A multi-layer three-dimensional prevention and control flying animal repelling system and method for a power transmission line

By constructing a multi-layered, three-dimensional prevention and control system, and combining radar detection, image acquisition, and data fusion, the identification and risk assessment of flying animals have been achieved, solving the problem of poor deterrence effect in existing technologies and realizing efficient and eco-friendly prevention and control of flying animals.

CN123100592APending Publication Date: 2026-11-17XUCHANG SIDA ELECTRIC POWER EQUIP
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611137081.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-29
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies struggle to achieve an efficient balance between "eco-friendliness" and "safety protection," lacking system-level coordination and resulting in poor effectiveness in repelling flying animals, as well as adaptation issues.

Method used

A multi-layered, three-dimensional prevention and control system is constructed, including a panoramic perception unit, an intelligent decision-making unit, and a collaborative execution unit. Combining radar detection, image acquisition, and data fusion, it enables the identification and risk assessment of flying animals, and drives them away through a collaborative architecture of ecological guidance, mid-level dynamic expulsion, and inner-level static barrier.

Benefits of technology

It realizes a repulsion strategy that shifts from passive defense to active guidance, effectively breaking the adaptability of flying animals, accurately classifying risk levels, ensuring safety as a safety net, and the system has a variety of non-contact repulsion methods and physical barriers to adapt to various environmental conditions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123100592A_ABST
    Figure CN123100592A_ABST
Patent Text Reader

Abstract

This invention discloses a multi-layered, three-dimensional system and method for controlling and repelling flying animals on power transmission lines. The system includes a panoramic perception unit, an intelligent decision-making unit, a collaborative execution unit, a remote control unit, and an energy management unit. The collaborative execution unit comprises an outer ecological guidance module, a middle dynamic repelling module, and an inner static barrier module. The method includes: acquiring target information through the panoramic perception unit; the intelligent decision-making unit using a flying animal recognition model to identify the species and determine the attitude of the target, and weighting the risk level based on species, distance, and behavioral patterns; when the risk level is medium or high, activating the middle dynamic repelling module to execute an alarm or strong repelling strategy; when it is determined that the target has entered an extremely high-risk area, triggering an emergency alarm and evidence collection mechanism, and simultaneously repelling the animal with the highest intensity. This invention achieves a balance between precise control and eco-friendliness through a combination of multi-level gradient response and randomized repelling, effectively reducing the bird-related flashover accident rate on power transmission lines.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of intelligent animal control technology, specifically to a multi-layered three-dimensional system and method for controlling and controlling flying animals on power transmission lines. Background Technology

[0002] With the improvement of the ecological environment and the enhancement of public awareness of protection, the population of flying animals has been increasing year by year, and their activities have increasingly intersected with power transmission lines. Large wading birds (such as Oriental white storks and black storks) perch and nest on power poles, and the flashover caused by their droppings and the phase-to-phase short circuits caused by nesting materials have become important factors threatening the safe and stable operation of the power grid.

[0003] Currently, the main methods for preventing and controlling flying animal pests along power transmission lines are divided into three categories: physical barriers, sound and light deterrence, and ecological guidance.

[0004] Physical barriers, such as bird spikes and bird shields, prevent flying animals from perching and causing bird droppings by occupying the space above the insulator or forming a physical barrier. These devices provide passive protection and are effective in the long term, but they only protect the installation point and cannot prevent flying animals from moving to other parts of the tower. Furthermore, large-scale installation is costly and may affect the accessibility for maintenance and repair.

[0005] Sound and light-based deterrent devices, such as laser deterrents, ultrasonic deterrents, and loudspeakers, actively drive away flying animals by emitting stimulus signals. These devices are highly proactive and have a wide coverage area; however, long-term use of a single stimulus method can lead to habituation and adaptation in flying animals, resulting in a decrease in deterrent effectiveness. Some devices lack target recognition capabilities, resulting in indiscriminate deterrence regardless of species, and may even disturb non-harmful flying animals and pedestrians.

[0006] Ecological guidance devices, such as artificial bird nests and perches, create a more suitable habitat in areas away from power lines, guiding flying animals to voluntarily move away from power facilities. This method aligns with ecological protection principles, but it has a long construction period, making it difficult to establish immediate protective capabilities, and its guiding effect is affected by multiple factors such as habitat selection and seasonal migration.

[0007] In existing technologies, various decoy methods are often deployed independently, lacking system-level coordination and linkage. Although some solutions attempt to combine radar detection with visual recognition, they still have shortcomings in multi-source data fusion, accurate risk assessment, and tiered response strategies, making it difficult to achieve an efficient balance between "eco-friendliness" and "security protection". Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of the prior art and provide a multi-layer three-dimensional system and method for preventing and controlling flying animals on power transmission lines.

[0009] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a multi-layer three-dimensional control and dispersal system for flying animals in power transmission lines, comprising a panoramic perception unit, an intelligent decision-making unit, and a collaborative execution unit; The panoramic perception unit is used to acquire environmental data and flying animal target information around the protected facility in real time. It includes a radar detector, an image acquisition device and a data fusion module. The data fusion module is used to synchronize the radar detection data and the image acquisition data in time and space to obtain the position and motion status information of the flying animal target. The intelligent decision-making unit is connected to the panoramic perception unit and is used to identify flying animals based on flying animal target information and assess risk levels. It has a built-in flying animal identification model and risk grading model. The flying animal identification model is used to determine the species and posture of flying animals based on perception data, and the risk grading model is used to calculate risk values ​​based on flying animal species, distance and behavior patterns, and output the corresponding risk level. The collaborative execution unit is connected to the intelligent decision-making unit, and includes an outer ecological guidance module, a middle dynamic expulsion module, and an inner static barrier module. The outer ecological guidance module is located in an area far from the protected facilities to provide a habitat for flying animals; The mid-level dynamic deterrence module is used to actively deter flying animals that approach the protected facility by combining multiple non-contact deterrence methods according to the risk level. The inner static barrier module is installed on the protected facility to form a physical barrier to prevent flying animals from perching on the protected facility.

[0010] As a further optimization of the multi-layer three-dimensional flight animal control and removal system for power transmission lines of the present invention: the middle layer dynamic removal module includes a laser removal device and an acoustic removal device.

[0011] As a further optimization of the multi-layer three-dimensional flight animal control and deterrence system for power transmission lines of the present invention: the inner static barrier module includes at least one of bird-proof umbrella cover, bird-proof spikes and bird-proof wrapping spikes. The bird-proof umbrella cover is used to be installed above the insulator string, and the bird-proof wrapping spikes are flexible structures used to be wrapped around the conductor or ground wire.

[0012] As a further optimization of the multi-layer three-dimensional control and flight animal deterrence system for power transmission lines of the present invention: the outer ecological guidance module includes an artificial bird nest platform and a bionic bird perch, and the artificial bird nest platform integrates a camera.

[0013] As a further optimization of the multi-layer three-dimensional flight animal control and removal system for power transmission lines of the present invention: the removal system also includes a remote control unit and an energy management unit. The remote control unit communicates with the intelligent decision-making unit to receive and display the system operating status, alarm information and removal records, and provides a remote parameter adjustment and strategy update interface. The energy management unit is used to convert solar energy into electrical energy and store it to power the system.

[0014] This invention also provides a multi-layered, three-dimensional method for controlling and repelling flying animals along power transmission lines. The method utilizes the aforementioned system for repelling animals and includes the following steps: S1. Obtain target information around the protected facility through the panoramic perception unit; S2. Identify flying animals in the target using an intelligent decision-making unit and assess the risk level in real time; S3. Based on the risk level, implement reciprocal expulsion actions: When the risk level is low, continue monitoring and do not initiate a eviction process. When the risk level is medium or high, the mid-level dynamic expulsion module is activated to execute a non-contact expulsion combination strategy corresponding to the risk level, and the expulsion strategy executed at high risk is stronger than that at medium risk. When the risk level is extremely high, the control mid-level dynamic expulsion module operates at the highest intensity, generates the highest priority intrusion alarm information, and uploads the marked on-site evidence data to the remote control terminal to request manual intervention.

[0015] As a further optimization of the multi-layer three-dimensional method for preventing and controlling flying animals in power transmission lines according to the present invention, step S1 specifically includes: The panoramic perception unit acquires the first motion data of the suspected target through a radar detector and obtains the visual image of the corresponding area through an image acquisition device. It then correlates and fuses the first motion data with the visual image to confirm whether the suspected target is a flying animal target. After confirmation, it generates the flight path and identity information of the flying animal target.

[0016] As a further optimization of the multi-layer three-dimensional control method for driving away flying animals in power transmission lines according to the present invention, the method of associating and fusing the first motion data with the visual image is as follows: the data frame acquired by the radar detector and the image acquisition device is timestamped by the hardware synchronization signal, and the spatial coordinate points of the radar detector are mapped to the image pixel coordinate system of the image acquisition device by the pre-established joint calibration parameters.

[0017] As a further optimization of the multi-layer three-dimensional method for controlling and repelling flying animals in power transmission lines according to the present invention, step S2 specifically includes: The intelligent decision-making unit receives the fused data output by the panoramic perception unit, calls the built-in flying animal recognition model to classify the target by species, and simultaneously extracts the key features of the target to determine its behavior pattern. Based on the species classification results of the target, the distance between the target and the protected facility, and the behavior pattern of the target, the species classification results are assigned a first weight coefficient, the distance is assigned a second weight coefficient, and the behavior pattern is assigned a third weight coefficient. The comprehensive risk value is calculated by weighted summation. The overall risk value is compared with multiple preset risk thresholds, and the corresponding low risk, medium risk, high risk or very high risk level is output.

[0018] As a further optimization of the multi-layer three-dimensional prevention and control method for driving away flying animals in power transmission lines according to the present invention, when the behavior pattern of the flying animal target is determined to be an attempt to perch, and its distance from the protected facility has entered the range of extremely high risk threshold, high-definition video and fused flight track data containing the flying animal target's attempt to perch are automatically captured and stored and uploaded as on-site evidence data. At the same time, an emergency intrusion alarm containing real-time images or video clips is pushed to the preset operation and maintenance terminal to prompt the operation and maintenance personnel to conduct remote confirmation or on-site handling.

[0019] This invention offers the following advantages: By constructing a collaborative architecture encompassing an outer layer of ecological guidance, a middle layer of dynamic active repulsion, and an inner layer of static barrier, coupled with AI-specific recognition and multi-dimensional risk assessment for large flying animals, this invention achieves a repulsion strategy that shifts from passive defense to active guidance, and from simple driving away to gradient response. The system randomly combines lasers, variable-frequency sound waves, and predator calls to effectively break the adaptability of flying animals; and it precisely classifies risk levels using hysteresis control and hard judgment mechanisms, automatically triggering emergency alarms and evidence collection in extremely high-risk situations to ensure a safety net. Attached Figure Description

[0020] Figure 1 A schematic diagram of the three-layer three-dimensional prevention and control system of the present invention. Detailed Implementation

[0021] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0022] Example 1

[0023] like Figure 1 As shown, a multi-layered three-dimensional system for preventing and controlling flying animals on power transmission lines includes a panoramic perception unit, an intelligent decision-making unit, and a collaborative execution unit.

[0024] The panoramic perception unit is used to acquire environmental data and flying animal target information around the protected facility in real time. It includes a radar detector, an image acquisition device, and a data fusion module. The data fusion module is used to synchronize the radar detection data and the image acquisition data in time and space to obtain the position and movement status information of the flying animal target.

[0025] The panoramic perception unit operates in two modes: a sleep mode and a wake-up mode. When the radar detector does not detect the suspected target, the panoramic perception unit is in sleep mode, and the image acquisition device is in standby or low-frequency inspection mode. When the radar detector detects the suspected target, the panoramic perception unit is triggered to enter wake-up mode, driving the image acquisition device to point towards the suspected target for locking and acquisition.

[0026] The intelligent decision-making unit is connected to the panoramic perception unit and is used to identify flying animals based on their target information and assess their risk level. It has a built-in flying animal identification model and a risk grading model. The flying animal identification model is used to determine the species and posture of the flying animal based on the perception data, while the risk grading model is used to calculate the risk value based on the species, distance, and behavior pattern of the flying animal and output the corresponding risk level.

[0027] The collaborative execution unit is connected to the intelligent decision-making unit, which includes an outer ecological guidance module, a middle dynamic expulsion module, and an inner static barrier module.

[0028] The outer ecological guidance module is located in an area far from the protected facilities to provide a habitat for flying animals.

[0029] The mid-level dynamic deterrence module is used to actively deter flying animals that approach the protected facility by combining various non-contact deterrence methods according to the risk level.

[0030] The mid-level dynamic deterrence module includes a laser deterrence device and an acoustic deterrence device. The laser deterrence device is used to track the location of flying animals and emit a scanning beam to warn or deter them. The acoustic deterrence device is used to emit variable frequency sound waves and / or play specific biological auditory stimulation signals.

[0031] The mid-level dynamic expulsion module is also equipped with a strategy combination and adaptive mechanism to store multiple expulsion strategy scripts and randomly combine different expulsion methods and modes to reduce the habitual adaptation of flying animals.

[0032] The inner static barrier module is installed on the protected facility to form a physical barrier to prevent flying animals from perching on the protected facility.

[0033] The inner static barrier module includes at least one of bird-proof umbrella cover, bird-proof spikes, bird-proof needle plate and bird-proof wrapping spikes. The bird-proof umbrella cover is used to be installed above the insulator string, and the bird-proof wrapping spikes are flexible structures used to be wrapped around the conductor or ground wire.

[0034] The bird protection umbrella is a split, butt-joint structure that can be installed around the insulator string; its edges have a reverse bending shape to prevent liquid backflow.

[0035] The bird spikes have a retractable structure, which is configured to automatically unfold to a protective state under gravity, retract to a maintenance state when subjected to external pulling force, and automatically reset after the pulling force is removed.

[0036] The outer ecological guidance module includes an artificial bird nest platform and a biomimetic bird perch, and the artificial bird nest platform is equipped with a camera.

[0037] The de-escape system also includes a remote control unit and an energy management unit. The remote control unit communicates with the intelligent decision-making unit to receive and display the system's operating status, alarm information, and de-escape records, and provides an interface for remote parameter adjustment and policy update. The energy management unit is used to convert solar energy into electrical energy and store it to power the system.

[0038] Example 2

[0039] like Figure 1 As shown, a multi-layered, three-dimensional method for controlling and repelling flying animals along power transmission lines utilizes the aforementioned system. This method includes the following steps: S1. Obtain target information around the protected facility through the panoramic perception unit; The panoramic perception unit acquires the first motion data of the suspected target through a radar detector and obtains the visual image of the corresponding area through an image acquisition device. It then correlates and fuses the first motion data with the visual image to confirm whether the suspected target is a flying animal target. After confirmation, it generates the flight path and identity information of the flying animal target.

[0040] The image acquisition device captures images of a designated area and automatically switches to infrared imaging mode in low-light environments to obtain visual images in all weather conditions.

[0041] When a suspected target is identified as a flying animal through visual images from an image acquisition device, it is upgraded to a confirmed target and its radar track and visual characteristics are linked. If no flying animal is identified through visual images and the target is determined to be a non-flying animal or interference object, the track of the suspected target is discarded and tracking is terminated.

[0042] The method for linking and fusing motion data with visual images is as follows: a hardware synchronization signal is used to timestamp each frame of data acquired by the radar detector and the image acquisition device, and the spatial coordinates of the radar detector are mapped to the image pixel coordinate system of the image acquisition device through pre-established joint calibration parameters.

[0043] S2. Identify flying animals in the target area through the intelligent decision-making unit and assess the risk level in real time.

[0044] The intelligent decision-making unit receives the fused data output by the panoramic perception unit, calls the built-in flying animal recognition model to classify the target into species, and simultaneously extracts the key features of the target to determine its behavior pattern.

[0045] The flying animal recognition model is adaptable to adverse weather and complex backgrounds. It is trained using flying animal sample data covering a variety of environmental conditions to maintain recognition capabilities in rain, fog, strong light, or backlight conditions.

[0046] The process of constructing the flying animal identification model is as follows: S201. Construct a dedicated sample library of flying animals that may cause damage to power transmission lines, covering extreme scenarios. This step provides a data foundation for the robustness of the model. The sample library covers multiple species, complex backgrounds, and all operating weather and lighting conditions.

[0047] The target species were selected as 50 harmful flying animals that pose a safety threat to power transmission lines, including Oriental white storks, black storks, magpies, large-billed crows, and common kestrels. The background environment covers typical power transmission scenarios such as power transmission line towers, substation structures, open skies, mountain forests, and farmland. Weather and lighting conditions cover sunny, cloudy, light rain, moderate rain, dense fog, strong light and backlight, twilight and low light, and nighttime infrared imaging modes to achieve full-scene working condition coverage.

[0048] Along multiple 220kV and above voltage level transmission lines operated and maintained in cooperation, 4K high-definition network cameras of the same model as the terminal equipment are deployed, with a frame rate of 15fps. Video data of flying animals are continuously collected in spring, summer, autumn and winter, covering the daytime and nighttime infrared periods. After the collection is completed, key frames in which flying animals appear clearly are manually extracted, and blurry or targetless invalid images are removed.

[0049] Image data matching the aforementioned 50 target species were selected from publicly available flight animal datasets. After deduplication and format standardization, supplementary samples were obtained, which together with real-world scene samples constituted the basic dataset.

[0050] All samples were labeled using the LabelMe annotation tool. The annotations included two types: first, object detection bounding boxes that completely enclosed the flying animal's body, with the corresponding species name as the classification label; second, flying animal keypoint annotations, where five keypoints were annotated for each flying animal: head apex, left wingtip, right wingtip, tail tip, and claw center point, providing supervisory signals for subsequent attitude determination. After annotation, the samples were randomly divided into training, validation, and test sets in an 8:1:1 ratio.

[0051] S202. YOLOv5s is selected as the basic detection network, which includes a backbone network consisting of a Focus slicing layer, a C3 feature extraction module and an SPPF pooling module, as well as a neck feature fusion network with an FPN+PAN structure.

[0052] After the three upsampled output nodes of the YOLOv5s neck PAN network, coordinate attention modules are embedded. These modules encode feature maps along both horizontal and vertical spatial directions, while capturing inter-channel dependencies and spatial location information. Through attention weighting, the model can focus on subtle discriminative features such as claws and beaks of flying animals in scenarios with small targets at a distance and partial occlusion, while suppressing background interference such as clouds and shadows.

[0053] The network end features a parallel dual-detection-head structure that shares features extracted from the backbone and neck, and performs object detection and keypoint prediction tasks respectively: Output three feature maps at different scales. Each feature map outputs the classification probability, bounding box coordinates, and confidence score, enabling species classification and target localization of 50 flying animals.

[0054] Key point detection head: Outputs in parallel at the same scale as the detection head. Each feature map outputs the two-dimensional coordinates of 5 key points of the flying animal, which are used for the accurate determination of the flying / stationary posture of the flying animal in the future.

[0055] The bounding box regression loss uses CIoU loss, the optimizer is AdamW, the weight decay coefficient is set to 0.0005, the learning rate adopts the cosine annealing decay strategy, the initial learning rate is set to 0.001, the minimum learning rate is 0.00001, and the cosine decay is completed during the training period.

[0056] S203. Pre-training stage: YOLOv5s weights pre-trained on the COCO general dataset are used as initial weights to enable the model to learn common primary visual features such as edges, textures, and shapes in advance, thereby accelerating convergence.

[0057] The fine-tuning phase is divided into two stages: the first stage freezes all parameters of the backbone network and trains only the neck network and dual detection heads for 30 epochs to enable the model to quickly adapt to the data distribution of flying animals along power transmission lines; the second stage unfreezes all network parameters, lowers the learning rate to 1 / 10 of the initial value, and performs joint training of the entire network for a total of 60 epochs to avoid catastrophic forgetting of backbone network parameters and suppress overfitting.

[0058] Based on the species classification results of the target, the distance between the target and the protected facility, and the target's behavior pattern, a first weighting coefficient is assigned to the species classification results, a second weighting coefficient is assigned to the distance, and a third weighting coefficient is assigned to the behavior pattern. The comprehensive risk value is calculated by weighted summation.

[0059] Based on the size, activity habits, and potential threat level of flying animals to power facilities, various potentially harmful flying animals are pre-classified into different danger levels, and a corresponding danger coefficient value is set for each danger level.

[0060] The overall risk value is compared with multiple preset risk thresholds, and the corresponding low risk, medium risk, high risk or very high risk level is output.

[0061] Specifically, the following steps are included: S205. Based on the size, activity habits, and potential threat level of flying animals to power transmission line insulators and tower structures, various potentially harmful flying animals are classified into different danger levels, and a corresponding danger factor value is set for each level, forming a pre-set danger factor database: Large, dangerous flying animals, including species with large wingspans such as Oriental White Stork and Black Stork, which are prone to causing phase-to-phase short circuits, are classified as Level 1, with a corresponding risk factor of 0.9. Medium-sized dangerous flying animals, including species such as magpies and crows that nest on poles and perch for extended periods, are classified as Level 2, with a corresponding risk factor of 0.6. Small flying animals or non-harmful flying animals pose an extremely low threat to power transmission lines, with a danger level of three and a corresponding danger coefficient of 0.3.

[0062] If the identification result is that a non-flying animal, such as a drone, has intruded into the target, an independent security alarm process will be triggered, and it will not be included in the calculation scope of this risk assessment model.

[0063] S206. Based on target trajectory fusion data, taking the location of the tower insulator string as the center of the core protection zone, calculate the real-time spatial straight-line distance between the flying animal target and the core protection zone, and assign quantified values ​​according to distance intervals: Close range: distance less than 30m, belongs to the core restricted area, distance factor is assigned to 1.0; Medium distance range: distance greater than or equal to 30m and less than 50m, distance factor is assigned a value of 0.7; Long-distance intervals: distances greater than or equal to 50m and less than 100m, with a distance factor of 0.4; Safe distance range: distance greater than or equal to 100m, distance factor is assigned a value of 0.1.

[0064] S207. Based on the key point coordinate sequence of the flying animal output by the aforementioned key point detection, and combined with the continuous frame trajectory trend, comprehensively determine the behavioral intention of the flying animal and assign corresponding quantitative values: Attempted perching behavior: When actions such as diving, hovering, circling descent, and decelerating claw extension are identified, the behavior pattern coefficient is assigned a value of 0.9; Low-altitude hovering or crossing behavior: When the target makes circular motion or crosses the route area laterally in the low-altitude range, the behavior pattern coefficient is assigned a value of 0.5; High-altitude straight-line flight behavior: When the target flies across the route area from high altitude along a straight trajectory without any tendency to approach, the behavior pattern coefficient is assigned a value of 0.2.

[0065] S208. Weighted Summation Calculation and Conventional Risk Level Classification: The comprehensive risk value is calculated using a weighted summation method, and the calculation formula is as follows:

[0066] In the formula, S is the species risk factor. It is the first weight coefficient and ( =0.4); D is the distance factor, The second weighting coefficient and ( =0.4); B is the behavioral pattern coefficient, It is the third weighting coefficient and ( =0.2).

[0067] The calculated comprehensive risk value R is compared with multiple preset risk thresholds, and initially divided into three conventional levels: low risk, medium risk, and high risk. The corresponding thresholds, typical scenarios, and response logic for each level are as follows: Low risk level: R < 0.45; Medium risk level: meets 0.45 <R<0.65; High risk level: Meets 0.65 <R<0.8; To avoid the lag in the response of weighted scoring, the system has set up an independent channel for determining extremely high risk, as a supplementary mechanism to the regular classification. If any of the following conditions are met, regardless of whether the comprehensive risk value R reaches the corresponding threshold, it will be directly classified as extremely high risk: Intent hard determination condition: The behavior pattern coefficient is 0.9, and the distance factor is greater than or equal to 0.7 at the same time; Distance hard criteria: The distance factor is 1.0, and the species danger coefficient is greater than or equal to 0.6.

[0068] The extremely high risk level also covers conventional computing scenarios with R≥0.8. Typical scenarios include medium to large flying animals diving and attempting to land within a 30m range, and large flying animals hovering and observing within a 50m range. Under this level, the system determines that conventional expulsion methods are at risk of failure, immediately triggers an emergency alarm, collects and preserves on-site evidence, and sends a manual intervention request to the superior operation and maintenance platform.

[0069] To prevent the risk level from repeatedly changing due to the target moving near the threshold boundary, the system introduces a hysteresis control mechanism when outputting the final risk level, setting differentiated judgment rules for upgrading and downgrading: Upgrade judgment rule: When the comprehensive risk value calculation result reaches a higher level threshold, or triggers the extremely high risk hard judgment condition, the system immediately reports the corresponding level alarm information and switches the expulsion response strategy. The level upgrade response time is less than 1 second.

[0070] Downgrade Decision Rules: When the overall risk value falls back to a lower threshold range, the system will not immediately perform a downgrade operation; two decision conditions must be met simultaneously: First, the target continues to move away from the core protected area, meaning that the distance from the core protected area continues to increase and the direction of flight speed is away from the protected area. Second, the overall risk value must remain below the corresponding downgrade threshold for a preset duration, ranging from 30 to 60 seconds. Only after these conditions are met will the system gradually reduce the risk level and adjust the removal strategy accordingly until a low-risk monitoring state is restored.

[0071] S3. Based on the risk level, implement reciprocal expulsion actions: When the risk level output by the intelligent decision-making unit is low risk, the system maintains continuous monitoring and does not activate any of the driving-off execution units of the mid-level dynamic driving-off module; it only keeps the front-end visual acquisition, target recognition and trajectory tracking links running normally, updates the species attributes, spatial distance and behavior patterns of the target in real time, continuously tracks the risk change trend, and triggers the corresponding level of driving-off response when the risk level rises.

[0072] When the risk level is determined to be medium or high, the mid-level dynamic expulsion module is activated to execute a non-contact expulsion combination strategy corresponding to the current risk level. Specifically, the expulsion intensity and density for high-risk levels are higher than those for medium-risk levels. The medium-risk level deportation operation employs a low-intensity warning-based combined deportation scheme. This involves activating the audible and visual warning unit within the mid-level dynamic deportation module, using intermittent low-frequency sound waves combined with low-brightness flashing lights to output a deportation signal. The sound wave frequency is set to 2-3 kHz, with each sound emission lasting 2 seconds and cycling at 3-second intervals. The light flashing frequency is 1 Hz. Each deportation cycle lasts 10 seconds, repeating every 30 seconds until the target moves to a safe distance or the risk level changes. This level prioritizes long-distance warnings to avoid excessive disturbance to the surrounding ecosystem and residential environment.

[0073] For high-risk level expulsion operations, a medium-to-high-efficiency targeted combined expulsion scheme is employed. In addition to the continuous operation of the audible and visual warning unit, a directional acoustic wave expulsion unit and a low-power laser expulsion unit are activated simultaneously. The directional acoustic wave unit adjusts its pitch and horizontal rotation angles according to the target's real-time coordinates, outputting a 115-120 dB directional expulsion acoustic wave towards the target. The low-power laser unit outputs visible laser light in a wavelength sensitive to flying animals, sweeping back and forth along the target's flight path. Each expulsion cycle lasts 15 seconds, and is repeated every 15 seconds. During the expulsion process, the device's direction is dynamically adjusted in real-time according to the target's flight path to achieve precise targeted expulsion and improve its effectiveness.

[0074] When the risk level is determined to be extremely high, the system simultaneously initiates a full-process handling procedure including the highest-intensity expulsion, emergency alarm generation, on-site evidence collection and uploading, and requests for manual intervention. Specifically: The control layer's dynamic deterrence module operates at full power and maximum intensity, with the three execution units—audio-visual warning, directional acoustic wave, and laser deterrence—simultaneously starting at full load. The directional acoustic wave output power is increased to its maximum value, and the laser deterrence unit switches to continuous directional irradiation mode, creating a strong deterrence field around the core protection zone of the tower, prioritizing the flight of targets away from the core area. Simultaneously, intrusion alarm information marked with the highest priority is generated. The alarm information includes structured fields such as target species identification results, real-time spatial location, risk triggering conditions, and current deterrence operation status. The alarm is simultaneously pushed to the remote management platform's monitoring screen and the preset mobile terminals of maintenance personnel, along with a real-time image snapshot of the site, prompting maintenance personnel to respond quickly.

[0075] 4K high-definition video clips, totaling 20 seconds, are extracted 10 seconds before and after the extremely high-risk trigger point. These clips are then simultaneously packaged with target fusion trajectory data, risk level change logs, and species identification confidence data for the corresponding time periods to form a field evidence collection data package with a unified timestamp. This data package is encrypted and stored locally on the terminal storage unit, while simultaneously being uploaded to a remote control terminal via the power grid's dedicated wireless network to retain maintenance and handling documentation.

[0076] When the system determines that the behavior of a flying animal target is an attempt to land, and its distance from the protected power transmission facility has entered the extremely high-risk threshold range, a special refined evidence collection process is triggered: continuous high-definition video containing the complete process of the flying animal diving, decelerating, extending its claws, and attempting to land is automatically captured, and real-time spatial coordinates, flight path direction, species name, and other labeling information are simultaneously superimposed on the video footage; in addition to being encrypted and stored and uploaded to the control platform, the special evidence collection file is directly pushed to the work terminal of the responsible maintenance personnel of the corresponding line, supporting the maintenance personnel to remotely confirm and subsequently handle on-site and trace the source of the fault.

[0077] After receiving an extremely high-risk alert and evidence data, the remote control terminal automatically generates a manual intervention work order and dispatches it to the corresponding maintenance team. If the target still does not leave the core protected area after the expulsion action is executed, the system will continue to push alarm reminders until the maintenance personnel complete remote confirmation or on-site handling.

[0078] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A multi-layered three-dimensional prevention and control flying animal repelling system for a power transmission line, characterized in that: It includes a panoramic perception unit, an intelligent decision-making unit, and a collaborative execution unit; The panoramic perception unit is used to acquire environmental data and flying animal target information around the protected facility in real time. It includes a radar detector, an image acquisition device and a data fusion module. The data fusion module is used to synchronize the radar detection data and the image acquisition data in time and space to obtain the position and motion status information of the flying animal target. The intelligent decision-making unit is connected to the panoramic perception unit and is used to identify flying animals based on flying animal target information and assess risk levels. It has a built-in flying animal identification model and risk grading model. The flying animal identification model is used to determine the species and posture of flying animals based on perception data, and the risk grading model is used to calculate risk values ​​based on flying animal species, distance and behavior patterns, and output the corresponding risk level. The collaborative execution unit is connected to the intelligent decision-making unit, and includes an outer ecological guidance module, a middle dynamic expulsion module, and an inner static barrier module. The outer ecological guidance module is located in an area far from the protected facilities to provide a habitat for flying animals; The mid-level dynamic deterrence module is used to actively deter flying animals that approach the protected facility by combining multiple non-contact deterrence methods according to the risk level. The inner static barrier module is installed on the protected facility to form a physical barrier to prevent flying animals from perching on the protected facility.

2. The multi-tiered aerial animal prevention system of claim 1, wherein, The mid-layer dynamic de-escape module includes a laser de-escape device and an acoustic de-escape device.

3. The multi-layer three-dimensional flight animal control and deterrence system for power transmission lines according to claim 1, characterized in that, The inner static barrier module includes at least one of bird-proof umbrella cover, bird-proof spikes, and bird-proof wrapping spikes. The bird-proof umbrella cover is used to be installed above the insulator string, and the bird-proof wrapping spikes are flexible structures used to be wrapped around the conductor or ground wire.

4. The multi-layer three-dimensional flight animal control and deterrence system for power transmission lines according to claim 1, characterized in that, The outer ecological guidance module includes an artificial bird nest platform and a bionic bird perch, and the artificial bird nest platform integrates a camera.

5. The multi-layer three-dimensional flight animal control and deterrence system for power transmission lines according to claim 1, characterized in that, The expulsion system also includes a remote control unit and an energy management unit. The remote control unit communicates with the intelligent decision-making unit to receive and display the system's operating status, alarm information, and expulsion records, and provides a remote parameter adjustment and strategy update interface. The energy management unit is used to convert solar energy into electrical energy and store it to power the system.

6. A multi-layered, three-dimensional method for controlling and repelling flying animals along power transmission lines, utilizing the system described in any one of claims 1-5 to repel flying animals, characterized in that... The method includes the following steps: S1. Obtain target information around the protected facility through the panoramic perception unit; S2. Identify flying animals in the target using an intelligent decision-making unit and assess the risk level in real time; S3. Based on the risk level, implement reciprocal expulsion actions: When the risk level is low, continue monitoring and do not initiate a eviction process. When the risk level is medium or high, the mid-level dynamic expulsion module is activated to execute a non-contact expulsion combination strategy corresponding to the risk level, and the expulsion strategy executed at high risk is stronger than that at medium risk. When the risk level is extremely high, the control mid-level dynamic expulsion module operates at the highest intensity, generates the highest priority intrusion alarm information, and uploads the marked on-site evidence data to the remote control terminal to request manual intervention.

7. The multi-layered three-dimensional method for controlling and repelling flying animals in power transmission lines according to claim 6, characterized in that, Step S1 is as follows: The panoramic perception unit acquires the first motion data of the suspected target through a radar detector and obtains the visual image of the corresponding area through an image acquisition device. It then correlates and fuses the first motion data with the visual image to confirm whether the suspected target is a flying animal target. After confirmation, it generates the flight path and identity information of the flying animal target.

8. The multi-layered three-dimensional method for controlling and repelling flying animals on power transmission lines according to claim 7, characterized in that, The method for linking and fusing motion data with visual images is as follows: a hardware synchronization signal is used to timestamp each frame of data acquired by the radar detector and the image acquisition device, and the spatial coordinates of the radar detector are mapped to the image pixel coordinate system of the image acquisition device through pre-established joint calibration parameters.

9. The multi-layered three-dimensional method for controlling and repelling flying animals in power transmission lines according to claim 7, characterized in that, Step S2 is as follows: The intelligent decision-making unit receives the fused data output by the panoramic perception unit, calls the built-in flying animal recognition model to classify the target by species, and simultaneously extracts the key features of the target to determine its behavior pattern. Based on the species classification results of the target, the distance between the target and the protected facility, and the behavior pattern of the target, the species classification results are assigned a first weight coefficient, the distance is assigned a second weight coefficient, and the behavior pattern is assigned a third weight coefficient. The comprehensive risk value is calculated by weighted summation. The overall risk value is compared with multiple preset risk thresholds, and the corresponding low risk, medium risk, high risk or very high risk level is output.

10. The multi-layer three-dimensional method for controlling and repelling flying animals in power transmission lines according to claim 9, characterized in that, When the behavior pattern of the flying animal target is determined to be an attempt to perch, and its distance from the protected facility has entered the extremely high risk threshold range, high-definition video and fused flight track data containing the flying animal target's attempt to perch are automatically captured, encrypted, stored and uploaded as on-site evidence data. At the same time, an emergency intrusion alarm containing real-time images or video clips is pushed to the preset operation and maintenance terminal to prompt operation and maintenance personnel to conduct remote confirmation or on-site handling.