Substation equipment data acquisition and state analysis model system based on side-end intelligence

Abstract

The invention discloses a substation equipment data acquisition and state analysis model system based on side-end intelligence, which comprises a data acquisition device, a side-end intelligent control processing device and a state analysis model, and is characterized in that the data acquisition device transmits data to the side-end intelligent control processing device through a communication networking protocol; the side-end intelligent control processing device summarizes and processes data acquired by the data acquisition device through an embedded computer or an industrial controller, and the state analysis model establishes a fault diagnosis model according to the data processed by the side-end intelligent control processing device to realize fault early warning and fault analysis on related equipment. According to the system, an automatic monitoring mode is used for replacing manual inspection, the working efficiency is improved, the collected data is deeply analyzed through the side end intelligent control processing device and the state analysis model, a scientific basis is provided for operation and maintenance, a maintenance plan can be made in advance, and the safety and stability of equipment operation are improved.

Description

Technical Field

[0001] The present invention relates to the field of substation equipment operation and maintenance technology, and in particular to a substation equipment data collection and status analysis model system based on edge intelligence. Background Art

[0002] At present, the maintenance method of key equipment in substations (such as main transformers, circuit breakers, and disconnectors) is usually "pre-test inspection" or "fault maintenance", which relies on monthly manual inspections or waiting for relay alarms. It is unable to capture changes in gas content in a timely manner, especially temporary "fault maintenance", which has caused many equipment to stop operating unplanned, increased the passivity of work, and seriously affected the reliability of power supply. At the same time, the online information collection form of the equipment is single, mainly relying on automatic measurement and control devices to collect telemetry information such as "current, voltage, and power" and telesignal information such as "switch position information", resulting in incomplete collection of equipment operation status information, and no systematic "big data correlation analysis" has been formed. The "online health" of the equipment is insufficiently evaluated, resulting in a lack of "early warning" and "precise intelligent diagnosis" of equipment failures, and insufficient guidance for production maintenance plans and processes. Summary of the invention

[0003] In order to overcome the above problems, the purpose of the present invention is to provide a substation equipment data acquisition and status analysis model system based on edge intelligence. The system realizes regular or continuous data acquisition through a data acquisition device, improves the monitoring frequency, replaces manual inspections with automated monitoring methods, improves work efficiency, and reduces labor intensity. The collected data is deeply analyzed through the edge intelligent control processing device and the status analysis model, which provides a scientific basis for operation and maintenance, can formulate maintenance plans in advance, and improves the safety and stability of equipment operation.

[0004] The technical solution adopted by the present invention is:

[0005] The substation equipment data collection and status analysis model system based on edge intelligence includes a data collection device, an edge intelligent control processing device, and a status analysis model. The data collection device transmits data to the edge intelligent control processing device through a communication networking protocol. The edge intelligent control processing device summarizes and processes the data collected by the data collection device through an embedded computer or an industrial controller to achieve data conversion, encryption and transmission. The status analysis model establishes a fault diagnosis model based on the data processed by the edge intelligent control processing device to achieve fault warning and fault analysis for related equipment.

[0006] The data acquisition device includes a gas content change data acquisition device for a gas relay, a transformer core leakage current data acquisition device, Equipment gas density monitoring sensor device, circuit breaker hydraulic energy storage mechanism pressure online monitoring sensor device, disconnector temperature monitoring sensor device, circuit breaker opening and closing coil resistance collection sensor device,

[0007] The edge intelligent control processing device includes a gas relay receiving gateway, an 8-channel RS485 gateway, a 4-channel RS485 gateway, an RS485 to TCP transparent transmission device, a networking router, an embedded computer, and a power module. The gas relay receiving gateway receives the image information collected by the gas content change data acquisition device of the gas relay. The 8-channel RS485 gateway receives data from the pressure online monitoring sensor device of the circuit breaker hydraulic energy storage mechanism and the circuit breaker opening and closing coil resistance acquisition sensor device. The 4-channel RS485 gateway receives data from the isolation switch temperature monitoring sensor device, the transformer core leakage current data acquisition device, The data of the equipment gas density monitoring sensor device, the gas relay receiving gateway, the 8-channel RS485 gateway transmits the data to the networking router, the 4-channel RS485 transmits the data to the networking router through the RS485 to TCP transparent transmission device, the networking router transmits the data to the embedded computer, and the power module provides power for the edge intelligent control processing device.

[0008] As a further description of the present invention, the software processing flow in the embedded computer of the edge intelligent exchange control processing device includes the following steps:

[0009] S01: Data collection, collecting various sensor data through RS485 gateway and RS485 to TCP transparent transmission device;

[0010] S02: Data processing, summarizing, converting, encrypting and transmitting the collected data;

[0011] S03: Data cleaning, screening out data that is not within the normal range or empty data or redundant data;

[0012] S04: Data transmission, transmitting the processed data to the backend server through the network;

[0013] S05: Visualization interface, which displays and analyzes data through the visualization interface on the background server.

[0014] As a further description of the present invention, the gas relay receiving gateway uses a Raspberry Pi 5A processor, is connected to an ESP32 camera, uses the OpenCV library to receive and analyze images from the ESP32 camera, and transmits the analysis results to an embedded computer via a network.

[0015] The embedded computer creates a Socket connection thereon, connects to the RS485 to TCP transparent transmission device, connects the RS485 to TCP transparent transmission device and the gas relay receiving gateway to the same network through a networking router, and configures an IP address and a port number.

[0016] As a further description of the present invention, the data processing in step S02 uses the struct module to convert the received binary data into floating point numbers;

[0017] The specific steps of data cleaning in step S03 are: first, define the normal range of each sensor data according to the specifications and actual needs of the equipment, screen out the data that is not within the normal range or empty data and redundant data, then traverse the converted data to check whether each data is within the normal range, and do not save the data that is not within the normal range;

[0018] In step S04, data transmission is performed by hybrid encryption and then transmitted to the backend server via the TCP protocol.

[0019] As a further description of the present invention, the steps of establishing the state analysis model are as follows:

[0020] Step 1: Determine the data source and collection indicators, including transformer electrical parameters, oil surface temperature, winding temperature, core leakage current, gas content of gas relay, characteristic gas composition of oil chromatogram, real-time liquid pressure, gas density, coil resistance value of circuit breaker, and temperature of disconnector;

[0021] Step 2: Classify all data sources into:

[0022] Electrical parameters: voltage, current, active power, reactive power,

[0023] Temperature: oil level temperature, winding temperature, disconnect switch temperature,

[0024] Electrical characteristics: core leakage current, coil resistance,

[0025] Chemical composition: gas content of gas relay, characteristic gas composition of oil chromatography,

[0026] Mechanical properties: real-time liquid pressure, gas density;

[0027] Step 3: Data preprocessing, including data cleaning, data standardization, and data fusion;

[0028] Step 4: Feature extraction, extracting time series features and combination features;

[0029] Step 5: Feature selection, using correlation analysis and principal component analysis methods to select the key features that best reflect changes in equipment status;

[0030] Step 6: Model selection: Select a suitable machine learning model;

[0031] Step 7: Model training: Use historical data sets to train the model, use k-fold cross validation to evaluate model performance, and tune hyperparameters through grid search or random search;

[0032] Step 8: Model validation;

[0033] Step 9: Real-time monitoring, input the newly collected data into the trained model for real-time prediction;

[0034] Step 10: Continuously optimize, update the model regularly, and retrain and evaluate the model using new data.

[0035] As a further description of the present invention, the gas content change data acquisition device of the gas relay adopts an ESP32-CAM module, which integrates a camera and a Wi-Fi function, and is used to take pictures of the observation window of the gas relay, and transmit the taken pictures to the edge intelligent control processing device or the backend server via Wi-Fi.

[0036] The software processing process of the gas content change data acquisition device of the gas relay is: image acquisition→selection of ROI area→image enhancement and noise reduction→marginalization monitoring→image graying→extraction of gas relay liquid level.

[0037] As a further description of the present invention, the transformer core leakage current data acquisition device, The software processing process of the equipment gas density monitoring sensor device, the circuit breaker hydraulic energy storage mechanism pressure online monitoring sensor device, the isolation switch temperature monitoring sensor device, and the circuit breaker opening and closing coil resistance acquisition sensor device are all:

[0038] Data acquisition (initialization, timing acquisition) → data processing (filtering, calibration) → data transmission (data packaging, data sending).

[0039] As a further description of the present invention, the hardware of the transformer core leakage current data acquisition device adopts a Hall effect sensor. The gas density monitoring sensor device of the equipment adopts a gas density sensor, the pressure online monitoring sensor device of the hydraulic energy storage mechanism of the circuit breaker adopts a pressure sensor, the temperature monitoring sensor device of the isolating switch adopts an infrared temperature sensor, and the circuit breaker opening and closing coil resistance collection sensor device adopts a coil resistance sensor.

[0040] The above-mentioned acquisition devices also include a signal conditioning circuit, an analog-to-digital converter, a microcontroller, a wireless transmission module, and a power module. The data collected by each sensor in the acquisition device are respectively connected to the corresponding signal conditioning circuit. The signal conditioning circuit amplifies and conditions the weak signal output by the sensor to make it suitable for processing. The analog-to-digital converter converts the conditioned analog signal into a digital signal. The microcontroller processes the digital signal and performs data processing and transmission. The wireless transmission module transmits data to the edge intelligent control processing device or the backend server via Wi-Fi, LoRa or Bluetooth wireless communication. The power module provides a stable battery power supply to ensure long-term stable operation of the equipment.

[0041] Beneficial effects of the present invention:

[0042] The present invention is based on the edge-end intelligent substation equipment data collection and status analysis model system, including a data collection device, an edge-end intelligent control processing device, and a status analysis model. The system realizes regular or continuous data collection through the data collection device, improves the monitoring frequency, replaces manual inspections with automated monitoring methods, improves work efficiency, and reduces labor intensity. The collected data is deeply analyzed through the edge-end intelligent control processing device and the status analysis model, which provides a scientific basis for operation and maintenance. Maintenance plans can be formulated in advance, early warning issues can be responded to in a timely manner, emergency shutdowns caused by sudden failures can be reduced, and the safety and stability of equipment operation can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is the overall principle diagram of the substation equipment data collection and status analysis model system based on edge intelligence proposed by the present invention;

[0044] Figure 2 It is a logic diagram of the data acquisition device of the substation equipment data acquisition and status analysis model system based on edge intelligence and the edge intelligent control processing device proposed by the present invention;

[0045] Figure 3 A software processing flow chart of the edge intelligent control processing device of the substation equipment data acquisition and status analysis model system based on edge intelligence proposed by the present invention;

[0046] Figure 4 The data transmission circuit design diagram of the data acquisition device of the substation equipment data acquisition and status analysis model system based on edge intelligence proposed by the present invention;

[0047] Figure 5 This is a wiring diagram of a gas content image acquisition camera for a gas relay in a substation equipment data acquisition and state analysis model system based on edge intelligence proposed by the present invention;

[0048] Figure 6 This is a diagram showing the overall interface of the state analysis model of the substation equipment data acquisition and state analysis model system based on edge intelligence proposed by the present invention;

[0049] Figure 7 This is a display diagram of the transformer digital analysis model interface of the substation equipment data acquisition and status analysis model system based on edge intelligence proposed by the present invention;

[0050] Figure 8 This is a display diagram of the circuit breaker digital analysis model interface of the substation equipment data acquisition and state analysis model system based on edge intelligence proposed by the present invention;

[0051] Fig. 9 This is a display diagram of the digital analysis model interface of the isolating switch of the substation equipment data acquisition and state analysis model system based on edge intelligence proposed by the present invention;

[0052] Fig.10 This is a physical diagram of the substation equipment data collection and status analysis model system based on edge intelligence proposed by the present invention;

[0053] Fig.11 This is a software processing flow chart of the gas content change data acquisition device of the gas relay in the edge-end intelligent substation equipment data acquisition and state analysis model system proposed by the present invention. DETAILED DESCRIPTION

[0054] The specific implementation of the present invention is described below in conjunction with the accompanying drawings and embodiments:

[0055] It should be noted that the structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the conditions under which the present invention can be implemented. Any structural modification, change in proportional relationship or adjustment of size should still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.

[0056] At the same time, the terms such as "upper", "lower", "left", "right", "middle" and "one" cited in this specification are only for the convenience of description and are not used to limit the scope of implementation of the present invention. Changes or adjustments to their relative relationships should be regarded as the scope of implementation of the present invention without substantially changing the technical content.

[0057] like Figures 1 to 11 As shown, it shows a specific embodiment of the present invention:

[0058] Example

[0059] The substation equipment data collection and status analysis model system based on edge intelligence disclosed in the present invention includes a data collection device, an edge intelligent control processing device, and a status analysis model. The data collection device transmits data to the edge intelligent control processing device through a communication networking protocol. The edge intelligent control processing device summarizes and processes the data collected by the data collection device through an embedded computer or an industrial controller to achieve data conversion, encryption and transmission. The status analysis model establishes a fault diagnosis model based on the data processed by the edge intelligent control processing device to achieve fault warning and fault analysis for related equipment.

[0060] In this embodiment, if Figure 1 As shown in the figure, the system realizes regular or continuous data collection through data acquisition devices, improves the monitoring frequency, replaces manual inspections with automated monitoring methods, improves work efficiency, reduces labor intensity, and conducts in-depth analysis of the collected data through edge intelligent control processing devices and status analysis models, providing a scientific basis for operation and maintenance. Maintenance plans can be formulated in advance, early warning issues can be responded to in a timely manner, emergency shutdowns caused by sudden failures can be reduced, and the safety and stability of equipment operation can be improved.

[0061] The data acquisition device includes a gas content change data acquisition device for a gas relay, a transformer core leakage current data acquisition device, Equipment gas density monitoring sensor device, circuit breaker hydraulic energy storage mechanism pressure online monitoring sensor device, disconnector temperature monitoring sensor device, circuit breaker opening and closing coil resistance collection sensor device,

[0062] In this embodiment, the data acquisition device selected from the above six devices is designed for key equipment (transformer, circuit breaker, disconnector) in the substation, and can monitor the parameters of the substation in real time to ensure stable and safe operation of the power grid.

[0063] Aiming at the typical fault temperature monitoring defects of transformers and gas signal action, we collect data on electrical parameters (current, voltage, active and reactive power), oil surface temperature, winding temperature, core leakage current, gas content of gas relays, characteristic gas components of oil chromatography, etc., and develop corresponding wireless transmission sensor collection front end, i.e. gas content change data collection device for gas relays. Figure 5 As shown, the transformer core leakage current data acquisition device, The equipment's gas density monitoring sensor device can directly collect and use data if it can already collect data on site, without the need to develop sensors.

[0064] Aiming at the typical faults of circuit breakers, such as frequent hydraulic mechanism pressure, low air pressure alarm and burning of opening and closing coils, the real-time liquid pressure, gas density and coil resistance data are collected, and the corresponding wireless transmission sensor collection front end is developed, namely, the circuit breaker hydraulic energy storage mechanism pressure online monitoring sensor device and the circuit breaker opening and closing coil resistance collection sensor device.

[0065] Aiming at the heating problem of the isolating switch, temperature collection is carried out, and a corresponding wireless transmission sensor collection front end, namely the isolating switch temperature monitoring sensor device, is developed.

[0066] The edge intelligent control processing device includes a gas relay receiving gateway, an 8-channel RS485 gateway, a 4-channel RS485 gateway, an RS485 to TCP transparent transmission device, a networking router, an embedded computer, and a power module. The gas relay receiving gateway receives the image information collected by the gas content change data acquisition device of the gas relay. The 8-channel RS485 gateway receives data from the pressure online monitoring sensor device of the circuit breaker hydraulic energy storage mechanism and the circuit breaker opening and closing coil resistance acquisition sensor device. The 4-channel RS485 gateway receives data from the isolation switch temperature monitoring sensor device, the transformer core leakage current data acquisition device, The data of the equipment gas density monitoring sensor device, the gas relay receiving gateway, the 8-channel RS485 gateway transmits the data to the networking router, the 4-channel RS485 transmits the data to the networking router through the RS485 to TCP transparent transmission device, the networking router transmits the data to the embedded computer, and the power module provides power for the edge intelligent control processing device.

[0067] In this embodiment, if Figure 2 As shown, the data collected by the data acquisition device is transmitted and processed by the edge intelligent control processing device, providing a real and reliable data source for the establishment of the state analysis model.

[0068] Specifically, Figure 3 As shown, the software processing flow in the embedded computer of the edge intelligent exchange control processing device includes the following steps:

[0069] S01: Data collection, collecting various sensor data through RS485 gateway and RS485 to TCP transparent transmission device;

[0070] S02: Data processing, summarizing, converting, encrypting and transmitting the collected data;

[0071] S03: Data cleaning, screening out data that is not within the normal range or empty data or redundant data. For example, if the temperature value is less than 0°C or greater than 100°C, the data is considered abnormal and will not be saved; if the current value is less than 0A or greater than 100A, the data is considered abnormal and will not be saved, etc.

[0072] S04: Data transmission, transmitting the processed data to the backend server through the network;

[0073] S05: Visualization interface, which displays and analyzes data through the visualization interface on the background server.

[0074] Specifically, the gas relay receiving gateway uses a Raspberry Pi 5A processor, connects to the ESP32 camera, uses the OpenCV library to receive and analyze the images from the ESP32 camera, and transmits the analysis results to the embedded computer through the network.

[0075] The embedded computer creates a Socket connection thereon, connects to the RS485 to TCP transparent transmission device, connects the RS485 to TCP transparent transmission device and the gas relay receiving gateway to the same network through a networking router, and configures an IP address and a port number.

[0076] Specifically, the data processing in step S02 uses the struct module to convert the received binary data into floating point numbers;

[0077] The specific steps of data cleaning in step S03 are: first, define the normal range of each sensor data according to the specifications and actual needs of the equipment, screen out the data that is not within the normal range or empty data and redundant data, then traverse the converted data to check whether each data is within the normal range, and do not save the data that is not within the normal range;

[0078] In step S04, data transmission is performed by hybrid encryption and then transmitted to the backend server via the TCP protocol.

[0079] Specifically, the steps for establishing the state analysis model are as follows:

[0080] Step 1: Determine the data source and collection indicators, including transformer electrical parameters, oil surface temperature, winding temperature, core leakage current, gas content of gas relay, characteristic gas composition of oil chromatogram, real-time liquid pressure, gas density, coil resistance value of circuit breaker, and temperature of disconnector;

[0081] Step 2: Classify all data sources into:

[0082] Electrical parameters: voltage, current, active power, reactive power,

[0083] Temperature: oil level temperature, winding temperature, disconnect switch temperature,

[0084] Electrical characteristics: core leakage current, coil resistance,

[0085] Chemical composition: gas content of gas relay, characteristic gas composition of oil chromatography,

[0086] Mechanical properties: real-time liquid pressure, gas density;

[0087] Step 3: Data preprocessing, including data cleaning, removing noise and outliers,

[0088] Data standardization, also known as normalized data, ensures that data of different dimensions are on the same scale.

[0089] Data fusion integrates the cleaned and standardized data into a structured data set for further analysis and modeling.

[0090] The sample program of data fusion is as follows:

[0091] {

[0092] "data": [

[0093] {

[0094] "timestamp": "2023-11-28T09:00:00Z",

[0095] "device_id": "T1",

[0096] "electrical_parameters": {

[0097] "current": 100.2,

[0098] "voltage": 110.5,

[0099] "active_power": 100.0,

[0100] "reactive_power": 50.0

[0101] },

[0102] "temperature": {

[0103] "oil_temperature": 60.0,

[0104] "winding_temperature": 70.0,

[0105] "switch_temperature": 55.0

[0106] },

[0107] "electrical_characteristics": {

[0108] "core_leakage_current": 0.5,

[0109] "coil_resistance": 1.2

[0110] },

[0111] "chemical_composition": {

[0112] "gas_content": {

[0113] "H2": 10,

[0114] "C2H2": 2,

[0115] "CH4": 30,

[0116] "C2H4": 5

[0117] }

[0118] },

[0119] "mechanical_characteristics": {

[0120] "liquid_pressure": 1000000,

[0121] "gas_density": 1.25

[0122] },

[0123] "other_information": {

[0124] "environmental_temperature": 25.0,

[0125] "humidity": 60

[0126] }

[0127] },

[0128] {

[0129] "timestamp": "2023-11-28T09:01:00Z",

[0130] "device_id": "T1",

[0131] "electrical_parameters": {

[0132] "current": 100.3,

[0133] "voltage": 110.6,

[0134] "active_power": 100.1,

[0135] "reactive_power": 50.1

[0136] },

[0137] "temperature": {

[0138] "oil_temperature": 60.1,

[0139] "winding_temperature": 70.1,

[0140] "switch_temperature": 55.1

[0141] },

[0142] "electrical_characteristics": {

[0143] "core_leakage_current": 0.6,

[0144] "coil_resistance": 1.3

[0145] },

[0146] "chemical_composition": {

[0147] "gas_content": {

[0148] "H2": 11,

[0149] "C2H2": 2.5,

[0150] "CH4": 31,

[0151] "C2H4": 5.5

[0152] }

[0153] },

[0154] "mechanical_characteristics": {

[0155] "liquid_pressure": 1000001,

[0156] "gas_density": 1.26

[0157] },

[0158] "other_information": {

[0159] "environmental_temperature": 25.1,

[0160] "humidity": 61

[0161] }

[0162] }

[0163] / / More data items... ]

[0165] }

[0166] Timestamp: timestamp, in ISO 8601 standard format.

[0167] device_id: device identifier, used to distinguish different devices.

[0168] electrical_parameters: electrical parameters, including current, voltage, active power and reactive power.

[0169] temperature: Temperature information, including oil level temperature, winding temperature and disconnector temperature.

[0170] electrical_characteristics: Electrical characteristics, including core leakage current and coil resistance.

[0171] chemical_composition: chemical composition, including gas content of gas relay and characteristic gas composition of oil chromatography.

[0172] mechanical_characteristics: Mechanical properties, including real-time pressure of liquids and density of gases.

[0173] other_information: Other relevant information, such as ambient temperature and humidity.

[0174] Step 4: Feature extraction: extract time series features, such as mean, variance, maximum and minimum values, etc., and combine features, such as the correlation between different features;

[0175] Step 5: Feature selection, using correlation analysis, principal component analysis and other methods to select the key features that best reflect the changes in equipment status;

[0176] Step 6: Model selection: Select a suitable machine learning model, such as random forest, support vector machine, neural network, etc.

[0177] Step 7: Model training: Use historical data sets to train the model, use k-fold cross validation to evaluate model performance, and tune hyperparameters through grid search or random search;

[0178] Step 8: Model verification: verify the model performance on an independent test set to ensure that the model has good generalization ability;

[0179] Step 9: Real-time monitoring: input the newly collected data into the trained model for real-time prediction. If the model predicts a potential fault, an early warning is issued and the fault is judged by setting the threshold of the characteristic value.

[0180] Step 10: Continuous optimization, update the model regularly, retrain and evaluate the model using new data collection, user feedback, and continuously improve model performance.

[0181] In this embodiment, after the model system is established, its overall interface is as follows: Figure 6 As shown in Figure 2, the monitoring parameters of the transformer are as follows: Figure 7 As shown in Figure 2, the monitoring parameters of the circuit breaker are as follows: Figure 8 As shown in Figure 2, the monitoring parameters of the isolating switch are as follows: Fig. 9 As shown, from the display diagram after the above model is established, the actual conditions of various parameters can be intuitively observed, the operation status of the substation can be understood in real time, and alarm information can be sent when abnormal conditions occur, and timely responses can be made to avoid emergency shutdowns caused by sudden failures, thereby improving the safety and stability of substation equipment operation.

[0182] Specifically, Figure 5 As shown, the gas content change data acquisition device of the gas relay adopts an ESP32-CAM module, which integrates a camera and Wi-Fi functions, and is used to take pictures of the observation window of the gas relay, and transmit the taken pictures to the edge intelligent control processing device or the backend server via Wi-Fi.

[0183] like Fig.11As shown, the software processing process of the gas content change data acquisition device of the gas relay is: image acquisition→select ROI area→image enhancement and noise reduction→marginalization monitoring→image graying→extracting gas relay liquid level.

[0184] In this embodiment, in the image acquisition step, first, a high-resolution camera installed near the observation window of the gas collecting box of the gas relay is used to periodically or continuously capture the dial image. The position of the camera is optimized and an LED fill light is installed to ensure the best shooting angle and clarity.

[0185] In the ROI selection step, the Rect function is used to extract the region of interest (ROI) containing the circular part of the Buchholz relay dial from the acquired image, reducing the amount of data to be processed and focusing on key information.

[0186] In the steps of image enhancement and noise reduction, edge monitoring, and image grayscale, the selected ROI area is subjected to image enhancement and noise reduction, including histogram equalization, contrast adjustment, brightness adjustment, sharpening, and grayscale, to highlight the details of the dial and remove unnecessary noise. The RGB color image is converted to the HSV color space to better extract the liquid level information. The conversion formula is as follows:

[0187]

[0188] In the edge monitoring and image graying steps, the Canny edge detection algorithm is combined with statistical laws to determine the liquid level and other important indicators. The Canny edge detection algorithm is based on multi-level edge detection and first grays the image:

[0189]

[0190] Then the number of edge pixels is calculated to represent the complexity of the regional structure, and threshold processing is used to determine whether the region is an oil-containing area.

[0191] To enhance the image through histogram equalization, first calculate the cumulative histogram of the image and calculate the probability of each gray level appearing in the image:

[0192]

[0193] Among them, N is the total number of pixels, pi is the probability of occurrence of the i-th grayscale, and ni is the total number of grayscale pixels at the i-th level. The cumulative statistical histogram of each level of grayscale is then calculated using the following formula:

[0194]

[0195] Based on the cumulative histogram, the original image is stretched and mapped in the grayscale space to achieve image histogram equalization.

[0196] In the step of extracting the liquid level of the gas relay, the ratio of the liquid level height to the total height can be used to estimate the volume proportion of the liquid. Assuming the liquid level height is h and the total height is H, the liquid proportion for:

[0197]

[0198] Specifically, the transformer core leakage current data acquisition device, The software processing process of the equipment gas density monitoring sensor device, the circuit breaker hydraulic energy storage mechanism pressure online monitoring sensor device, the isolation switch temperature monitoring sensor device, and the circuit breaker opening and closing coil resistance acquisition sensor device are all:

[0199] Data acquisition (initialization, timing acquisition) → data processing (filtering, calibration) → data transmission (data packaging, data sending).

[0200] Specifically, the hardware of the transformer core leakage current data acquisition device adopts a Hall effect sensor. The gas density monitoring sensor device of the equipment adopts a gas density sensor, the pressure online monitoring sensor device of the hydraulic energy storage mechanism of the circuit breaker adopts a pressure sensor, the temperature monitoring sensor device of the isolating switch adopts an infrared temperature sensor, and the circuit breaker opening and closing coil resistance collection sensor device adopts a coil resistance sensor.

[0201] The above acquisition devices also include a signal conditioning circuit, an analog-to-digital converter, a microcontroller, a wireless transmission module, and a power supply module. Figure 4 As shown, the data collected by each sensor in the acquisition device are respectively connected to the corresponding signal conditioning circuit, the signal conditioning circuit amplifies and conditions the weak signal output by the sensor to make it suitable for processing, the analog-to-digital converter converts the conditioned analog signal into a digital signal, the microcontroller processes the digital signal, performs data processing and transmission, the wireless transmission module transmits the data to the edge intelligent control processing device or the backend server via Wi-Fi, LoRa or Bluetooth wireless communication, and the power module provides a stable battery power supply to ensure long-term stable operation of the equipment.

[0202] like Fig.10As shown in the figure, the system has produced corresponding physical objects and applied them in actual sites. In general, the system realizes regular or continuous data collection through data acquisition devices, improves the monitoring frequency, replaces manual inspections with automated monitoring methods, improves work efficiency, and reduces labor intensity. The collected data is deeply analyzed through edge intelligent control processing devices and status analysis models, which provides a scientific basis for operation and maintenance. Maintenance plans can be formulated in advance, early warning issues can be responded to in a timely manner, emergency shutdowns caused by sudden failures can be reduced, and the safety and stability of equipment operation can be improved.

[0203] The preferred embodiments of the present invention are described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge scope of ordinary technicians in this field without departing from the purpose of the present invention.

[0204] Many other changes and modifications may be made without departing from the concept and scope of the present invention.It should be understood that the present invention is not limited to the specific embodiments, and the scope of the present invention is defined by the appended claims.

Claims

1. The substation equipment data collection and status analysis model system based on edge intelligence is characterized by: It includes a data acquisition device, an edge intelligent exchange control processing device, and a status analysis model. The data acquisition device transmits data to the edge intelligent exchange control processing device through a communication networking protocol. The edge intelligent exchange control processing device summarizes and processes the data collected by the data acquisition device through an embedded computer or an industrial controller to achieve data conversion, encryption and transmission. The status analysis model establishes a fault diagnosis model based on the data processed by the edge intelligent exchange control processing device to achieve fault warning and fault analysis for related equipment. The data acquisition device includes a gas content change data acquisition device for a gas relay, a transformer core leakage current data acquisition device, Equipment gas density monitoring sensor device, circuit breaker hydraulic energy storage mechanism pressure online monitoring sensor device, disconnector temperature monitoring sensor device, circuit breaker opening and closing coil resistance collection sensor device, The edge intelligent control processing device includes a gas relay receiving gateway, an 8-channel RS485 gateway, a 4-channel RS485 gateway, an RS485 to TCP transparent transmission device, a networking router, an embedded computer, and a power module. The gas relay receiving gateway receives the image information collected by the gas content change data acquisition device of the gas relay. The 8-channel RS485 gateway receives data from the pressure online monitoring sensor device of the circuit breaker hydraulic energy storage mechanism and the circuit breaker opening and closing coil resistance acquisition sensor device. The 4-channel RS485 gateway receives data from the isolation switch temperature monitoring sensor device, the transformer core leakage current data acquisition device, The data of the equipment gas density monitoring sensor device, the gas relay receiving gateway, the 8-channel RS485 gateway transmits the data to the networking router, the 4-channel RS485 transmits the data to the networking router through the RS485 to TCP transparent transmission device, the networking router transmits the data to the embedded computer, and the power module provides power for the edge intelligent control processing device.

2. According to claim 1, the substation equipment data acquisition and status analysis model system based on edge intelligence is characterized in that: The software processing flow in the embedded computer of the edge intelligent exchange control processing device includes the following steps: S01: Data collection, collecting various sensor data through RS485 gateway and RS485 to TCP transparent transmission device; S02: Data processing, summarizing, converting, encrypting and transmitting the collected data; S03: Data cleaning, screening out data that is not within the normal range or empty data or redundant data; S04: Data transmission, transmitting the processed data to the backend server through the network; S05: Visualization interface, which displays and analyzes data through the visualization interface on the background server.

3. According to claim 1, the substation equipment data acquisition and status analysis model system based on edge intelligence is characterized in that: The gas relay receiving gateway adopts a Raspberry Pi 5A processor, connects to the ESP32 camera, uses the OpenCV library to receive and analyze the images from the ESP32 camera, and transmits the analysis results to the embedded computer through the network. The embedded computer creates a Socket connection thereon, connects to the RS485 to TCP transparent transmission device, connects the RS485 to TCP transparent transmission device and the gas relay receiving gateway to the same network through a networking router, and configures an IP address and a port number.

4. The substation equipment data acquisition and status analysis model system based on edge intelligence according to claim 2 is characterized in that: In the step S02, the data processing uses the struct module to convert the received binary data into floating point numbers; The specific steps of data cleaning in step S03 are: first, define the normal range of each sensor data according to the specifications and actual needs of the equipment, screen out the data that is not within the normal range or empty data and redundant data, then traverse the converted data to check whether each data is within the normal range, and do not save the data that is not within the normal range; In step S04, data transmission is performed by hybrid encryption and then transmitted to the backend server via the TCP protocol.

5. The substation equipment data collection and status analysis model system based on edge intelligence according to claim 1 is characterized in that: The steps for establishing the state analysis model are as follows: Step 1: Determine the data source and collection indicators, including transformer electrical parameters, oil surface temperature, winding temperature, core leakage current, gas content of gas relay, characteristic gas composition of oil chromatogram, real-time liquid pressure, gas density, coil resistance value of circuit breaker, and temperature of disconnector; Step 2: Classify all data sources into: Electrical parameters: voltage, current, active power, reactive power, Temperature: oil level temperature, winding temperature, disconnect switch temperature, Electrical characteristics: core leakage current, coil resistance, Chemical composition: gas content of gas relay, characteristic gas composition of oil chromatography, Mechanical properties: real-time liquid pressure, gas density; Step 3: Data preprocessing, including data cleaning, data standardization, and data fusion; Step 4: Feature extraction, extracting time series features and combination features; Step 5: Feature selection, using correlation analysis and principal component analysis methods to select the key features that best reflect changes in equipment status; Step 6: Model selection: Select a suitable machine learning model; Step 7: Model training: Use historical data sets to train the model, use k-fold cross validation to evaluate model performance, and tune hyperparameters through grid search or random search; Step 8: Model validation; Step 9: Real-time monitoring, input the newly collected data into the trained model for real-time prediction; Step 10: Continuously optimize, update the model regularly, and retrain and evaluate the model using new data.

6. The substation equipment data acquisition and status analysis model system based on edge intelligence according to claim 1 is characterized in that: The gas content change data acquisition device of the gas relay adopts the ESP32-CAM module, which integrates the camera and Wi-Fi functions, and is used to take pictures of the observation window of the gas relay, and transmit the taken pictures to the edge intelligent control processing device or the backend server via Wi-Fi. The software processing process of the gas content change data acquisition device of the gas relay is: image acquisition→selection of ROI area→image enhancement and noise reduction→marginalization monitoring→image graying→extraction of gas relay liquid level.

7. The substation equipment data acquisition and status analysis model system based on edge intelligence according to claim 1 is characterized in that: The transformer core leakage current data acquisition device, The software processing process of the equipment gas density monitoring sensor device, the circuit breaker hydraulic energy storage mechanism pressure online monitoring sensor device, the isolation switch temperature monitoring sensor device, and the circuit breaker opening and closing coil resistance acquisition sensor device are all: Data acquisition (initialization, timing acquisition) → data processing (filtering, calibration) → data transmission (data packaging, data sending).

8. The substation equipment data acquisition and status analysis model system based on edge intelligence according to claim 1 is characterized in that: The hardware of the transformer core leakage current data acquisition device adopts a Hall effect sensor. The gas density monitoring sensor device of the equipment adopts a gas density sensor, the pressure online monitoring sensor device of the hydraulic energy storage mechanism of the circuit breaker adopts a pressure sensor, the temperature monitoring sensor device of the isolating switch adopts an infrared temperature sensor, and the circuit breaker opening and closing coil resistance collection sensor device adopts a coil resistance sensor. The above-mentioned acquisition devices also include a signal conditioning circuit, an analog-to-digital converter, a microcontroller, a wireless transmission module, and a power module. The data collected by each sensor in the acquisition device are respectively connected to the corresponding signal conditioning circuit. The signal conditioning circuit amplifies and conditions the weak signal output by the sensor to make it suitable for processing. The analog-to-digital converter converts the conditioned analog signal into a digital signal. The microcontroller processes the digital signal and performs data processing and transmission. The wireless transmission module transmits data to the edge intelligent control processing device or the backend server via Wi-Fi, LoRa or Bluetooth wireless communication. The power module provides a stable battery power supply to ensure long-term stable operation of the equipment.

Images