Metal particle morphological characteristic online diagnosis system, method, equipment and medium

Abstract

The invention provides a metal particle morphological characteristic on-line diagnosis system, method and device and a medium, and relates to the field of power systems, and the system is characterized in that a core photosensitive element of an optical sensing module is a silicon photomultiplier tube array; the direct current power supply module comprises a direct current-direct current conversion circuit and is connected with the optical sensing module; the power supply phase measurement module comprises a voltage dividing capacitor and a voltage transformer, and the voltage dividing capacitor is connected with the voltage transformer; the input end of the data acquisition module is connected with the optical sensing module and the voltage transformer; the edge computing device comprises a data processing module and a particle morphological feature recognition module, the input end of the data processing module is connected with the output end of the data acquisition module, and the output end of the data processing module is connected with the input end of the particle morphological feature recognition module. The particle morphological feature recognition module is used for outputting morphological feature parameters of the metal particles causing creeping discharge.

Description

Technical Field

[0001] This invention relates to the field of power system technology, specifically to an online diagnostic system, method, device, and medium for the morphological characteristics of metal particles. Background Technology

[0002] Gas-insulated switchgear (GIS) is widely used in modern power systems due to its high reliability, compact structure, and excellent insulation performance. However, the stability of the internal insulation performance of GIS is susceptible to various factors, among which contamination by metal particles is one of the main causes of surface discharge and even insulation breakdown. These particles may originate from residues during manufacturing and assembly processes, or from mechanical wear during long-term operation. Under the influence of a strong electric field, metal particles may migrate and adhere to the insulator surface, causing local electric field distortion and subsequently initiating surface discharge, seriously threatening the safe operation of GIS.

[0003] In the field of GIS internal insulation defect detection, traditional detection methods mainly employ ultra-high frequency (UHF) and ultrasonic methods. The UHF method primarily classifies defects by detecting electromagnetic wave signals in the 0.3-1.5 GHz frequency band, while the ultrasonic method identifies defects by capturing sound wave signals in the 20-80 kHz frequency band. In recent years, optical detection technology has received widespread attention due to its non-contact and high sensitivity characteristics. With the development of artificial intelligence technology, intelligent diagnostic methods for partial discharge based on machine learning and deep learning algorithms have become a research hotspot. Current research focuses on identifying the types of GIS insulation defects, but lacks detailed characterization and diagnostic studies for specific defects. There is also a lack of systematic research on the discharge characteristics of different morphologies of metal particles. Simultaneously, existing metal particle detection technologies mainly focus on the detectability and severity diagnosis of particles, lacking methods for accurately identifying the specific shape and size characteristics of particles, which are crucial for assessing discharge risk and predicting insulation degradation trends. Summary of the Invention

[0004] This invention addresses the problems existing in the prior art by providing an online diagnostic system, method, device, and medium for the morphological characteristics of metal particles, which can accurately identify and quantify the specific morphological characteristics of surface-discharged metal particles in GIS.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: The optical sensing module, wherein the core photosensitive element of the optical sensing module is a silicon photomultiplier tube array; A DC power supply module, the DC power supply module including a DC-DC conversion circuit, is connected to the optical sensing module; A power phase measurement module, comprising a voltage divider capacitor and a voltage transformer, wherein the voltage divider capacitor is connected to the voltage transformer; A data acquisition module, the input terminal of which is connected to the optical sensing module and the voltage transformer; An edge computing device includes a data processing module and a particle morphology feature recognition module. The input of the data processing module is connected to the output of the data acquisition module, and the output of the data processing module is connected to the input of the particle morphology feature recognition module. The particle morphology feature recognition module is used to output the morphological feature parameters of metal particles that induce surface discharge.

[0006] In some embodiments, the input terminal of the silicon photomultiplier tube array is connected to the GIS, and the output terminal of the silicon photomultiplier tube array is connected to a low-noise transimpedance amplifier circuit.

[0007] In some embodiments, the input terminal of the voltage divider capacitor is connected to the GIS main circuit, and the output terminal of the voltage divider capacitor is connected to the input terminal of the voltage transformer, which is used to output a low voltage signal that is phase-synchronized with the power supply.

[0008] In some embodiments, the data processing module is used to perform signal preprocessing and construct phase-resolved surface discharge spectrum samples.

[0009] In some embodiments, the particle morphology feature recognition module includes a multi-task learning model for extracting features and recognizing patterns from the phase-resolved surface discharge spectrum samples, establishing a correlation mapping relationship between discharge pulse features and particle morphology, and outputting morphological feature parameters of the metal particles that trigger surface discharge.

[0010] In some embodiments, the data acquisition module is connected to the optical sensing module and the voltage transformer via a 4-channel high-speed synchronous data acquisition card, and to the edge computing device via a USB 3.0 interface.

[0011] In some embodiments, the DC power supply module is used to provide the optical sensing module with a DC operating voltage of 25.2 V to 30.7 V.

[0012] This invention proposes an online diagnostic method for the morphological characteristics of metal particles, comprising: The optical sensing module and the power phase measurement module start synchronously to collect the optical pulse signal generated by surface discharge and the phase signal of the high voltage power supply. The data acquisition module performs synchronous analog-to-digital conversion on the optical pulse signal generated by the surface discharge and the phase signal of the high-voltage power supply at a preset sampling frequency, and transmits the digitized signal data to the edge computing device. The data processing module preprocesses the digitized signal data and constructs phase-resolved surface discharge spectrum samples. The particle morphology feature recognition module performs deep feature extraction and analysis on the phase-resolved surface discharge spectrum sample, and outputs the morphological feature parameters of the metal particles that trigger surface discharge, wherein the morphological feature parameters include shape category and size information.

[0013] This invention proposes a computer device, comprising: At least one processor; and a memory storing a computer program executable on the processor, wherein the processor executes the program to perform the steps of the online diagnostic method for the morphological characteristics of metal particles.

[0014] The present invention proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the online diagnostic method for the morphological characteristics of metal particles.

[0015] Compared with the prior art, the present invention has the following beneficial effects: This invention proposes an online diagnostic system, method, device, and medium for the morphological characteristics of metal particles. The system includes: an optical sensing module, the core photosensitive element of which is a silicon photomultiplier tube array; a DC power supply module, including a DC-DC conversion circuit connected to the optical sensing module; a power phase measurement module, comprising a voltage divider capacitor and a voltage transformer, the voltage divider capacitor being connected to the voltage transformer; a data acquisition module, the input of which is connected to the optical sensing module and the voltage transformer; and an edge computing device, including a data processing module and a particle morphological characteristic recognition module. The input of the data processing module is connected to the output of the data acquisition module, and the output of the data processing module is connected to the input of the particle morphological characteristic recognition module. The particle morphological characteristic recognition module is used to output the morphological characteristic parameters of metal particles that induce surface discharge.

[0016] This invention can diagnose and quantify the specific morphological characteristics of surface-discharged metal particles in GIS online, overcoming the limitations of focusing only on detectability and severity while ignoring the fine identification of characteristic parameters. This provides key technical support for accurately assessing discharge risk and predicting insulation degradation trends, thereby improving the accuracy and reliability of power equipment insulation condition monitoring.

[0017] This invention employs a highly sensitive and interference-resistant silicon photomultiplier tube to detect partial discharge, significantly improving the ability to detect weak partial discharge signals.

[0018] This invention constructs a sample database based on PRSD maps of surface discharge, which can provide an effective basis for identifying the morphological characteristics of particles that induce surface discharge.

[0019] This invention efficiently extracts key features from PRSD maps using a lightweight convolutional module and leverages a multi-task learning model to achieve joint learning and recognition of particle shape and size. Compared to traditional single-task learning models, this device can simultaneously complete shape and size recognition tasks in a single run, demonstrating significant advantages in overall recognition performance and efficiency. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these drawings without creative effort.

[0021] Figure 1 A module diagram of an online diagnostic system for the morphological characteristics of metal particles provided by the present invention.

[0022] Figure 2 The flowchart of an online diagnostic method for the morphological characteristics of metal particles provided by the present invention is shown.

[0023] Figure 3 A schematic diagram of the structure of an embodiment of the computer device provided by the present invention.

[0024] Figure 4 This is a schematic diagram of an embodiment of the computer-readable storage medium provided by the present invention.

[0025] Figure 5 The present invention provides a detailed flowchart of an online diagnostic method for the morphological characteristics of metal particles. Detailed Implementation

[0026] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be construed as limiting the scope of protection of the present invention. It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of this application.

[0027] It should be noted that all uses of "first" and "second" in the embodiments of the present invention are for the purpose of distinguishing two entities or parameters with the same name but different names. It is clear that "first" and "second" are only for the convenience of expression and should not be construed as limiting the embodiments of the present invention. Subsequent embodiments will not explain this in detail.

[0028] This invention proposes an online diagnostic system for the morphological characteristics of metal particles. Please refer to [link / reference]. Figure 1 ,include: The optical sensing module, wherein the core photosensitive element of the optical sensing module is a silicon photomultiplier tube array; A DC power supply module, the DC power supply module including a DC-DC conversion circuit, is connected to the optical sensing module; A power phase measurement module, comprising a voltage divider capacitor and a voltage transformer, wherein the voltage divider capacitor is connected to the voltage transformer; A data acquisition module, the input terminal of which is connected to the optical sensing module and the voltage transformer; An edge computing device includes a data processing module and a particle morphology feature recognition module. The input of the data processing module is connected to the output of the data acquisition module, and the output of the data processing module is connected to the input of the particle morphology feature recognition module. The particle morphology feature recognition module is used to output the morphological feature parameters of metal particles that induce surface discharge.

[0029] This technical solution specifically relates to an online diagnostic device for identifying the morphological characteristics of metal particles in GIS. The device mainly consists of an optical sensing module, a DC power supply module, a power phase measurement module, a data acquisition module, an edge computing device, a data processing module, and a particle morphological feature identification module. Through the collaborative work of these multiple modules, it achieves real-time acquisition, feature extraction, and intelligent identification of surface discharge light pulse signals, ultimately outputting the precise morphological characteristics of the particles. The specific implementation methods of each module are as follows: The optical sensing module uses a high-sensitivity silicon photomultiplier tube (SiPM) array as the core photosensitive element, which possesses single-photon detection capability, high gain, and fast response characteristics, effectively capturing the weak light pulse signals generated by the discharge of metal particles. A low-noise transimpedance amplifier circuit is connected behind the array to convert the weak current signal output by the SiPM into a voltage signal and perform preliminary amplification. This module can operate continuously and stably for extended periods, responsible for real-time acquisition of light pulse signals from surface discharges induced by metal particles in GIS (Gas Insulation System).

[0030] The DC power supply module employs a DC-DC conversion circuit to provide a highly stable, low-ripple DC bias voltage for the optical sensing module. Through multi-stage filtering and shielding, the operating voltage supplied to the SiPM array is ensured to be stable, maintaining the optimal operating state of the optical sensing module and guaranteeing the accuracy and reliability of optical pulse signal acquisition.

[0031] The power supply phase measurement module includes a high-voltage divider capacitor and a precision voltage transformer. The divider capacitor couples in the main circuit of the GIS (Gas Insulation System) to obtain the power frequency high-voltage signal. After signal isolation and proportional transformation by the voltage transformer, it outputs a low-voltage signal that is strictly synchronized with the power supply phase. This module can achieve real-time and accurate measurement of the power frequency high-voltage power supply phase, providing a phase reference for subsequent data processing modules.

[0032] The data acquisition module uses a multi-channel synchronous data acquisition card as its core, featuring high sampling rate and high resolution. The data acquisition card simultaneously receives discharge light pulse signals from the optical sensing module and phase signals from the power phase measurement module, and transmits these signals in real time to the edge computing device for further processing at a specific data sampling rate.

[0033] The edge computing device adopts an embedded industrial computer architecture, integrating a multi-core processor and a high-performance graphics processing unit (GPU) to provide ample computing power. Internally, the device deploys a data processing module and a particle morphology feature recognition module, enabling on-site processing and analysis of data at the edge, meeting the real-time requirements of online monitoring.

[0034] The data processing module uses digital signal processing algorithms to preprocess the collected discharge pulse and phase information, including normalization, pulse extraction, and key feature statistics. Based on this, it constructs a phase-resolved surface discharge spectrum (PRSD). This spectrum includes characteristic dimensions such as discharge amplitude-phase distribution and discharge repetition rate-phase distribution, providing data samples and a database for the particle morphology identification module.

[0035] The particle morphology feature recognition module employs a multi-task learning model to extract features and recognize patterns from PRSD spectrum samples. It establishes a correlation mapping between discharge pulse characteristics and particle morphology, enabling accurate identification and quantification of the shape features and size parameters of metal particles.

[0036] This invention can synchronously identify the shape and size of metal particles that induce surface discharge in GIS in real time. On the test set, its shape recognition accuracy is 99.10%, and its size recognition accuracy is 100%. The average time to complete data processing and analysis is approximately 1.6 seconds, providing an effective technical means for assessing the insulation status of GIS equipment.

[0037] This invention integrates multiple dedicated modules, including optical sensing, stable DC power supply, phase synchronization measurement, multi-channel data acquisition, and edge computing processing, to form an integrated, online diagnostic device. Through the coordinated operation of these modules, the device achieves full automation from signal perception, preprocessing, feature extraction to intelligent recognition, meeting the high reliability and real-time requirements of GIS equipment status monitoring.

[0038] In some embodiments, please refer to Figure 1 The input terminal of the silicon photomultiplier tube array is connected to the GIS, and the output terminal of the silicon photomultiplier tube array is connected to a low-noise transimpedance amplifier circuit.

[0039] The optical sensing module uses a high-sensitivity single-sided array silicon photomultiplier tube with 22,292 micro-units as the core photosensitive element. It covers a wavelength range of 200 nm to 900 nm, with a peak response wavelength of 420 nm and a conversion gain of 4.2 × 10⁻⁶. 6 The array's rear end is connected to a low-noise transimpedance amplifier circuit, which converts the weak current signal output by the SiPM into a voltage signal and amplifies it to 0V-5V. This module is mounted on the optical observation window of a full-scale GIS device via a flange interface, enabling real-time acquisition of surface discharge light pulse signals induced by metal particles in the GIS.

[0040] This invention employs a silicon photomultiplier tube with single-photon detection capability combined with a low-noise transimpedance amplifier circuit to achieve high signal-to-noise ratio capture of weak surface discharge light pulses induced by metal particles in GIS; simultaneously, through a dedicated power phase measurement module, phase information that is strictly synchronized with the discharge signal is obtained, providing an accurate phase reference for constructing PRSD map samples.

[0041] In some embodiments, please refer to Figure 1 The input terminal of the voltage divider capacitor is connected to the GIS main circuit, and the output terminal of the voltage divider capacitor is connected to the input terminal of the voltage transformer. The voltage transformer is used to output a low voltage signal that is phase-synchronized with the power supply.

[0042] The power phase measurement module includes a high-voltage divider capacitor and a precision voltage transformer with an input range of -500 V to 500 V. The divider capacitor obtains the power frequency high-voltage signal from the GIS main circuit. After signal isolation and proportional conversion by the voltage transformer, it outputs a voltage signal of 0 V to 10 V, which is strictly synchronized with the power supply phase.

[0043] The input terminals of the voltage divider capacitors are directly connected to the GIS main circuit, directly obtaining the power frequency high-voltage signal from the main circuit. The GIS main circuit carries the critical power transmission task in the power system, and its power frequency high-voltage signal directly reflects the real-time status of the power supply, ensuring timely and accurate signal acquisition.

[0044] The output of the voltage divider capacitor is connected to the input of the voltage transformer. The voltage transformer provides signal isolation, effectively separating the high-voltage side from the low-voltage side and avoiding potential dangers caused by high voltage. Simultaneously, the voltage transformer also features a proportional conversion function, converting the high-voltage signal from the voltage divider capacitor into a low-voltage signal according to a specific ratio. After conversion, the output voltage signal ranges from 0V to 10V, making it easier for subsequent processing equipment to receive and analyze, greatly improving the convenience and operability of the measurement.

[0045] In some embodiments, please refer to Figure 1 The data processing module is used to perform signal preprocessing and construct phase-resolved surface discharge spectrum samples.

[0046] The edge computing device uses an embedded edge computing development board (model: NVIDIA Jetson TX2) as the hardware core for data processing and analysis. It is equipped with a dual-core NVIDIA Denver2 64-bit CPU, a quad-core Arm Cortex-A57 MPCore processor, and a 256-core NVIDIA Pascal architecture high-performance graphics processing unit (GPU), running the Ubuntu 18.04 operating system. This enables real-time processing, analysis, and display of signal data. Within the embedded edge computing device, digital signal processing algorithms are used to construct PRSD (Pressure Probability of Surge) map samples in real time, incorporating multi-dimensional features such as amplitude-phase and repetition rate-phase. A multi-task learning model is employed to extract deep features from the PRSD map samples, simultaneously and jointly achieving accurate identification and output of the shape, type, and size parameters of metal particles.

[0047] The data processing module, based on the Python 3.11 programming code library, normalizes the optical pulse data and phase data, extracts discharge pulse events by using a sliding translation window (window time span set to 500 ns) and threshold detection method, and finally counts the amplitude and number of discharge events within each phase window (resolution set to 1°) to construct PRSD spectrum samples.

[0048] Normalization can unify data of different dimensions and ranges to a relatively consistent scale, eliminating the differences in magnitude between data.

[0049] The sliding window technique, with a window time span of 500 ns, captures transient features during surface discharge, ensuring no important discharge pulse information is missed. This extracts discharge pulse events from complex data, improving the accuracy and efficiency of discharge pulse identification.

[0050] The amplitude and number of discharge events within each phase window were statistically analyzed, with a phase resolution of 1°. This clearly reflects the specific characteristics and variations of surface discharge at different phases. Based on these statistical results, a PRSD spectrum sample was constructed to analyze the surface discharge phenomenon.

[0051] In some embodiments, please refer to Figure 1 The particle morphology feature recognition module includes a multi-task learning model, which is used to extract features and recognize patterns from the phase-resolved surface discharge spectrum samples, establish the correlation mapping relationship between discharge pulse features and particle morphology, and output the morphological feature parameters of the metal particles that trigger surface discharge.

[0052] The particle morphology feature recognition module is based on the PyTorch 1.12 deep learning codebase. The constructed online diagnostic model for particle morphology features based on multi-task learning mainly consists of three networks: a feature encoding network, a feature extraction network, and a classification network. To fully exploit the edge morphological features of PRSD map samples, the first four convolutional modules in the feature encoding network are composed of 3×3 convolutional layers, ReLU activation functions, and max pooling layers, with 16, 32, 64, and 128 channels, respectively. The fifth lightweight convolutional module contains 18×18 deep convolutional layers, 1×1 point convolutional layers, and ReLU activation functions, which convert the 18×18×128 feature matrix into 1×64 encoded features with lower time and space complexity. In the feature extraction network, task-specific expert A, task-specific expert B, and shared expert all adopt a two-layer fully connected network architecture with ReLU activation functions. Both gate networks A and B, based on a dynamic gating mechanism, contain a single-layer fully connected network with a Softmax activation function to output weighted combined weight values. In the classification network, both tower networks A and B employ a three-layer fully connected network architecture with ReLU and Softmax activation functions, which independently output prediction results for particle shape (e.g., sheet-like, line-like, elliptical sheet-like) and size (e.g., 10 mm, 13 mm, 16 mm).

[0053] Surface discharge patterns contain information about the discharge process. A multi-task learning model simultaneously processes multiple related tasks, mining features from different perspectives and establishing a mapping relationship between discharge pulse characteristics and particle morphology. Surface discharge is closely related to the morphology of metal particles; different particle morphologies lead to different discharge pulse characteristics. This model enables accurate identification of the morphology of metal particles that trigger surface discharge and outputs morphological feature parameters of these particles. These parameters include key information such as particle size, shape, and surface roughness. In some embodiments, please refer to Figure 1The data acquisition module is connected to the optical sensing module and the voltage transformer via a 4-channel high-speed synchronous data acquisition card, and to the edge computing device via a USB 3.0 interface.

[0054] The data acquisition module uses a 4-channel high-speed synchronous data acquisition card with a sampling rate of 41.7 MSa / s and a resolution of 8 bits. The acquisition card simultaneously receives the discharge light pulse signal from the optical sensing module and the power phase signal from the power phase measurement module, and transmits the data to the edge computing device via a USB 3.0 interface.

[0055] The four channels can work simultaneously and in parallel, and can be precisely connected to the optical sensing module and voltage transformer respectively to achieve synchronous acquisition of multiple signals.

[0056] In some embodiments, please refer to Figure 1 The DC power supply module is used to provide a DC operating voltage of 25.2 V to 30.7 V for the optical sensing module.

[0057] The DC power supply module adopts an isolated DC-DC conversion circuit to provide the optical sensing module with a DC operating voltage of 25.2 V~30.7 V, ensuring its stable operation in the complex field environment of GIS.

[0058] Isolation circuits can effectively block the electrical connection between input and output, preventing the optical sensing module from being affected by factors such as voltage fluctuations and interference signals on the input side.

[0059] This invention proposes an online diagnostic method for the morphological characteristics of metal particles. Please refer to [link / reference]. Figure 2 and Figure 5 ,include: S1. The optical sensing module and the power phase measurement module are started synchronously to collect the optical pulse signal generated by the surface discharge and the phase signal of the high voltage power supply. S2. The data acquisition module performs synchronous analog-to-digital conversion on the optical pulse signal generated by the surface discharge and the phase signal of the high-voltage power supply at a preset sampling frequency, and transmits the digitized signal data to the edge computing device. S3. The data processing module preprocesses the digitized signal data and constructs a phase-resolved surface discharge spectrum sample. S4. The particle morphology feature recognition module performs deep feature extraction and analysis on the phase-resolved surface discharge spectrum sample and outputs the morphology feature parameters of the metal particles that trigger surface discharge, wherein the morphology feature parameters include shape category and size information.

[0060] The workflow of the diagnostic system proposed in this invention is as follows: First, after startup, the optical sensing module and the power phase measurement module start synchronously, respectively acquiring the optical pulse signal generated by surface discharge and the phase signal of the high voltage power supply in real time, and ensuring the timing consistency between the signals.

[0061] Subsequently, the data acquisition module performs synchronous analog-to-digital conversion on the aforementioned optical and electrical signals at a preset sampling frequency, and transmits the digitized signal data to the edge computing device.

[0062] Secondly, in the edge computing device, the data processing module performs preprocessing operations such as normalization, pulse extraction and feature statistics on the input data, and on this basis, constructs phase-resolved surface discharge (PRSD) spectrum samples.

[0063] Finally, the particle morphology feature recognition module performs in-depth feature extraction and intelligent analysis on the PRSD spectrum samples, and outputs the morphological feature parameters of the metal particles that trigger surface discharge, including shape category and size information.

[0064] Once the above process is completed, the next round of data collection, processing, and analysis will begin.

[0065] More specifically, after startup, under the stable power supply of the DC power supply module, the optical sensing module acquires the light pulse signal of the surface discharge in real time, while the power phase measurement module simultaneously acquires the power phase signal. The data acquisition module performs analog-to-digital conversion on the two signals at a sampling rate of 41.7 MSa / s and transmits the data to the edge computing device NVIDIA Jetson TX2 via a USB 3.0 interface. In the NVIDIA Jetson TX2, the data processing module preprocesses the received data: first, it performs normalization; then, it extracts the discharge pulses using a sliding window and threshold detection method; finally, it counts the amplitude and number of discharge events within each phase window to construct a PRSD map sample. The particle morphology feature recognition module uses the PRSD map sample as input to the online diagnostic model for particle morphology features. It performs deep feature extraction and analysis through a feature encoding network, a feature extraction network, and a classification network, and outputs the predicted results of particle morphology features. After a single analysis is completed, it automatically enters the next working cycle, realizing continuous online monitoring of the GIS equipment.

[0066] Based on the same inventive concept, according to another aspect of the present invention, such as Figure 3 As shown, an embodiment of the present invention also provides a computer device 30, which includes a processor 310 and a memory 320. The memory 320 stores a computer program 321 that can run on the processor. When the processor 310 executes the program, it performs the steps of the system described above.

[0067] Based on the same inventive concept, according to another aspect of the present invention, such as Figure 4 As shown, embodiments of the present invention also provide a computer-readable storage medium 40, which stores a computer program 410 that executes the system described above when executed by a processor.

[0068] Embodiments of the present invention may also include corresponding computer devices. The computer device includes a memory, at least one processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it performs any of the aforementioned systems.

[0069] The memory, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as program instructions / modules in the embodiments of this application. The processor executes various functional applications and data processing of the device by running the non-volatile software programs, instructions, and modules stored in the memory, thereby realizing the above-described system.

[0070] The memory may include a program storage area and a data storage area, wherein the program storage area may store operating methods and application programs required for at least one function; the data storage area may store data created based on the use of the device, etc. Furthermore, the memory may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In embodiments, the memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the local module via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0071] Finally, it should be noted that those skilled in the art will understand that all or part of the processes in the systems described above can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the embodiments of the systems above. The storage medium for the program can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc. The embodiments of the computer program described above can achieve the same or similar effects as any of the corresponding system embodiments.

[0072] Those skilled in the art will also understand that the various exemplary logic blocks, modules, circuits, and algorithm steps described in conjunction with the disclosure herein can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this interchangeability between hardware and software, the functionality of the various illustrative components, blocks, modules, circuits, and steps has been generally described. Whether this functionality is implemented as software or as hardware depends on the specific application and the design constraints imposed on the overall method. Those skilled in the art can implement the functionality in various ways for each specific application, but such implementation decisions should not be construed as departing from the scope of the embodiments disclosed herein.

[0073] The above are exemplary embodiments disclosed in this invention. However, it should be noted that various changes and modifications can be made without departing from the scope of the embodiments of this invention as defined by the claims. The functions, steps, and / or actions of the system claims according to the disclosed embodiments described herein do not need to be performed in any particular order. The sequence numbers of the disclosed embodiments of this invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments. Furthermore, although the elements disclosed in the embodiments of this invention may be described or claimed individually, they may be understood as multiple unless explicitly limited to a singular number.

[0074] It should be understood that, as used herein, the singular form “a” is intended to include the plural form as well, unless the context clearly supports an exception. It should also be understood that, as used herein, “and / or” refers to any and all possible combinations of one or more of the associated listed items.

[0075] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of the invention (including the claims) is limited to these examples. Within the framework of the invention, technical features of the above embodiments or different embodiments can be combined, and many other variations of different aspects of the invention exist, which are not provided in the details for the sake of brevity. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the invention should be included within the protection scope of the invention.

Claims

1. A metal fine particle morphology characteristic on-line diagnosis system characterized by comprising: It comprises: an optical sensing module, the core photosensitive element of which is a silicon photomultiplier array; a direct current power supply module, which comprises a direct current-direct current conversion circuit connected to the optical sensing module; a power supply phase measurement module, which comprises a voltage divider capacitor and a voltage transformer, the voltage divider capacitor being connected to the voltage transformer; a data acquisition module, the input end of which is connected to the optical sensing module and the voltage transformer; an edge computing device, which comprises a data processing module and a particle morphology feature recognition module, the input end of the data processing module being connected to the output end of the data acquisition module, the output end of the data processing module being connected to the input end of the particle morphology feature recognition module, and the particle morphology feature recognition module being used to output the morphology feature parameters of the metal particles that induce the surface discharge.

2. The metal particle morphology online diagnostic system of claim 1, wherein, The input end of the silicon photomultiplier array is connected to a GIS, and the output end of the silicon photomultiplier array is connected to a low-noise transimpedance amplification circuit.

3. The metal particle morphology online diagnostic system of claim 1, wherein, The input end of the voltage divider capacitor is connected to a GIS main circuit, and the output end of the voltage divider capacitor is connected to the input end of the voltage transformer, which is used to output a low-voltage signal synchronized with the power supply phase.

4. The metal particle morphology online diagnostic system of claim 1, wherein The data processing module is used to perform signal preprocessing and construct a phase-resolved surface discharge spectrum sample.

5. The metal particle morphology online diagnostic system of claim 4, wherein, The particle morphology feature recognition module comprises a multi-task learning model, which is used to perform feature extraction and pattern recognition on the phase-resolved surface discharge spectrum sample, establish a correlation mapping relationship between the discharge pulse feature and the particle morphology, and output the morphology feature parameters of the metal particles that induce the surface discharge.

6. The metal particle morphology online diagnostic system of claim 1, wherein The data acquisition module is connected to the optical sensing module and the voltage transformer through a 4-channel high-speed synchronous data acquisition card, and is connected to the edge computing device through a USB 3.0 interface.

7. The metal particle morphology online diagnostic system of claim 1, wherein The direct current power supply module is used to provide a direct current working voltage of 25.2 V~30.7 V for the optical sensing module.

8. A method of on-line diagnosis of metal particle morphology characteristics, characterized by, It comprises: The optical sensing module and the power supply phase measurement module are started synchronously to collect the light pulse signals generated by the surface discharge and the phase signals of the high-voltage power supply. The data acquisition module performs synchronous analog-to-digital conversion on the light pulse signals generated by the surface discharge and the phase signals of the high-voltage power supply at a preset sampling frequency, and transmits the digitized signal data to the edge computing device. The data processing module pre-processes the digitized signal data and constructs a phase-resolved surface discharge spectrum sample. The particle morphology feature recognition module performs deep feature extraction and analysis on the phase-resolved surface discharge spectrum sample and outputs the morphology feature parameters of the metal particles that induce the surface discharge, wherein the morphology feature parameters include shape categories and size information.

9. A computer device comprising: at least one processor; and a memory, the memory storing a computer program executable on the processor, characterized in that the processor executes the program to perform the steps of the metal particle morphology feature online diagnosis method according to claim 8.

10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. The computer program is executed by a processor to perform the steps of the metal particle morphology characteristic online diagnosis method according to claim 8.

Images