clarify the detailed information of the target flight area, specific threshold standards for monitoring and early warning are analyzed based on historical data. Simultaneously, it is also necessary to accurately determine the specific location range of the UAV inspection flight, including its latitude and longitude coordinates, flight altitude, and coverage area, among other key parameters. This ensures the accuracy of the monitoring and early warning system and the efficient execution of UAV inspection tasks, thereby achieving timely monitoring and early warning of floating objects on the sea surface.

[0019] Step S20: Determine the drone inspection path and inspection strategy based on the monitoring and early warning threshold and the location range of the drone inspection flight area.

[0020] Based on pre-set monitoring and early warning threshold standards and the specific geographical location of the drone inspection flight area, and taking into account the basic conditions of climate, drones, and floating objects, the drone inspection path planning and corresponding inspection strategy are determined. By accurately analyzing the characteristics of the early warning threshold and the flight area, it is ensured that the drone can perform inspection tasks efficiently and accurately, maximize inspection results, and promptly identify and address potential problems.

[0021] Step S30: Conduct inspections according to the inspection strategy and the UAV inspection path, and obtain graphic data of the target inspection path.

[0022] Specifically, the drones utilize radar monitoring to conduct inspections based on inspection strategies and drone inspection paths. Following pre-set inspection strategies and meticulously planned drone inspection paths, the inspection work is systematically carried out. During this process, various graphic data along the target inspection path are accurately acquired and recorded, ensuring data integrity and accuracy, and providing reliable basic information for subsequent analysis and processing.

[0023] Step S40: Recognize and process the graphic information of the target inspection path to obtain the recognition result.

[0024] By employing advanced image recognition algorithms and radar data processing technology, the acquired target inspection path graphic information is analyzed in depth and accurately identified. Key feature information is extracted and transformed into an understandable and usable data form, thereby obtaining accurate and reliable identification results.

[0025] Step S50: Determine whether the identification result meets the preset alarm conditions. If it does, issue an early warning.

[0026] The system needs to evaluate the identification results to determine whether they meet the preset alarm conditions. If the identification results do meet the preset alarm conditions, the system will immediately activate the early warning mechanism and take corresponding early warning measures to promptly notify relevant personnel and take necessary countermeasures, thereby effectively preventing the occurrence of potential risks.

[0027] In one embodiment, step S10 includes: Requirements for obtaining monitoring and early warning of floating objects on the sea surface.

[0028] By communicating with relevant departments of the nuclear power plant, reviewing historical monitoring data, and referring to industry standards and regulations, we comprehensively and meticulously obtained the requirements for monitoring and early warning of floating objects on the sea surface. This includes the types of floating objects, such as large pieces of garbage, oil spills, and algal blooms; defining size thresholds for different types of floating objects, such as setting thresholds for large garbage exceeding a certain length or oil spills reaching a specific area; determining thresholds for the quantity of floating objects, such as triggering an early warning if a certain number of specific floating objects appear in a unit of sea area; and considering the movement trends of the floating objects, such as whether they are moving rapidly towards the nuclear power plant's water intake.

[0029] According to the requirements for monitoring and early warning of floating objects on the sea surface, obtain the horizontal projected area of ​​the floating objects that need to be warned.

[0030] Using radar equipment mounted on drones, the sea surface is scanned in all directions with high precision. Advanced algorithms are used to process the radar echo data to accurately calculate the horizontal projected area of ​​floating objects requiring warning. This process fully considers the influence of the shape, material, and other factors of different types of floating objects on radar reflection characteristics, ensuring the accuracy and reliability of the area calculation.

[0031] The monitoring and early warning threshold is determined based on the horizontal projected area.

[0032] Based on the obtained horizontal projected area of ​​the floating debris requiring warning, combined with the actual conditions of the nuclear power plant's intake sea area, such as the size of the sea area and water flow conditions, and referring to the degree of risk caused by similar floating debris in the past, the monitoring and warning thresholds corresponding to different types of floating debris are comprehensively determined. For large pieces of waste that may seriously affect the safety of nuclear power plant water intake, a warning threshold is set when their horizontal projected area reaches a relatively small but potentially hazardous value; for oil pollution, the corresponding area warning threshold is determined based on its diffusion characteristics and the degree of pollution it may cause to water quality; for algal aggregates, considering that their large-scale accumulation may block the intake, an appropriate area warning threshold is set based on past experience. By accurately determining the monitoring and warning thresholds in this way, potential risks can be detected more promptly and accurately, ensuring the safe operation of the nuclear power plant's intake.

[0033] In one embodiment, step S20 includes: The drone inspection path is determined according to the location range of the drone inspection flight area and the drone radar detection range, based on a preset overlap.

[0034] The planning and determination of drone inspection paths are based on meticulous consideration of the specific location and range of the drone's inspection flight area, combined with the detection range of the radar system on board the drone, and precisely set according to preset overlap standards. This ensures that the drone can cover the designated inspection area without blind spots, while also fully utilizing the maximum effectiveness of radar detection. Through reasonable path planning, each flight maximizes the drone's inspection efficiency while ensuring safety, thereby guaranteeing the comprehensiveness and accuracy of the inspection data.

[0035] In some scenarios, the target area of ​​the drone flight is determined by removing the boundary of the target area monitoring range and thus determining the effective detection range of the drone radar.

[0036] The drone inspection path is parallel or perpendicular to the baseline direction of the drone inspection flight area.

[0037] In practice, setting the drone inspection path to be parallel or perpendicular to the baseline of the drone's inspection flight area offers several advantages. When the path is parallel to the baseline, it provides a more regular coverage of the inspection area, facilitating continuous monitoring of targets in the same direction. This helps identify anomalies along this direction and creates a clearer data sequence, improving data processing efficiency. Conversely, when the path is perpendicular to the baseline, it allows for a comprehensive scan of the inspection area from another dimension, complementing the parallel path and ensuring no area is missed. This enhances the ability to capture various conditions within the inspection area, making the monitoring of the nuclear power plant's intake area more comprehensive, detailed, and reliable.

[0038] In some scenarios, the drone inspection path is determined by a certain overlap amount based on the target monitoring area and the drone's radar detection range. The scanning path can be parallel or perpendicular to the target area's baseline. The amount of overlap depends on the drone's flight speed and the possible drift speed and direction of floating objects, and is set based on empirical values ​​determined by weather and wind direction. The drone's control strategy can be configured flexibly by the inspection personnel, using automatic, manual, or a combination of both methods, depending on the inspection path.

[0039] Understandably, since floating objects move with the waves and wind, the aircraft scanning process may result in some objects being missed (drifting from undetected areas to detected areas). Therefore, an overlap setting is necessary. The overlap is estimated based on the product of the estimated drift speed and the flight time of a single scan, converted into a coefficient according to an acceptable missed detection rate. The missed detection rate is related to the product of the drift speed ratio and the ratio of the scan width to the length. The faster the drift speed ratio and the larger the product of the speed ratio and the ratio of the scan width to the length, the lower the missed detection rate. Due to the uncertainty of the direction and drift speed of floating objects, an empirical value can be simplified based on environmental conditions such as weather during engineering implementation. This value can then be adjusted according to the actual missed detection situation. If there are many missed detections, the overlap is increased; conversely, the overlap is decreased.

[0040] In one embodiment, the flight path of the UAV is planned using geometric path planning based on the UAV takeoff area, the inspection area of ​​the furthest target, the radar scan coverage area, and the set overlap area. Figure 2 As shown, the flight direction can be adjusted to be horizontal or vertical according to the target area, planned from far to near. The strategy mainly considers the balance between scanning efficiency and drone flight time limits, how to return when the drone's power supply and weather conditions are unfavorable, which scanning path is more efficient, and which takes into account the safety of returning when the drone's power is low. These strategies can be determined by the operator. The strategy ensures full coverage of the inspection area and can return to the starting position for charging or to avoid severe weather based on the remaining power and weather changes. The flight area can be determined in advance by the operator based on the inspection task.

[0041] In one embodiment, step S30 includes: Acquire single-shot radar monitoring data and spatial location of the drone.

[0042] After acquiring single-shot radar monitoring data and spatial location information from a UAV, preliminary analysis can be performed on the data to determine if it contains any abnormal signal characteristics. Simultaneously, based on the UAV's spatial location information, the specific area of ​​the nuclear power plant's water intake sea area corresponding to that monitoring data can be accurately determined. The analyzed radar monitoring data is then associated and stored with the corresponding sea area for subsequent comprehensive analysis and processing. By integrating multiple acquisitions of single-shot radar monitoring data and spatial location information, a radar monitoring data map of the nuclear power plant's water intake sea area can be constructed, providing a more intuitive representation of the situation at different locations within the sea area.

[0043] A single radar monitoring data point is fused with spatial location data to obtain a single radar image.

[0044] By deeply fusing the data acquired from a single radar monitoring session with specific spatial coordinates, a complete and accurate single-session radar monitoring image is generated. This ensures the accuracy and real-time performance of the radar data.

[0045] In some scenarios, a single radar scan is treated as a line (or one frame). The drone's flight is equivalent to scanning the radar data frame by frame. Since the radar is fixed within the drone and radar waves travel at high speeds, the drone's speed can be ignored. Knowing the drone's spatial coordinates (including altitude) and flight direction allows the radar image to be aligned according to these coordinates. After alignment, the image can be converted to grayscale or rendered according to specific requirements. Drone radar detects by transmitting and receiving radar signals. Different objects on the sea surface reflect waves with different characteristics. By identifying these reflected waves and their characteristics (such as distance-time difference and the spectrum of the reflected waves), they can be equivalently represented as grayscale images.

[0046] Based on the UAV inspection path, the single radar images are stitched together according to the required image size and spatial location registration to obtain the target inspection path graphic data.

[0047] Based on the pre-set UAV inspection path, and according to the specific size and precise spatial location registration requirements of the required images, the radar images acquired in a single instance are stitched together with high precision to ultimately obtain complete graphic data that perfectly matches the target inspection path.

[0048] Furthermore, since the UAV can obtain information such as spatial position and flight speed through spatial position sensors such as GPS and gyroscope during the scanning process, the spatial position of each radar image in the sea area can be obtained based on the fixed position of the radar on the UAV. The images are aligned and stitched together by using the spatial position of the images. The stitching process will perform image fusion on the overlapping areas to eliminate ghosting.

[0049] In one embodiment, step S40 includes: The identification results are obtained by identifying the type, size, and location of floating objects on the target inspection path graphic data.

[0050] By using advanced feature recognition technology, the graphic data on the target inspection path is analyzed in detail to accurately identify the specific type, size, and precise location information of floating objects in the graphic data, thereby obtaining comprehensive and accurate identification results.

[0051] In one embodiment, step S40 includes: Establish a radar image dataset of floating objects.

[0052] This project aims to construct a radar image dataset specifically for floating objects. The dataset will cover various types of floating objects, including but not limited to common water surface debris such as plastic waste, branches, and duckweed. By collecting and organizing radar images of these floating objects under different environmental conditions, this dataset will provide valuable training material for the development and optimization of radar image recognition algorithms, and also offer important reference data for researchers in related fields.

[0053] An artificial neural network algorithm was trained using a dataset of floating object radar images to obtain a recognition model.

[0054] During training, the floating object radar image dataset was divided into training, validation, and test sets. Radar image data from the training set was input into the artificial neural network algorithm, allowing the algorithm to continuously learn and adjust its parameters to gradually improve the accuracy of identifying the type, size, and location of floating objects. The validation set was used to evaluate the model in real time during training, and the training strategy was adjusted promptly based on the evaluation results to prevent overfitting or underfitting. After extensive training and optimization, the test set was used to comprehensively evaluate the final recognition model, ensuring that the model can accurately and efficiently identify floating object information detected by UAV radar in the waters near the nuclear power plant intake area in practical applications.

[0055] The target inspection path graphic data is identified and detected by the recognition model to obtain the recognition results.

[0056] By using a recognition model, the graphic data corresponding to the preset target inspection path is accurately identified and detected, thereby obtaining detailed recognition results and ensuring the accuracy and reliability of the data.

[0057] In some embodiments, image signal processing methods are used to identify the type, size, and location of floating objects through features. Alternatively, machine learning methods can be used to establish a floating object radar image dataset and train an artificial neural network algorithm in the dataset. The artificial neural network algorithm is used to identify and detect floating objects. After identifying floating objects using image processing or machine learning methods, the type and projected area of ​​the floating object are marked on the radar image, and a warning is issued based on the warning threshold of the projected area of ​​the floating object.

[0058] The types of floating objects are primarily identified by forming images based on the time-frequency characteristics of the reflected waves from different floating objects. These images are compared with images without floating objects, and traditional image signal processing methods such as image entropy or gradient descent contour recognition can be used to identify the type of floating object and its pixel location in the image, such as using the centroid position of the contour. Machine learning methods mainly involve manually annotating simulated or acquired radar images of floating objects. After training an artificial neural network on these datasets to achieve the required recognition accuracy, the artificial neural network can then be used to identify floating objects.

[0059] In one embodiment, step S50 includes: Obtain the projected area from the recognition results.

[0060] Determine if the projected area is greater than the preset area; if so, determine that the preset alarm conditions are met.

[0061] If the projected area is not greater than the preset area, further analysis of other relevant parameters or characteristics will continue to determine whether there are other situations that may meet the alarm conditions. When it is determined that the preset alarm conditions are met, the alarm mechanism will be triggered immediately, and the potential risk information of the floating objects will be promptly conveyed to relevant personnel through preset alarm methods, such as sound alarms, flashing lights, or sending notifications to the mobile devices of relevant personnel.

[0062] In one embodiment, it further includes: Step S60: Determine the abnormal area based on the early warning information.

[0063] The system identifies specific areas where abnormal situations occur based on early warning information, so that appropriate countermeasures can be taken in a timely manner.

[0064] Step S70: Determine whether the drone's battery life information meets the requirements for abnormal inspection; if it does, replan the drone inspection path.

[0065] If the drone's battery life requirements are not met, the drone will be immediately returned to its home base and replaced with a drone that meets the battery life requirements, or the original drone will be allowed to recharge before the anomaly inspection task can be performed. When replanning the drone inspection route, the location and extent of the anomaly area, as well as surrounding environmental factors such as wind direction and water flow, will be taken into account to ensure that the planned route can efficiently and comprehensively inspect the anomaly area, so as to obtain accurate and detailed anomaly data in a timely manner.

[0066] In one embodiment, it includes: Obtain meteorological and remaining flight information for abnormal areas.

[0067] Based on meteorological and flight endurance information, determine whether the drone meets the requirements for abnormal inspection.

[0068] If the conditions are met, the drone inspection route will be replanned based on the abnormal area.

[0069] If the requirements are not met, the inspection will continue according to the drone inspection path.

[0070] Understandably, the system needs to acquire comprehensive and detailed meteorological information and drone endurance information for the abnormal area. This includes, but is not limited to, meteorological parameters such as wind speed, wind direction, temperature, and humidity, as well as key data such as the drone's current battery level and estimated flight time.

[0071] Based on the collected meteorological and flight endurance information, the system will conduct a comprehensive analysis and evaluation to determine whether the drone is capable of performing inspection missions in abnormal areas. This judgment process needs to consider the impact of meteorological conditions on the drone's flight and whether the drone's current flight endurance is sufficient to complete the expected inspection mission.

[0072] If the assessment results show that the drone meets the requirements for anomaly inspection, the system will immediately replan and adjust the drone's inspection path based on the specific conditions of the anomaly area. The new path planning will ensure that the drone can efficiently and safely cover the anomaly area and complete the inspection task.

[0073] Conversely, if the assessment results indicate that the drone does not meet the requirements for abnormal inspection, the system will maintain the original drone inspection path and continue inspection along the predetermined route. This ensures that the drone can still complete inspection tasks in other areas under the current conditions, avoiding resource waste and potential risks.

[0074] In one embodiment, the system determines whether the inspection path needs to be reset based on the identification results and performs the corresponding inspection. When an anomaly is detected during the inspection, personnel are allowed to interrupt the previous flight path and conduct manual or automatic inspections of the selected area (not all areas) to confirm the anomaly. The inspection path is replanned by personnel. Since the inspection process also needs to consider weather conditions and the drone's own endurance, personnel determine whether it is necessary to address the anomaly or a specific area (manually designated during the inspection process).

[0075] Furthermore, after completing the inspection path, the UAV returns to the termination position, records the inspection data and analysis results, generates a UAV radar monitoring task sheet, and stores the monitoring data and task sheet to complete the UAV radar monitoring task.

[0076] The present invention also provides a radar monitoring device for unmanned aerial vehicles, including a memory and a processor.

[0077] The memory contains computer programs.

[0078] The processor executes the steps of the unmanned aerial vehicle (UAV) radar monitoring method for the sea area of ​​the nuclear power plant intake, as described above, by calling a computer program stored in memory.

[0079] It is understood that the above embodiments only illustrate preferred embodiments of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can freely combine the above technical features without departing from the concept of the present invention, and can also make several modifications and improvements, all of which fall within the protection scope of the present invention. Therefore, all equivalent transformations and modifications made with respect to the scope of the claims of the present invention should fall within the scope of the claims of the present invention.

Claims

1. A method for unmanned aerial vehicle (UAV) radar monitoring of the sea area near the intake of a nuclear power plant, characterized in that, include: Step S10: Obtain the requirements for monitoring and early warning of floating objects on the sea surface and the target flight area, and obtain the monitoring and early warning threshold and the location range of the UAV inspection flight area; Step S20: Determine the UAV inspection path and inspection strategy based on the monitoring and early warning threshold and the location range of the UAV inspection flight area; Step S30: Perform inspection according to the inspection strategy and the UAV inspection path, and obtain the target inspection path graphic data; Step S40: The target inspection path graphic information is processed for identification to obtain the identification result; Step S50: Determine whether the identification result meets the preset alarm conditions. If it does, issue an early warning.

2. The UAV radar monitoring method according to claim 1, characterized in that, Step S10 includes: Obtain the aforementioned requirements for monitoring and early warning of floating objects on the sea surface; The horizontal projected area of ​​the floating objects requiring early warning is obtained according to the aforementioned requirements for monitoring and early warning of floating objects on the sea surface; The monitoring and early warning threshold is determined based on the horizontal projected area.

3. The UAV radar monitoring method according to claim 1, characterized in that, Step S20 includes: The drone inspection path is determined according to the location range of the drone inspection flight area and the drone radar detection range, based on a preset overlap. The UAV inspection path is parallel or perpendicular to the baseline direction of the UAV inspection flight area.

4. The UAV radar monitoring method according to claim 1, characterized in that, Step S30 includes: Acquire single-shot radar monitoring data and spatial location of the UAV; The single radar monitoring data is fused with the spatial location to obtain a single radar image; Based on the UAV inspection path, the single radar image is stitched together according to the required image size and the spatial position registration to obtain the target inspection path graphic data.

5. The UAV radar monitoring method according to claim 1, characterized in that, Step S40 includes: The identification result is obtained by identifying the type, size, and location of floating objects on the target inspection path graphic data through feature recognition.

6. The UAV radar monitoring method according to claim 1, characterized in that, Step S40 includes: Establish a dataset of floating object radar images; An artificial neural network algorithm was trained using the aforementioned floating object radar image dataset to obtain a recognition model; The recognition model is used to identify and detect the graphic data of the target inspection path to obtain the recognition result.

7. The UAV radar monitoring method according to claim 1, characterized in that, Step S50 includes: The projected area is obtained from the recognition result; Determine whether the projected area is greater than a preset area; if so, determine that the preset alarm condition is met.

8. The UAV radar monitoring method according to claim 1, characterized in that, Also includes: Step S60: Determine the abnormal area based on the warning information; Step S70: Determine whether the drone's battery life information meets the requirements for abnormal inspection; If the conditions are met, the drone inspection path will be replanned.

9. The pressurization method according to claim 8, characterized in that, Step S70 includes: Obtain meteorological information and remaining battery information for the abnormal area; Based on the meteorological information and the remaining flight information, determine whether the drone meets the abnormal inspection requirements; If the conditions are met, the UAV inspection path will be replanned based on the abnormal area. If the conditions are not met, the inspection will continue according to the described drone inspection path.

10. A radar monitoring device for unmanned aerial vehicles (UAVs), characterized in that, Including memory and processor; The memory contains computer programs; The processor executes the steps of the UAV radar monitoring method for the sea area of ​​the nuclear power plant intake as described in any one of claims 1 to 9 by calling the computer program stored in the memory.

Images