Power distribution overhead line insulation defect monitoring system and method based on multi-node data fusion

Abstract

The application discloses a multi-node data fusion power distribution overhead line insulation defect monitoring system and method, and relates to the technical field of power systems, comprising: a Rogowski coil 1, an electromagnetic induction coil 2, a voltage coupler 3, a low-power ARM processor 4, a signal processing circuit 5, an energy collection circuit 6, a low-power NB-IOT module 7 and a cloud platform 8. The application has the advantages that 1) a peak value holding circuit is used to perform frequency reduction processing on the partial discharge high-frequency signal generated by the insulation defect, thereby avoiding the need for an expensive and power-consuming high-speed data acquisition system; and 2) energy is obtained from the line through the electromagnetic induction coil and the energy collection circuit, and low-line current starting and low-power data acquisition processing are realized based on a low-power working strategy and low-power device selection, so that the monitoring system can be installed at any position of the overhead power distribution line and is not limited by external power supply and the size of the current in the line.

Description

Multi-node data fusion-based monitoring system and method for insulation defects in overhead power distribution lines Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a monitoring system and method for insulation defects in overhead distribution lines based on multi-node data fusion. Background Technology

[0002] Overhead distribution lines are the lifeblood of the power system, numerous and complex, connecting power plants to end users. Because they are widely used in rural areas, forests, and some urban areas, and are exposed to the elements year-round, they are prone to insulation faults that are difficult to locate. If faults in overhead distribution lines are not addressed promptly, they can escalate into serious power outages, especially in forested areas where they can trigger wildfires and cause significant economic losses. Therefore, monitoring and diagnosing insulation defects in overhead distribution lines is of great practical importance.

[0003] Currently, domestic power grid companies mainly use ultrasonic and infrared detection instruments to inspect insulation defects in overhead distribution lines, achieving good operation and maintenance results. However, these detection methods are not sensitive to internal insulation defects, easily leading to missed detections. Moreover, for widely distributed overhead distribution lines, manual inspection along the lines is very time-consuming, labor-intensive, and unsafe. In recent years, scholars have proposed a high-frequency partial discharge detection technology based on traveling waves. This technology can detect insulation defects over a large area of ​​overhead distribution lines, thereby improving the efficiency and effectiveness of overhead distribution line operation and maintenance. However, in practical applications, this technology still faces challenges when monitoring insulation defects on multiple overhead distribution lines, mainly due to the following two reasons. First, embedded solutions for traveling wave detection require expensive and power-intensive high-speed data acquisition equipment (e.g., equipment with a sampling rate of at least 20 Msps). Second, these detection devices have specific power supply requirements and may need to be installed near line towers, which limits the installation and deployment location of the monitoring system. Therefore, further research into insulation defect monitoring technology for overhead power distribution lines to achieve efficient, flexible, and low-cost insulation defect detection and diagnosis is an important topic in the research, product design, and application of this type of technology.

[0004] Therefore, proposing a multi-node data fusion-based insulation defect monitoring system and method for overhead power distribution lines to address the difficulties in existing technologies is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides a multi-node data fusion system and method for monitoring insulation defects in overhead power distribution lines, which can realize low-cost, flexible and efficient online monitoring and diagnosis of insulation status of overhead power distribution lines.

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

[0007] A multi-node data fusion-based overhead power distribution line insulation defect monitoring system includes a Rogowski coil, an electromagnetic induction coil, a voltage coupler, a low-power ARM processor, a signal processing circuit, an energy harvesting circuit, a low-power NB-IoT communication module, and a cloud platform.

[0008] The Rogowski coil is electrically connected to the signal processing circuit and is used to detect partial discharge signals generated by insulation defects on the line through the principle of high-frequency current induction.

[0009] The electromagnetic induction coil is electrically connected to the energy harvesting circuit and is used to obtain the power frequency voltage from the power frequency current in the circuit through the principle of magnetic field induction energy harvesting.

[0010] The energy harvesting circuit is electrically connected to both the low-power ARM processor and the signal processing circuit, and is used to provide power to both the low-power ARM processor and the signal processing circuit.

[0011] The voltage coupler is electrically connected to the low-power ARM processor; it is used to detect the phase information of the line power frequency voltage based on the principle of voltage mutual inductance.

[0012] The low-power NB-IoT communication module is electrically connected to both the low-power ARM processor and the cloud platform; it is used to monitor communication and data transmission between the system and the cloud platform.

[0013] The cloud platform is used to monitor the system's operational status at different nodes and to perform fusion analysis and diagnosis of monitoring data from multiple monitoring nodes.

[0014] The low-power ARM processor is used to implement the core control and algorithm functions of the monitoring system and coordinate the stable operation of various modules.

[0015] The signal processing circuit is used to preprocess the partial discharge signal detected by the Rogowski coil.

[0016] Optionally, in the above system, the signal processing circuit consists of a filter amplifier circuit and a peak hold circuit, used to improve the signal-to-noise ratio and reduce the signal frequency to avoid the use of high-speed data acquisition devices.

[0017] Optionally, the energy harvesting circuit in the above system consists of an overvoltage protection circuit, a rectifier circuit, an over-energy release circuit, and an energy management circuit, used for overvoltage protection, rectification and voltage regulation, and energy management of the power frequency voltage collected by the electromagnetic induction coil.

[0018] Optionally, in the above system, the voltage coupler is made of a semi-annular copper plate to provide a phase reference for the arrival time of partial discharge and the PRPD spectrum.

[0019] Optionally, in the above system, the low-power ARM processor is an STM32L476RG low-power processor.

[0020] Optionally, in the above system, the low-power NB-IoT communication module uses the BC26 communication module.

[0021] Optionally, the cloud platform in the above system can perform fusion analysis and diagnosis of monitoring data at different monitoring nodes in two ways: single-node edge monitoring diagnosis and multi-node data fusion diagnosis.

[0022] In single-node edge monitoring and diagnosis, a single monitoring node uploads amplitude and partial discharge phase spectrum information to the cloud platform. The cloud platform diagnoses the severity and type of defects based on the amplitude and partial discharge phase spectrum detected by the node.

[0023] The cloud platform integrates and analyzes the detection amplitude information uploaded by multiple monitoring nodes, calculates the amplitude offset Δm of each node using the following formula, and compares the Δm at different monitoring nodes to determine the defect location and severity.

[0024]

[0025]

[0026] Where: U is the amplitude at a certain monitoring node, U av is the mean of the amplitudes of all monitoring nodes, and k is the number of monitoring nodes.

[0027] A method for monitoring insulation defects in overhead distribution lines using multi-node data fusion, employing any of the above-mentioned multi-node data fusion systems for monitoring insulation defects in overhead distribution lines, includes the following steps:

[0028] Step 1: After installing the monitoring device on the overhead power line, use the electrical energy obtained from the magnetic induction coil to start the energy harvesting circuit and the low-power ARM processor;

[0029] Step 2: Start the low-power NB-IoT communication module and determine if the system communication is normal. If the communication is abnormal, reset the low-power NB-IoT communication module; otherwise, proceed to Step 3.

[0030] Step 3: Check the charging status of the supercapacitor at regular intervals and wait for the supercapacitor to complete charging to enable normal system startup under low line current. If the supercapacitor voltage is less than 3.3V, continue to wait; otherwise, proceed to Step 4.

[0031] Step 4: Start the ADC and signal processing module to detect partial discharge signals generated by insulation defects on the overhead power distribution line, collect partial discharge and power frequency voltage data, calculate the amplitude and phase information of the partial discharge, and upload the detection results through the low-power NB-IoT communication module.

[0032] As can be seen from the above technical solution, compared with the prior art, the present invention provides a multi-node data fusion system and method for monitoring insulation defects in overhead power distribution lines:

[0033] 1) This invention uses a peak hold circuit to down-frequency process the high-frequency partial discharge signal generated by insulation defects, thereby avoiding the need for expensive and power-consuming high-speed data acquisition systems.

[0034] 2) This invention obtains energy from the line through electromagnetic induction coils and energy harvesting circuits, and realizes low line current start-up and low power data acquisition and processing based on low power operation strategy and low power device selection, so that the monitoring system can be installed at any location on the overhead power distribution line, without being limited by the external power supply and the current in the line.

[0035] 3) This invention uses multi-node data fusion diagnosis and horizontal comparative analysis of monitoring information from different monitoring nodes to obtain a more comprehensive insulation defect status, which can provide new reference for the location of insulation defects and the determination of the severity of defects. Attached Figure Description

[0036] 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 embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0037] Figure 1 is an overall framework diagram of the multi-node data fusion-based overhead power line insulation defect monitoring system provided by the present invention.

[0038] Figure 2 is a schematic diagram of the principle of insulation defect detection provided in an embodiment of the present invention;

[0039] Figure 3 is a schematic diagram of the energy harvesting principle provided by an embodiment of the present invention;

[0040] Figure 4 is a flowchart of the multi-node data fusion method for monitoring insulation defects in overhead power distribution lines provided by the present invention;

[0041] Figure 5 is a schematic diagram illustrating the principle of multi-node data fusion diagnosis provided by the present invention;

[0042] Figure 6 is a schematic diagram of the application of the multi-node data fusion overhead distribution line insulation defect monitoring system provided in the overhead distribution line system of the present invention.

[0043] Among them: 1-Rogowski coil, 2-Electromagnetic induction coil, 3-Voltage coupler, 4-Low-power ARM processor, 5-Signal processing circuit, 6-Energy harvesting circuit, 7-Low-power NB-IoT communication module, 8-Cloud platform, 51-Filtering and amplification circuit, 52-Peak hold circuit, 61-Overvoltage protection circuit, 62-Rectifier circuit, 63-Over-energy release circuit, 64-Energy management circuit. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] Referring to Figure 1, the present invention discloses a multi-node data fusion overhead power distribution line insulation defect monitoring system, including a Rogowski coil 1, an electromagnetic induction coil 2, a voltage coupler 3, a low-power ARM processor 4, a signal processing circuit 5, an energy harvesting circuit 6, a low-power NB-IoT communication module 7, and a cloud platform 8.

[0046] Rogowski coil 1 is electrically connected to signal processing circuit 5 and is used to detect partial discharge signals generated by insulation defects on the line through the principle of high-frequency current induction.

[0047] Electromagnetic induction coil 2 is electrically connected to energy harvesting circuit 6 and is used to obtain power frequency voltage from the power frequency current in the circuit through the principle of magnetic field induction energy harvesting.

[0048] The energy harvesting circuit 6 is electrically connected to the low-power ARM processor 4 and the signal processing circuit 5, respectively, and is used to provide power to the low-power ARM processor 4 and the signal processing circuit 5.

[0049] Voltage coupler 3 is electrically connected to low-power ARM processor 4; it is used to detect the phase information of the line power frequency voltage based on the principle of voltage mutual inductance.

[0050] The low-power NB-IoT communication module 7 is electrically connected to the low-power ARM processor 4 and the cloud platform 8 respectively; it is used to monitor the communication and data transmission between the system and the cloud platform 8.

[0051] Cloud Platform 8 is used to monitor the system's operational status at different nodes and to perform fusion analysis and diagnosis of monitoring data from multiple monitoring nodes;

[0052] The low-power ARM processor 4 is used to implement the core control and algorithm functions of the monitoring system and coordinate the stable operation of various modules.

[0053] Signal processing circuit 5 is used to preprocess the partial discharge signal detected by Rogowski coil 1.

[0054] Furthermore, the signal processing circuit 5 consists of a filter amplifier circuit 51 and a peak hold circuit 52, which are used to improve the signal-to-noise ratio and reduce the signal frequency to avoid the use of high-speed data acquisition devices.

[0055] Specifically, as shown in Figure 2, the insulation defect detection includes a Rogowski coil 1 and a signal processing circuit 5; the Rogowski coil 1 is used to detect partial discharge signals generated by insulation defects on the line; the signal processing circuit 5 consists of a filter amplifier circuit 51 and a peak hold circuit 52, which are used to preprocess the partial discharge signals detected by the Rogowski coil 1.

[0056] Furthermore, the energy harvesting circuit 6 consists of an overvoltage protection circuit 61, a rectifier circuit 62, an over-energy release circuit 63, and an energy management circuit 64, which are used to perform overvoltage protection, rectification and voltage regulation, and energy management on the power frequency voltage collected by the electromagnetic induction coil 2.

[0057] Specifically, as shown in Figure 3, the energy acquisition includes a magnetic field induction coil 2 and an energy harvesting circuit 6; the magnetic field induction coil 2 collects the power frequency voltage in the power frequency current of the circuit through magnetic field induction; the energy harvesting circuit 6 consists of an overvoltage protection circuit 61, a rectifier circuit 62, an over-energy release circuit 63 and an energy management circuit 64, which are used to perform overvoltage protection, rectification and voltage regulation and energy management on the power frequency voltage collected by the electromagnetic induction coil 2.

[0058] Furthermore, voltage coupler 3 is made of a semi-annular copper plate to provide a phase reference for the arrival time of partial discharge and the PRPD spectrum.

[0059] Furthermore, the low-power ARM processor 4 uses the STM32L476RG low-power processor.

[0060] Furthermore, the low-power NB-IoT communication module 7 uses the BC26 communication module.

[0061] Furthermore, the cloud platform 8 performs integrated analysis and diagnosis of monitoring data at different monitoring nodes, including two methods: single-node edge monitoring diagnosis and multi-node data fusion diagnosis.

[0062] In single-node edge monitoring and diagnosis, a single monitoring node uploads amplitude and partial discharge phase spectrum information to cloud platform 8. Cloud platform 8 diagnoses the severity and type of defects based on the amplitude and partial discharge phase spectrum detected by the node.

[0063] The cloud platform 8 integrates and analyzes the detection amplitude information uploaded by multiple monitoring nodes, calculates the amplitude offset Δm of each node using the following formula, and compares the Δm at different monitoring nodes to determine the defect location and severity.

[0064]

[0065]

[0066] Where: U is the amplitude at a certain monitoring node, U av is the mean of the amplitudes of all monitoring nodes, and k is the number of monitoring nodes.

[0067] In one specific embodiment, the monitoring system measures 10cm*10cm*15cm and weighs 800g, allowing for convenient installation and implementation on the line. The Rogowski coil 1, electromagnetic induction coil 2, and voltage coupler 3 all employ an open design, allowing for non-intrusive installation at any location on the overhead power line without affecting the circuit's operation. The Rogowski coil 1 detects partial discharge signals generated by insulation defects on the line using the principle of high-frequency current induction. The electromagnetic induction coil 2 induces power frequency voltage from the power frequency current in the line using the principle of magnetic field induction. The voltage coupler 3 detects the phase information of the line's power frequency voltage based on the principle of voltage mutual inductance, providing a phase reference for the arrival time of partial discharge and the PRPD spectrum. The low-power ARM processor 4 uses an STM32L476RG low-power processor with an operating voltage of 3.3V. It supports a low-power advanced reduced instruction set and has an ADC peripheral with a sampling rate of 5Msps. Under normal operation, its power consumption is approximately 10mW, and when the ADC is enabled, it consumes approximately 80mW. It implements the core control and algorithm functions of the monitoring system, coordinating the stable operation of various modules. The signal processing circuit 5 preprocesses the partial discharge signal detected by the Rogowski coil 1, mainly including a filter amplification circuit 51 and a peak hold circuit 52, to improve the signal-to-noise ratio and reduce the signal frequency to avoid the use of a high-speed data acquisition system. The energy harvesting circuit 6 provides overvoltage protection, rectification, voltage regulation, and energy management for the power frequency voltage collected by the electromagnetic induction coil 2. It mainly consists of an overvoltage protection circuit 61, a rectification circuit 62, an over-energy release circuit 63, and an energy management circuit 64 to meet the power supply requirements of the low-power ARM processor 4 and the signal processing circuit 5. The low-power NB-IoT communication module 7 uses the low-power BC26 communication module manufactured by Quectel, supporting 4G narrowband IoT communication with an operating power consumption of approximately 5mW. It is used for communication and data transmission between the monitoring system and the cloud platform 8. The low-power NB-IoT communication module 7 and the cloud platform 8 communicate based on narrowband IoT 4G. The low-power NB-IoT communication module 7 can upload detection data from the monitoring nodes, such as amplitude and PRPD spectrum information, to the cloud platform 8. The cloud platform 8 can monitor the system's operating status at different nodes and perform fusion analysis and diagnosis of monitoring data from multiple monitoring nodes.

[0068] As shown in Figure 2, the Rogowski coil 1 adopts an open-type design, consisting of two semi-rings with inner diameter d1 and outer diameter d2 of 40mm and 70mm respectively, a thickness h1 of 15mm, 20 turns, and a passband bandwidth frequency of 0.2-50MHz. It can be easily installed on overhead power lines for non-invasive detection. Since the partial discharge signal of insulation defects detected by the Rogowski coil 1 is generally in the mV range and contains harmonics and high-frequency interference, the original detection signal needs to be filtered and amplified to improve the signal-to-noise ratio. The signal processing circuit 5 mainly consists of a filter amplifier circuit 51 and a peak hold circuit 52. The filter amplifier circuit 51 has a passband frequency of 1-20MHz and a gain of 20dB, which can improve the signal-to-noise ratio to effectively extract the partial discharge information generated by insulation defects. The peak hold circuit 52 is used to down-convert the high-frequency partial discharge signal to meet the 5Msps sampling rate requirement of the low-power STM32L476RG ARM processor ADC peripheral, avoiding the use of a high-speed data acquisition system and reducing system cost. Furthermore, a simple thresholding method is used to extract information about the amplitude and arrival time of the partial discharge pulse. The threshold is defined as m + 3σ, where m and σ are the average value and standard deviation of the sampled signal, respectively. The calculated arrival time is then converted into a phase value to obtain accurate phase information of the partial discharge pulse within the power frequency voltage signal period.

[0069] As shown in Figure 3, the electromagnetic induction coil 2 adopts an open design, consisting of two semi-rings, and is made of permalloy with high magnetic permeability. Its inner diameter d3 and outer diameter d4 are 40mm and 80mm respectively, and its thickness h2 is 20mm. The coil has 750 turns, allowing for easy installation on overhead power lines to initially harvest energy. The energy harvesting circuit 6 mainly consists of an overvoltage protection circuit 61, a rectifier circuit 62, an over-energy release circuit 63, and an energy management circuit 64, enabling the monitoring system to safely and stably harvest energy from the line. The overvoltage protection circuit is implemented by a surge arrester RV1, a gas discharge tube GTD, an inductor L, and a Zener diode S1, protecting the circuit board from transient voltages in the power system. The rectifier circuit consists of four diodes D1, D2, D3, and D4, and a capacitor C1, converting the power frequency voltage harvested by the electromagnetic induction coil 2 into DC voltage. The over-energy release circuit consists of a Zener diode S2, resistors R1 and R2, a metal-oxide-semiconductor field-effect transistor M (SQD50N10-8m9L, Vishay Inter Technology), and a valve resistor r. When the load current in the overhead power distribution line is in the range of several hundred A, it can limit the voltage to within 10V to protect downstream circuits. When the voltage exceeds 10V, the field-effect transistor M turns on, and the excess energy is consumed by resistor r. The energy management circuit consists of an energy management chip (LTC3355, Analog Devices Inc.), a 120F supercapacitor Csuper, and peripheral electronic components. During startup, the energy management chip first charges the supercapacitor to store energy. During operation, when the energy collected by the preceding stage circuit is low, the supercapacitor supplies power to the subsequent load circuit; when the energy collected by the preceding stage circuit is high, it directly supplies power to the subsequent load circuit and stores the excess energy in the supercapacitor. Using this design, the maximum voltage drop of the monitoring system in each measurement cycle is only 1mV.

[0070] Table 1 Power Consumption of Each Module in the Monitoring System

[0071]

[0072] As shown in Table 1, the energy harvesting circuit, the NB-IoT module, and the low-power ARM processor with the analog-to-digital converter (ADC) disabled consume very little power, while the low-power ARM processor with the ADC enabled and the signal processing circuit consume significantly more power. Therefore, to ensure the monitoring system operates normally under low line current, a low-power management algorithm, as shown in Figure 4, was developed and programmed into the low-power ARM processor 4. The method is described below:

[0073] 1) After the monitoring system is installed on the overhead power line, the energy harvesting circuit 6 and the low-power ARM processor 4 are started using the electrical energy obtained by the electromagnetic induction coil 2.

[0074] 2) Start the low-power NB-IoT communication module 7 and determine if the system communication is normal. If the communication is abnormal, reset the low-power NB-IoT communication module 7; otherwise, proceed to the next step.

[0075] 3) Check the charging status of the supercapacitor every 5 minutes and wait for it to complete charging to enable normal system startup under low line current. If the supercapacitor voltage is less than 3.3V, continue to wait; otherwise, proceed to the next step.

[0076] 4) Start the ADC and signal processing module to detect the partial discharge signal generated by insulation defects on the overhead power distribution line, collect partial discharge and power frequency voltage data, calculate the amplitude and phase information of the partial discharge, and upload the detection results through the low-power NB-IOT communication module 7.

[0077] By adopting the above-mentioned low-power management algorithm, the monitoring system can start normally with a line current as low as 5A, and can collect, process and transmit 50 power frequency voltage cycles (20ms), or 1s, of detection data at a time.

[0078] As shown in Figure 5, due to the attenuation characteristics of the partial discharge signal generated by the insulation defects on the overhead power distribution line during the propagation process, the signal strength will continuously weaken as the propagation distance increases. The cloud platform 8 will perform fusion analysis on the detection amplitude information uploaded by multiple monitoring nodes, calculate the amplitude offset Δm of each node using the following formula, and compare the Δm at different monitoring nodes laterally to determine the defect location and the severity of the defect.

[0079]

[0080]

[0081] Where: U is the amplitude at a certain monitoring node, U av is the mean of the amplitudes of all monitoring nodes, and k is the number of monitoring nodes.

[0082] By comparing Δm at different monitoring nodes laterally, a reference can be provided for defect location and severity determination. In Figure 5, Δmr is a reference value for lateral comparison set based on experimental and empirical values. According to the relationship between Δm and Δmr at each monitoring node, the status of the monitoring node can be defined as normal, warning, and dangerous. A normal state is defined as Δm < Δmr, a warning state is defined as Δmr ≤ Δm < 2Δmr, and a dangerous state is defined as Δm ≥ Δmr. Figure 5 shows n monitoring nodes. According to the above definitions, nodes 3 and n are in normal state, node 1 is in warning state, and node 2 is in dangerous state. Since the abnormal signal generated by partial discharge from insulation defects attenuates continuously with increasing propagation distance on the line, it is not difficult to determine that there may be an insulation defect between node 1 and node 2, and the insulation defect is relatively close to node 1. Through the above multi-node fusion diagnostic analysis, the operating status of multiple monitoring nodes can be analyzed in parallel from a lateral perspective, and the diagnosis and location of line insulation defects can be achieved to a certain extent. In addition, each monitoring node will upload the detection amplitude and PRPD spectrum information at that node. After the location of the defect is initially determined through multi-node fusion diagnostic analysis, the detection information at the individual monitoring node can be used to diagnose and analyze the severity and type of the defect, and to more comprehensively assess the operating status of the overhead power distribution line system.

[0083] The monitoring system of this invention can be installed at any location on the distribution network line, requires no external power supply, and can operate normally with line currents as low as 5A. As shown in Figure 6, the monitoring system can be installed at the line tower switch and at any location on the line conductor. Furthermore, the monitoring system of this invention performs frequency reduction processing on high-frequency partial discharge signals, avoiding the use of expensive high-speed data acquisition systems, saving costs, and is suitable for large-scale installation and deployment in overhead distribution lines, demonstrating strong feasibility. Monitoring systems at different nodes upload monitoring data (amplitude, PRPD spectrum) and other information from that node to the cloud platform 8. The cloud platform 8 performs fusion processing and analysis on the monitoring data from multiple nodes, enabling the location and diagnosis of insulation defects in the distribution line.

[0084] Referring to Figure 4, the present invention also provides a method for monitoring insulation defects in overhead distribution lines using multi-node data fusion, for the specific implementation of the system in Figure 1. The method for monitoring insulation defects in overhead distribution lines using multi-node data fusion provided in this embodiment can be applied to computer terminals or various mobile devices, specifically including:

[0085] Step 1: After installing the monitoring device on the overhead power line, use the electrical energy obtained from the magnetic induction coil to start the energy harvesting circuit and the low-power ARM processor;

[0086] Step 2: Start the low-power NB-IoT communication module and determine if the system communication is normal. If the communication is abnormal, reset the low-power NB-IoT communication module; otherwise, proceed to Step 3.

[0087] Step 3: Check the charging status of the supercapacitor at regular intervals and wait for the supercapacitor to complete charging to enable normal system startup under low line current. If the supercapacitor voltage is less than 3.3V, continue to wait; otherwise, proceed to Step 4.

[0088] Step 4: Start the ADC and signal processing module to detect partial discharge signals generated by insulation defects on the overhead power distribution line, collect partial discharge and power frequency voltage data, calculate the amplitude and phase information of the partial discharge, and upload the detection results through the low-power NB-IoT communication module.

[0089] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A multi-node data fusion-based insulation defect monitoring system for overhead power distribution lines, characterized in that, The system includes a Rogowski coil (1), an electromagnetic induction coil (2), a voltage coupler (3), a low-power ARM processor (4), a signal processing circuit (5), an energy harvesting circuit (6), a low-power NB-IoT communication module (7), and a cloud platform (8). The Rogowski coil (1) is electrically connected to the signal processing circuit (5) and is used to detect partial discharge signals generated by insulation defects on the line through the principle of high-frequency current induction. The electromagnetic induction coil (2) is electrically connected to the energy harvesting circuit (6) and is used to induce power frequency voltage from the power frequency current in the line through the principle of magnetic field induction. The energy harvesting circuit (6) is connected to the low-power ARM processor (7), a signal processing circuit (5), an energy harvesting circuit (6), a low-power NB-IoT communication module (7), and a cloud platform (8). 4) Electrically connected to the signal processing circuit (5); used to provide power to the low-power ARM processor (4) and the signal processing circuit (5); the voltage coupler (3) is electrically connected to the low-power ARM processor (4); used to detect the phase information of the line power frequency voltage based on the principle of voltage mutual inductance; the low-power NB-IOT communication module (7) is electrically connected to the low-power ARM processor (4) and the cloud platform (8) respectively; used to monitor the communication and data transmission between the monitoring system and the cloud platform (8); the cloud platform (8) is used to monitor the operating status of the system at different nodes, and to perform fusion analysis and diagnosis of the monitoring data of multiple monitoring nodes; The low-power ARM processor (4) is used to implement the core control and algorithm functions of the monitoring system and coordinate the stable operation of each module; the signal processing circuit (5) is used to preprocess the partial discharge signal detected by the Rogowski coil (1).

2. The multi-node data fusion-based overhead power line insulation defect monitoring system according to claim 1, characterized in that, The signal processing circuit (5) consists of a filter amplifier circuit (51) and a peak hold circuit (52), which is used to improve the signal-to-noise ratio and reduce the signal frequency to avoid the use of high-speed data acquisition systems.

3. The multi-node data fusion-based overhead power line insulation defect monitoring system according to claim 1, characterized in that, The energy harvesting circuit (6) consists of an overvoltage protection circuit (61), a rectifier circuit (62), an over-energy release circuit (63), and an energy management circuit (64), and is used to perform overvoltage protection, rectification and voltage regulation, and energy management on the power frequency voltage collected by the electromagnetic induction coil (2).

4. The multi-node data fusion-based overhead power line insulation defect monitoring system according to claim 1, characterized in that, The voltage coupler (3) is made of a semi-annular copper plate and is used to provide a phase reference for the arrival time of partial discharge and the PRPD spectrum.

5. The multi-node data fusion-based overhead power line insulation defect monitoring system according to claim 1, characterized in that, The low-power ARM processor (4) uses the STM32L476RG low-power processor.

6. The multi-node data fusion-based overhead power line insulation defect monitoring system according to claim 1, characterized in that, The low-power NB-IoT communication module (7) uses the BC26 communication module.

7. The multi-node data fusion-based overhead power line insulation defect monitoring system according to claim 1, characterized in that, The cloud platform (8) performs fusion analysis and diagnosis on monitoring data at different monitoring nodes, including single-node edge monitoring diagnosis and multi-node data fusion diagnosis. In single-node edge monitoring diagnosis, a single monitoring node uploads amplitude and partial discharge phase spectrum information to the cloud platform (8). The cloud platform (8) diagnoses the severity and type of defects based on the amplitude and partial discharge phase spectrum detected by the node. The cloud platform (8) performs fusion analysis on the detection amplitude information uploaded by multiple monitoring nodes, calculates the amplitude offset Δm of each node using the following formula, and compares the Δm at different monitoring nodes horizontally to determine the location and severity of defects. Where: U is the amplitude at a certain monitoring node, U av is the mean of the amplitudes of all monitoring nodes, and k is the number of monitoring nodes.

8. A method for monitoring insulation defects in overhead distribution lines using multi-node data fusion, employing the multi-node data fusion system for monitoring insulation defects in overhead distribution lines as described in any one of claims 1-7, characterized in that, Includes the following steps: Step 1: After installing the monitoring system on the overhead power distribution line, use the electrical energy obtained from the magnetic induction coil to start the energy harvesting circuit and the low-power ARM processor. Step 2: Start the low-power NB-IoT communication module and determine if the system communication is normal. If communication is abnormal, reset the low-power NB-IoT communication module; otherwise, proceed to Step 3. Step 3: Check the charging status of the supercapacitor at regular intervals, waiting for the supercapacitor to complete charging to achieve normal system startup under low line current. If the supercapacitor voltage is less than 3.3V, continue to delay; otherwise, proceed to Step 4. Step 4: Start the ADC and signal processing module to detect partial discharge signals generated by insulation defects on the overhead power distribution line, collect partial discharge and power frequency voltage data, calculate the amplitude and phase information of the partial discharge, and upload the detection results through the low-power NB-IoT communication module.

Images