Intelligent detection device for partial discharge of power equipment based on multispectral imaging

Abstract

A kind of power equipment partial discharge intelligent detection device based on multispectral imaging, including multispectral image acquisition module, detection calculation module and data storage and transmission module;The multispectral image acquisition module is electrically connected with detection calculation module, for collecting the visible light image and ultraviolet light image of power equipment and transmission to detection calculation module;The detection calculation module is electrically connected with data storage and transmission module, for processing to detect partial discharge to the image received, and the detection result is transmitted to data storage and transmission module.The utility model is used to solve the problem of low efficiency and safety risk caused by dependence on artificial judgment in the existing power equipment partial discharge detection, poor detection stability in complex environment, visible light and ultraviolet light image registration difficult, difficult to adapt to multiple scenes.

Description

Technical Field

[0001] This utility model relates to an intelligent detection device for partial discharge of power equipment based on multispectral imaging. Background Technology

[0002] In the field of power equipment condition monitoring, partial discharge detection is a crucial step in ensuring the safe operation of equipment. However, existing partial discharge detection technologies have several shortcomings: Manual discharge detectors rely on personnel using handheld devices to scan at close range, collecting signals through ultrasonic or electric field sensors, and then relying on human experience to determine whether a discharge exists. This method is not only inefficient, but also poses safety risks due to the need for personnel to be in close contact with live equipment. Furthermore, the detection results are affected by the personnel's experience, resulting in insufficient stability and reliability.

[0003] Single-spectrum discharge detection devices are equipped with only single-spectrum sensors such as ultraviolet or visible light sensors. For example, an ultraviolet light sensor can only receive ultraviolet light signals generated by partial discharge. It cannot obtain visible condition information of the equipment surface (such as appearance damage, dirt, etc.), making it difficult to comprehensively assess the equipment's operating status. Furthermore, in scenarios such as strong light and complex electromagnetic environments, single-spectrum signals are easily interfered with, leading to false alarms or missed alarms. Utility Model Content

[0004] The purpose of this invention is to provide an intelligent detection device for partial discharge of power equipment based on multispectral imaging, which solves the problems of low efficiency and safety risks caused by reliance on manual judgment in the detection of partial discharge of power equipment, poor detection stability in complex environments, difficulty in registering visible light and ultraviolet light images, and difficulty in adapting to multiple scenarios.

[0005] To solve the above problems, the technical solution of this utility model is as follows: A smart detection device for partial discharge of power equipment based on multispectral imaging includes a multispectral image acquisition module, a detection calculation module, and a data storage and transmission module; The multispectral image acquisition module is electrically connected to the detection and calculation module, and is used to acquire visible light and ultraviolet light images of the power equipment and transmit them to the detection and calculation module. The detection calculation module is electrically connected to the data storage and transmission module, and is used to process the received image to detect partial discharge and transmit the detection result to the data storage and transmission module. The multispectral image acquisition module has a common optical path structure, including an optical path system. The optical path system is equipped with a beam splitter, which is used to split the incident light into two parts: visible light and ultraviolet light, and the two beams are coaxial.

[0006] Furthermore, the multispectral image acquisition module has an adjustable neutral density filter in front of its lens, which is used to adjust the filter density according to the light intensity.

[0007] Furthermore, the multispectral image acquisition module includes an optical window and a housing. The optical window is made of dustproof and waterproof material, and the housing has a sealed structure.

[0008] Furthermore, a shielding cover is provided on the circuit board of the multispectral image acquisition module, and the electronic components on the circuit board are components with strong anti-electromagnetic interference capabilities.

[0009] Furthermore, the base of the multispectral image acquisition module is equipped with a gimbal, which is used to monitor and compensate for changes in the module's attitude.

[0010] Furthermore, the detection computing module includes an NVIDIA Jetson Nano computing module, which is used for feature extraction, fusion, and partial discharge detection of visible light and ultraviolet light images.

[0011] Furthermore, the data storage and transmission module includes a terminal storage unit, a cloud storage unit, and a transmission unit; the terminal storage unit includes a solid-state drive and an SD card, the cloud storage unit is a distributed storage system, and the transmission unit is a 4G / 5G module.

[0012] The beneficial effects of this utility model are as follows: 1. By using the common optical path and beam splitter structure of the multispectral image acquisition module, visible light and ultraviolet light images are propagated along the same optical axis, eliminating image mismatch caused by parallax and greatly improving image registration accuracy and efficiency.

[0013] 2. The device is designed to withstand strong light overexposure through an adjustable neutral density filter, resist dust and moisture with a sealed dustproof and moisture-proof design, reduce electromagnetic interference with electromagnetic shielding, and compensate for platform vibration with a gimbal to ensure stable operation in complex environments.

[0014] 3. The detection and calculation module is equipped with a processor, which can automatically process images and identify partial discharges, reduce manual intervention, improve detection efficiency and result accuracy, and reduce safety risks.

[0015] 4. Supports deployment on multiple platforms including fixed installations (such as substations), handheld devices, and drones, combined with flexible data storage and transmission strategies to meet the detection needs of different scenarios. Attached Figure Description

[0016] The present invention will be further described below with reference to the accompanying drawings: Figure 1 This is a schematic diagram of the overall structure of this utility model. Figure 2 This is a schematic diagram of the structure of the multispectral image acquisition module of this utility model.

[0017] In the diagram: Neutral density filter 1, dustproof mirror 2, beam splitter 3, image processor 4, edge computer 5, ultraviolet filter 6. Detailed Implementation

[0018] The technical solutions of the present utility model will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present utility model, and not all embodiments. Based on the embodiments of the present utility model, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of the present utility model.

[0019] like Figure 1 and 2 As shown, a smart detection device for partial discharge of power equipment based on multispectral imaging includes a multispectral image acquisition module, a detection calculation module, and a data storage and transmission module. The multispectral image acquisition module is electrically connected to the detection and calculation module, and is used to acquire visible light and ultraviolet light images of the power equipment and transmit them to the detection and calculation module. The detection calculation module is electrically connected to the data storage and transmission module. After receiving the image, the detection calculation module processes the image through its built-in processor, identifies the partial discharge area, and sends the detection result to the data storage and transmission module. The multispectral image acquisition module has a common optical path structure. The module includes an optical path system containing a beam splitter 3 (such as a Thorlabs-BS013). The beam splitter 3 splits the incident light into visible and ultraviolet light components, with both beams coaxial. This common optical path structure enables real-time registration of the two spectral images at the hardware level, avoiding image mismatch caused by parallax in traditional dual-optical-path systems and ensuring accurate correspondence between the discharge location and the surface state of the device during subsequent detection. The data storage and transmission module is connected to the terminal or cloud and is used to store the detection data and transmit it to the terminal or cloud, so as to realize the reliable preservation and remote monitoring of the data.

[0020] Furthermore, the multispectral image acquisition module has an adjustable neutral density filter 1 (such as Edmund Optics' NT59-938) in front of its lens. This neutral density filter 1 is used to adjust the filter density according to the light intensity. In a strong light environment, the filter density increases to reduce the incident light intensity and prevent image overexposure; in a low light environment, the filter density decreases to ensure moderate image brightness, ensuring that images acquired under different lighting conditions are clear and usable.

[0021] Furthermore, the multispectral image acquisition module includes an optical window and a housing. The optical window is made of dustproof and waterproof material (JGS1 type - quartz glass), and the housing has a sealed structure. This structure can effectively prevent dust from entering the optical components, resist the effects of humid environments on the module, ensure the normal operation of the optical system in dusty or high-humidity environments such as outdoors and underground substations, and extend the service life of the device.

[0022] Furthermore, a shielding cover is provided on the circuit board of the multispectral image acquisition module, and the electronic components on the circuit board are components with strong electromagnetic interference resistance. The shielding cover can block external electromagnetic signals from interfering with the circuit board, and the anti-interference components reduce the degree of electromagnetic influence on the internal circuit, enabling the module to stably acquire images in strong electromagnetic environments such as substations, and avoiding data distortion caused by electromagnetic interference.

[0023] Furthermore, the base of the multispectral image acquisition module is equipped with a gimbal (such as the DJI Ronin-SC), which incorporates a gyroscope and accelerometer. The gimbal is used to monitor and compensate for changes in the module's attitude. When the device is deployed on mobile platforms such as drones and handheld devices, the gimbal uses a closed-loop control system to quickly compensate for jitter, ensuring module stability during image acquisition and preventing image blurring or shifting caused by platform shaking, thus guaranteeing the consistency of the acquired data.

[0024] Furthermore, the detection computing module includes an NVIDIA Jetson Nano computing module, which contains a quad-core ARM Cortex-A57 CPU and a 128-core NVIDIA Maxwell GPU. The computing module is used to receive images transmitted by the multispectral image acquisition module, automatically perform feature extraction, fusion and other processing, identify the partial discharge area of ​​the power equipment, and send the results to the data storage and transmission module to realize the automation of the detection process.

[0025] Furthermore, the data storage and transmission module includes a terminal storage unit, a cloud storage unit, and a transmission unit; the terminal storage unit includes a solid-state drive (such as the Samsung 870 EVO) and an SD card; the cloud storage unit is a distributed storage system that achieves high data availability and security through a redundant backup strategy to prevent data loss; the transmission unit is a 4G / 5G module, and the use of 4G / 5G networks for drone inspections ensures efficient data transmission to the cloud or user terminal, enabling remote monitoring and management.

[0026] The working process of this utility model is as follows: The working process of this utility model mainly includes three stages: image acquisition, data processing, and result storage and transmission. Each stage works together to achieve intelligent detection of partial discharge in power equipment. The specific process is as follows: S1. Image Acquisition Stage (based on the multispectral image acquisition module): When the device is deployed on fixed platforms in substations, drones, or handheld devices, the multispectral image acquisition module first receives optical signals reflected from power equipment and generated by partial discharge through a common optical path structure. The beam splitter in the optical path system separates the incident light into visible light (400-760nm) and ultraviolet light (200-400nm), and the two beams propagate along the same optical axis, achieving initial registration at the hardware level.

[0027] In response to environmental interference, the module automatically initiates adaptive adjustments: In a strong light environment, the adjustable neutral density filter in front of the lens will increase the optical density in real time according to the light intensity, reducing the incident light intensity to avoid image overexposure; in a dusty or humid environment, the sealed optical window (made of quartz glass) blocks dust intrusion and isolates moisture, preventing damage to optical components; when the device shakes due to platform movement (such as drone flight), the base gimbal monitors attitude changes through built-in gyroscopes and accelerometers, and drives the motor in real time to compensate for the shaking, ensuring stable image acquisition. Finally, the visible light and ultraviolet light images are geometrically corrected (combining parameters such as lens focal length and image field distortion) and pixel-level aligned by the image processor before being transmitted to the detection and calculation module. 2. Data Processing Stage (based on the detection and calculation module): After receiving image data, the NVIDIA Jetson Nano computing module (quad-core ARM Cortex-A57 CPU + 128-core Maxwell GPU) in the detection computing module starts the preset algorithm process: First, the dual-branch CNN encoder of the MobileNetV2 backbone network extracts the surface features of the device in the visible light image (such as appearance damage and dirt) and the discharge signal features in the ultraviolet light image, respectively; then, the Squeeze-and-Excitation (SE) attention module enhances the two types of features, dynamically increasing the weight of the discharge region features; finally, the multi-level progressive decoder upsamples and refines the fused features, and outputs the detection results such as the location and intensity of the local discharge. The whole process does not require manual intervention and realizes automated analysis. 3. Result Storage and Transmission Stage (based on the data storage and transmission module): After the test results are generated, the data storage and transmission module simultaneously starts a dual storage mechanism: the solid-state drive inside the terminal (such as a 1TB Samsung 870 EVO) prioritizes saving the original images, videos and test results to ensure fast data reading and writing; in the case of temporary inspection scenarios, the SD card (such as a 256GB SanDisk Extreme Pro) simultaneously backs up critical data to prevent loss. Meanwhile, 4G / 5G modules (such as the Huawei ME909s-821 4G module) upload data to a cloud-based distributed storage system (such as Ceph). After receiving the data, the cloud server establishes a real-time connection with the user terminal via the WebSocket protocol, displaying information such as discharge status and device surface condition through a visual interface, supporting remote monitoring and management decisions. Through the above process, the device achieves fully automated operation from multispectral image acquisition to intelligent detection and data interaction. It can adapt to complex environments and efficiently output detection results, meeting the diverse needs of partial discharge monitoring of power equipment.

[0028] The embodiments described in this specification are merely examples of implementations of the inventive concept. The scope of protection of this utility model should not be considered as limited to the specific forms described in the embodiments. The scope of protection of this utility model also extends to equivalent technical means that can be conceived by those skilled in the art based on the inventive concept.

Claims

1. A multi-spectral imaging-based power equipment partial discharge intelligent detection device, characterized in that, It includes a multispectral image acquisition module, a detection and calculation module, and a data storage and transmission module; The multispectral image acquisition module is electrically connected to the detection and calculation module, and is used to acquire visible light and ultraviolet light images of the power equipment and transmit them to the detection and calculation module. The detection calculation module is electrically connected to the data storage and transmission module, and is used to process the received image to detect partial discharge and transmit the detection result to the data storage and transmission module. The multispectral image acquisition module has a common optical path structure, including an optical path system. The optical path system is equipped with a beam splitter, which is used to split the incident light into two parts: visible light and ultraviolet light, and the two beams are coaxial.

2. The power equipment partial discharge intelligent detection device based on multispectral imaging of claim 1, wherein, The multispectral image acquisition module has an adjustable neutral density filter in front of its lens, which is used to adjust the filter density according to the light intensity. 3.The power equipment partial discharge intelligent detection device based on multispectral imaging of claim 1, wherein, The multispectral image acquisition module includes an optical window and a housing. The optical window is made of dustproof and waterproof material, and the housing has a sealed structure.

4. The power equipment partial discharge intelligent detection device based on multispectral imaging of claim 1, wherein, The circuit board of the multispectral image acquisition module is equipped with a shield, and the electronic components on the circuit board are components with strong anti-electromagnetic interference capabilities.

5. The power equipment partial discharge intelligent detection device based on multispectral imaging of claim 1, wherein, The base of the multispectral image acquisition module is equipped with a gimbal, which is used to monitor and compensate for changes in the module's attitude.

6. The power equipment partial discharge intelligent detection device based on multispectral imaging of claim 1, wherein, The detection and computing module includes an NVIDIA Jetson Nano computing module, which is used for feature extraction, fusion, and partial discharge detection of visible light and ultraviolet light images.

7. The multispectral imaging-based power equipment partial discharge intelligent detection device according to claim 1, characterized in that, The data storage and transmission module includes a terminal storage unit, a cloud storage unit, and a transmission unit; the terminal storage unit includes a solid-state drive and an SD card, the cloud storage unit is a distributed storage system, and the transmission unit is a 4G / 5G module.

Images