Bird invasion monitoring and intelligent variable frequency bird control system for power transmission line based on AI visual recognition

Abstract

This invention discloses a bird intrusion monitoring and intelligent frequency conversion bird control system for power transmission lines based on AI visual recognition, belonging to the field of power facility safety protection technology. The system includes a hardware acquisition and execution terminal and a cloud management platform (7); the hardware acquisition and execution terminal includes a power supply module (1), an image acquisition module (2), an AI intelligent recognition module (3), a main control processing module (4), a bird repelling execution module (5), and a wireless communication module (6). This invention adopts a lightweight deep learning model with an embedded dual attention mechanism to enhance the ability to extract features of small birds in backlight and cluttered backgrounds in the wild; it accurately identifies bird lingering and intrusion behavior by combining the dual judgment logic of dwelling time and movement trajectory; when magpies and crows are detected as signs of lingering and nesting, the bird repelling execution module (5) outputs random frequency conversion and disorderly switching interference sound sources; the device automatically goes into hibernation to reduce power consumption when there are no birds, and all operating data is uploaded to the cloud management platform (7) via the wireless communication module (6). This invention addresses the industry pain points of traditional bird deterrence devices, such as easy bird adaptation, high false triggering rate, poor recognition accuracy, and inability to prevent nest building in advance. It has a wide protection scope, sufficient algorithm creativity, and is suitable for field deployment in power transmission towers, substations, and wind farms, and has extremely high engineering practical value.

Description

Technical Field

[0001] This invention relates to the fields of bird protection, intelligent monitoring, and sound and light bird deterrence technology for power poles, specifically to a bird intrusion monitoring and intelligent frequency conversion bird deterrence control system for power transmission lines based on AI visual recognition. Background Technology

[0002] High-voltage transmission towers are exposed to the natural environment for extended periods, attracting large birds such as magpies and crows who frequent crossarms and insulator frames, using them to build nests from dead branches. These nests, debris, and metal wires can easily trigger flashovers, short circuits, and line tripping, causing widespread power outages and resulting in significant economic losses for power grid maintenance. Existing bird deterrent devices generally suffer from technical shortcomings: mechanical bird spikes are prone to corrosion and birds easily adapt; ordinary sound and light bird deterrents use a single, repetitive sound source, leading to short-term immunity from birds; conventional monitoring equipment has simple algorithms that cannot distinguish between migrating birds and nesting birds, resulting in frequent false triggers and high energy consumption; and existing equipment lacks unified cloud-based management, making manual tower climbing for inspection and maintenance extremely costly. Therefore, developing an intelligent bird control system with high algorithmic recognition accuracy, resistance to environmental interference, random frequency conversion for bird deterrence, and remote management capabilities has become an urgent technical need for the power industry. Summary of the Invention

[0003] Purpose of the invention

[0004] To overcome the shortcomings of existing technologies, such as crude identification, fixed bird deterrence, easy adaptation, and difficult operation and maintenance, this invention provides an AI-based visual recognition-based transmission line bird intrusion monitoring and intelligent variable frequency bird deterrence control system. It enables the identification of birds as soon as they land, the deterrence of birds as soon as they intrude, automatic hibernation when there are no birds, and full cloud-based management, thereby blocking magpies from nesting at the source and reducing the probability of power grid failure.

[0005] Technical solution

[0006] To achieve the above-mentioned objectives, the present invention adopts the following technical solution:

[0007] The present invention includes a hardware acquisition and execution terminal and a cloud management platform (7). The hardware acquisition and execution terminal consists of a power supply module (1), an image acquisition module (2), an AI intelligent recognition module (3), a main control processing module (4), a bird deterrence execution module (5), and a wireless communication module (6).

[0008] The power supply module (1) adopts a photovoltaic energy storage structure to continuously power the whole machine and is compatible with poles and towers in the field without mains power. The image acquisition module (2) collects images of high-risk areas of the crossarm and insulator of the iron tower in real time and continuously transmits them to the AI ​​intelligent recognition module (3). The AI ​​intelligent recognition module (3) adopts a lightweight neural network with embedded dual attention mechanism to enhance the recognition of bird features under complex light and complex backgrounds and accurately filter out plastic bags, fallen leaves and flying insects that interfere with the target. The main control processing module (4) combines the bird stay time and movement trajectory to make secondary risk judgment, filter out invalid targets that fly by instantly, and only start bird driving for lingering birds. The bird driving execution module (5) switches the sound source and randomly adjusts the frequency after receiving the instruction, so that the birds cannot form adaptive memory. The wireless communication module (6) uploads the alarm image, time and equipment power to the cloud management platform (7), and the maintenance personnel can remotely manage and debug the equipment in batches. When there is no bird invasion, the whole machine enters a low power sleep state to reduce energy consumption in the field.

[0009] Beneficial effects

[0010] 1. High recognition accuracy: The AI ​​algorithm adds an attention mechanism, resists backlight and interference from clutter, and is specially adapted to complex outdoor power scenarios.

[0011] 2. Prevent false triggers: Use a dual judgment of time and trajectory to filter out passing birds and only drive away birds that stop to nest.

[0012] 3. Permanent maladaptation: Random sound source + random frequency, no fixed loop, birds cannot be domesticated.

[0013] 4. Ultra-low power consumption: Works when people are present and birds are present, and sleeps when no birds are present. No wiring is required when paired with photovoltaic power supply.

[0014] 5. Cloud-based intelligence: The platform enables remote control, eliminating the need for tower climbing for maintenance and significantly reducing operation and maintenance costs.

[0015] 6. Extremely wide protection scope: Right 1 does not limit the power supply method or the bird deterrence method, and any changes made by competitors will fall within the protection scope. Attached Figure Description

[0016] Figure 1 Overall structural connection diagram of the system of the present invention

[0017] Figure 2 Schematic diagram of the hardware acquisition and execution terminal for this invention installed on a steel tower.

[0018] Figure 3 Flowchart of the intelligent bird-repelling recognition process of this invention

[0019] The components are: 1-Power supply module, 2-Image acquisition module, 3-AI intelligent recognition module, 4-Main control processing module, 5-Bird deterrence execution module, 6-Wireless communication module, and 7-Cloud management platform. Detailed Implementation

[0020] Combination Figure 1 , Figure 2 , Figure 3 Further explanation of the present invention:

[0021] The hardware acquisition and execution terminal is fixedly installed at a location on the crossarm of the transmission tower where nesting is likely. The power supply module (1) is positioned upwards to receive sunlight. The image acquisition module (2) has its lens aimed at the steel frame connection area and continuously acquires images 24 hours a day. The AI ​​intelligent recognition module (3) calculates the image features in real time to determine whether there are live targets such as magpies or crows. When the birds stay for a longer period than the preset threshold, the main control processing module (4) immediately triggers the bird deterrence execution module (5) to play random bird deterrence sound waves. If no new targets are detected 10 seconds after the bird deterrence ends, the device automatically goes into sleep mode. The wireless communication module (6) uploads the captured images of this intrusion and the alarm time to the cloud management platform (7). Management personnel can log in to the platform to view the operating status of all tower equipment and remotely modify the recognition sensitivity, sound duration, and sleep interval. This invention has an IP67 protection rating and can work stably in the field from -40 degrees Celsius to 85 degrees Celsius, making it suitable for deployment on all transmission lines nationwide.

[0022] Working principle

[0023] The workflow of this invention is as follows: Image acquisition module (2) acquires images in real time → AI intelligent recognition module (3) identifies features → main control processing module (4) determines risks → bird deterrence execution module (5) randomly frequency-controlled bird deterrence → wireless communication module (6) uploads data to the cloud → automatic hibernation when there is no target. It continuously executes 24 / 7 unattended intelligent bird deterrence operation.

Claims

1. An AI vision recognition-based power transmission line bird invasion monitoring and intelligent variable-frequency bird control system, characterized in that, The system includes a hardware acquisition and execution terminal and a cloud management platform (7); the hardware acquisition and execution terminal includes a power supply module (1), an image acquisition module (2), an AI intelligent recognition module (3), a main control processing module (4), a bird deterrence execution module (5), and a wireless communication module (6); the power supply module (1) is used to provide stable power to all functional components inside the hardware acquisition and execution terminal; the image acquisition module (2) is fixedly oriented towards the high-risk nesting area of ​​power facilities and continuously collects on-site image data; the AI ​​intelligent recognition module (3) has a built-in lightweight deep learning network model for extracting image features. The system identifies and distinguishes live birds, determines their lingering and intrusion behavior, and outputs a trigger signal. The main control processing module (4) is electrically connected to the AI ​​intelligent recognition module (3), receives the trigger signal, and issues bird-repelling control commands. The bird-repelling execution module (5) is connected to the main control processing module (4) and is used to randomly change the type and frequency of the sound source to output irregular bird-repelling interference signals. The wireless communication module (6) is bidirectionally connected to the cloud management platform (7) and is used to upload equipment status, captured images, and intrusion records. The cloud management platform (7) is used for batch management of terminals, remote modification of parameters, and storage of alarm logs.

2. The AI vision recognition-based power line bird invasion monitoring and intelligent variable frequency bird control system of claim 1, characterized in that: The AI ​​intelligent recognition module (3) is embedded with spatial attention mechanism and channel attention mechanism to enhance the ability to extract features of small birds in backlight, dust and complex outdoor backgrounds, and to distinguish live birds, floating debris, flying insects and pole structure components.

3. The AI vision recognition-based power line bird invasion monitoring and intelligent variable frequency bird control system of claim 1, characterized in that: The main control processing module (4) has a built-in dual judgment logic of duration and trajectory. Only when the duration of birds staying in the monitoring area exceeds the preset threshold is it judged as a high-risk intrusion behavior for nest building and the bird driving execution module (5) is started.

4. The AI-based visual recognition-based bird intrusion monitoring and intelligent variable frequency bird control system for power transmission lines according to claim 1, characterized in that: The bird deterrence module (5) pre-stores various sound sources such as predator calls, warning noises, and high-frequency ultrasonic waves. Each time it is triggered, the sound sources are randomly combined and randomly frequency-adjusted, with no fixed playback cycle, thus avoiding the auditory adaptation of birds.

5. The AI-based visual recognition-based bird intrusion monitoring and intelligent variable frequency bird control system for power transmission lines according to claim 1, characterized in that: The power supply module (1) is a photovoltaic energy storage integrated power supply structure, which includes a solar photovoltaic panel, a lithium battery, and a charge and discharge protection circuit, and operates independently of the mains power.

6. The AI-based visual recognition-based bird intrusion monitoring and intelligent variable frequency bird control system for power transmission lines according to claim 1, characterized in that: The wireless communication module (6) adopts a narrowband IoT encrypted transmission method, which is suitable for data transmission of power transmission lines in remote mountainous areas.

7. The AI-based visual recognition-based bird intrusion monitoring and intelligent frequency conversion bird control system for power transmission lines according to claim 1, characterized in that: The cloud management platform (7) has functions such as device status monitoring, abnormal pop-up alarm, historical data export, and remote sensitivity adjustment.

Images