A self-diagnostic modular surge arrester condition monitoring system

Abstract

The application discloses a self-diagnosis modular lightning arrester state monitoring system, which comprises a sensing module, an extension module, a power supply module, a communication module and a processing module; the sensing module is used for collecting full current signals, surface temperature signals and partial discharge signals of the lightning arrester; the extension module is used for accessing environmental humidity signals and internal pressure signals of the lightning arrester through a controller area network bus interface; the power supply module is used for extracting energy from induced current of a grounding down lead of the lightning arrester to provide electric energy required by system operation; the communication module is used for sending monitoring data to a remote operation and maintenance server; and the processing module is used for receiving various sensor signals, extracting a resistive current component by using a fast Fourier transform algorithm, constructing a three-dimensional feature vector containing a resistive current fundamental component, a partial discharge pulse number and a temperature change rate, and outputting a fault type recognition result through a honeybee algorithm optimization self-adaptive network fuzzy reasoning system classifier.

Description

Technical Field

[0001] This invention belongs to the field of power equipment online monitoring technology, and relates to a self-diagnostic modular surge arrester condition monitoring system. Background Technology

[0002] Surge arresters are critical equipment in power systems. Their main function is to limit overvoltage and protect the insulation of electrical equipment. During long-term operation, surge arresters are subject to the combined effects of electrical stress and environmental factors. This can lead to aging or moisture absorption of their internal varistors. Timely monitoring of the operating status of surge arresters is crucial for ensuring power grid safety. Currently, surge arrester condition monitoring mainly relies on single-parameter detection. Common methods include monitoring only the total current or only the temperature. This single-dimensional monitoring approach has significant limitations and cannot comprehensively reflect the complex operating conditions of the equipment. When a sensor malfunctions, the system often cannot effectively identify it, which can easily lead to misjudgments or missed diagnoses.

[0003] Existing monitoring systems also suffer from poor scalability. Most systems employ a fixed architecture, requiring large-scale modifications to both hardware and software when new monitoring parameters need to be added. This not only increases upgrade costs but also extends the construction period. Furthermore, traditional systems have limited data processing capabilities, typically failing to achieve intelligent identification and classification of fault types.

[0004] To address the aforementioned issues, those skilled in the art urgently need a new technical solution to achieve accurate assessment of the status of surge arresters. Summary of the Invention

[0005] To address the problems existing in the background technology, this invention proposes a self-diagnostic modular surge arrester condition monitoring system.

[0006] The first aspect of this application provides a self-diagnostic modular surge arrester condition monitoring system, comprising:

[0007] Sensing module, expansion module, power supply module, communication module and processing module;

[0008] The sensing module is used to collect the full current signal, surface temperature signal and partial discharge signal of the surge arrester.

[0009] The expansion module is used to access ambient humidity signals and surge arrester internal pressure signals through the controller area network bus interface.

[0010] The power supply module is used to extract energy from the induced current in the grounding lead of the surge arrester to provide the electrical energy required for system operation.

[0011] The communication module is used to send monitoring data to a remote operation and maintenance server;

[0012] The processing module is used to receive signals from various sensors, extract resistive current components using the Fast Fourier Transform algorithm, construct a three-dimensional feature vector containing the fundamental component of resistive current, the number of partial discharge pulses, and the rate of temperature change, and output the fault type identification result through an adaptive network fuzzy inference system classifier optimized by the Bee Algorithm.

[0013] Optionally, the expression for the three-dimensional feature vector F constructed by the processing module is:

[0014] ;

[0015] Where F is, To represent the fundamental component of resistive current, To represent the fundamental reference value of resistive current, To represent the number of partial discharge pulses, To indicate the maximum permissible number of partial discharges, To represent the amount of temperature change, To show the change over time, This represents the threshold for the rate of temperature change.

[0016] Optionally, the processing module is further configured to execute a self-diagnosis mechanism, which includes calculating the characteristic vector imbalance of the three-phase surge arrester, and generating a three-phase abnormal alarm signal when the characteristic vector imbalance exceeds a preset threshold and continues for a preset duration.

[0017] Optionally, the processing module is used to compare the relationship between the resistive current growth rate and the temperature change rate. When the resistive current growth rate exceeds a preset ratio and the temperature change rate is lower than a preset value, the current sensor is determined to be faulty and a device self-test alarm signal is generated.

[0018] Optionally, the processing module is used to compare the relationship between the resistive current growth rate and the temperature change rate. When the resistive current growth rate exceeds a preset ratio and the temperature change rate is lower than a preset value, the current sensor is determined to be faulty and a device self-test alarm signal is generated.

[0019] Optionally, the power supply module includes a micro-current energy harvesting unit and a supercapacitor energy storage unit. The micro-current energy harvesting unit is used to convert the current in the grounding lead into DC power, and the supercapacitor energy storage unit is used to store power and maintain system operation when the main power supply is disconnected.

[0020] Optionally, the expansion module adopts a plug-and-play structure, and the processing module automatically configures the sampling frequency and alarm threshold according to the type of sensor connected.

[0021] Optionally, the adaptive network fuzzy inference system classifier optimized by the bee algorithm optimizes the initial parameters and is trained using a training dataset that includes normal operating conditions, internal damp operating conditions, valve plate aging operating conditions, and external dirty operating conditions, so as to output fault type identification results.

[0022] Optionally, the processing module performs the Fast Fourier Transform algorithm with a transformation length of 4096 points and calculates the resistive current value in conjunction with the reference voltage phase.

[0023] Optionally, the sensing module includes a zero flux current sensor, an infrared temperature sensor, and a partial discharge sensor. The zero flux current sensor is used to acquire a full current signal, the infrared temperature sensor is used to acquire a surface temperature signal, and the partial discharge sensor is used to acquire a local discharge signal.

[0024] Compared with the prior art, the present invention has the following beneficial effects:

[0025] This invention provides a self-diagnostic modular surge arrester condition monitoring system. This system represents a significant technological advancement and has practical value. Firstly, the system integrates a sensing module and an expansion module. This enables multi-dimensional monitoring of total current, surface temperature, partial discharge, ambient humidity, and internal pressure. The monitoring data is more comprehensive and can accurately reflect the operating status of the surge arrester.

[0026] Secondly, the system employs a power supply module. This module draws energy from the induced current in the grounding lead. This solves the problem of traditional monitoring devices relying on external power sources, improving the system's power supply reliability and installation flexibility, making it particularly suitable for long-term monitoring of field or high-altitude equipment.

[0027] Furthermore, the processing module incorporates an optimized Fast Fourier Transform (FFT) algorithm. This optimized FFT algorithm can accurately extract the resistive current component. Combined with a three-dimensional feature vector and a classifier based on an optimized adaptive network fuzzy inference system using the Bee Algorithm, it achieves intelligent identification of fault types, significantly improving diagnostic accuracy.

[0028] Finally, the system has a complete self-diagnostic mechanism. By calculating the imbalance of the three-phase characteristic vector and analyzing the rationality of sensor data, it can detect abnormalities in the device itself in a timely manner, effectively avoiding false alarms caused by sensor failures and ensuring the stable operation of the monitoring system. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of a self-diagnostic modular surge arrester condition monitoring system according to an embodiment of the present invention. Detailed Implementation

[0030] 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.

[0031] In one embodiment, such as Figure 1 As shown, a self-diagnostic modular surge arrester condition monitoring system is provided. This system includes: a sensing module, an expansion module, a power supply module, a communication module, and a processing module. Detailed descriptions of each functional module are as follows:

[0032] The sensing module is used to collect the total current signal, surface temperature signal, and partial discharge signal of the surge arrester. The expansion module is used to access the ambient humidity signal and the internal pressure signal of the surge arrester through the controller area network bus interface. The power supply module is used to extract energy from the induced current of the surge arrester's grounding lead to provide the power required for system operation. The communication module is used to send the monitoring data to a remote operation and maintenance server. The processing module is used to receive the signals from each sensor, extract the resistive current component using the fast Fourier transform algorithm, construct a three-dimensional feature vector containing the fundamental component of the resistive current, the number of partial discharge pulses, and the temperature change rate, and output the fault type identification result through an adaptive network fuzzy inference system classifier optimized by the bee algorithm.

[0033] In one embodiment, the sensing module is used to collect the full current signal, surface temperature signal, and partial discharge signal of the surge arrester.

[0034] The sensing module is fundamental to this system's acquisition of the surge arrester's operating status. This module consists of a zero-flux current sensor, an infrared temperature sensor, and a partial discharge sensor. These sensors work together to ensure the accuracy and reliability of the monitoring data. The zero-flux current sensor is used to acquire the surge arrester's full current signal. Employing the magnetic balance principle, it features high precision and a wide bandwidth, effectively suppressing external magnetic field interference and ensuring accurate current data acquisition even in complex electromagnetic environments. The zero-flux current sensor is installed on the surge arrester's grounding lead to monitor changes in leakage current in real time. The infrared temperature sensor is used to acquire the surge arrester's surface temperature signal. Measuring the infrared radiation intensity of the equipment casing non-contactly, it quickly responds to temperature changes, reflecting the thermal state of the internal varistor. An abnormal increase in surface temperature often indicates internal aging or moisture defects. The partial discharge sensor is used to acquire the surge arrester's partial discharge signal. Its detection frequency range is 10 kHz to 10 MHz. Within this band, it can effectively capture partial discharge pulses caused by insulation degradation. The sensor adopts a high-frequency current transformer structure and is installed on the grounding wire, which has good anti-interference ability.

[0035] For example, the three sensors described above acquire data simultaneously. The zero-flux current sensor outputs an analog current signal, which is converted into a digital quantity by a signal conditioning circuit. The infrared temperature sensor directly outputs a digital temperature value. The partial discharge sensor amplifies the weak discharge pulse before sampling. In this invention, the simultaneous acquisition of multiple parameters improves the comprehensiveness of fault identification. The mutual verification between different physical quantities enhances the reliability of the diagnosis. The application of high-precision sensors improves the overall measurement accuracy of the system. The wideband partial discharge detection capability enhances the sensitivity to early insulation defects. Non-contact temperature measurement avoids affecting the insulation performance of the equipment. The entire module has a compact structure, is easy to install, and does not affect the normal operation of the surge arrester.

[0036] Furthermore, the sensing module includes a zero flux current sensor, an infrared temperature sensor, and a partial discharge sensor. The zero flux current sensor is used to acquire the full current signal, the infrared temperature sensor is used to acquire the surface temperature signal, and the partial discharge sensor is used to acquire the discharge signal in the frequency band of 10 kHz to 10 MHz.

[0037] The sensing module is fundamental to this system's acquisition of the surge arrester's operating status. This module consists of a zero-flux current sensor, an infrared temperature sensor, and a partial discharge sensor. These sensors work together to ensure the accuracy and reliability of the monitoring data. The zero-flux current sensor is used to acquire the surge arrester's full current signal. This sensor employs the magnetic balance principle, featuring high precision and wide bandwidth characteristics, effectively suppressing external magnetic field interference, ensuring accurate current data acquisition even in complex electromagnetic environments. The sensor is installed on the surge arrester's grounding lead to monitor changes in leakage current in real time. The infrared temperature sensor is used to acquire the surge arrester's surface temperature signal. This sensor measures the infrared radiation intensity of the equipment casing in a non-contact manner, rapidly responding to temperature changes and reflecting the thermal state of the internal varistor. An abnormal increase in surface temperature often indicates internal aging or moisture defects. The partial discharge sensor is used to acquire the surge arrester's partial discharge signal. This sensor's detection frequency range is 10 kHz to 10 MHz. Within this frequency band, it can effectively capture partial discharge pulses caused by insulation degradation. The sensor adopts a high-frequency current transformer structure, installed on the grounding wire, and has good anti-interference capabilities. For example, the three sensors described above acquire data simultaneously. The zero-flux current sensor outputs an analog current signal, which is converted into a digital value by a signal conditioning circuit. The infrared temperature sensor directly outputs a digital temperature value. The partial discharge sensor amplifies the weak discharge pulse before sampling.

[0038] In this invention, simultaneous acquisition of multiple parameters improves the comprehensiveness of fault identification. Mutual verification between different physical quantities enhances the reliability of the diagnosis. The application of high-precision sensors improves the overall measurement accuracy of the system. Wideband partial discharge detection capability enhances sensitivity to early insulation defects. Non-contact temperature measurement avoids impacting the insulation performance of the equipment. The entire sensing module has a compact structure, is easy to install, and does not affect the normal operation of the surge arrester.

[0039] In one embodiment, the expansion module is used to access ambient humidity signals and surge arrester internal pressure signals via a controller area network bus interface.

[0040] The expansion module enhances the system's adaptability and functionality. It enables flexible access to external sensors via a Controller Area Network (CLAN) bus interface, primarily for acquiring ambient humidity and surge arrester internal pressure signals. The CLAN bus interface is a serial communication protocol with high reliability and real-time performance, widely used in industrial control. It supports multi-node connections, allowing for the addition or removal of devices without interrupting system operation. An ambient humidity sensor is connected to the system via this interface. Deployed around the surge arrester, it monitors the external humidity level. Simultaneously, the surge arrester's internal pressure sensor is also connected via the CLAN bus interface. This internal pressure sensor, installed within the surge arrester's sealed structure, senses changes in internal gas pressure; abnormal pressure is typically associated with deteriorated sealing performance or internal discharge.

[0041] For example, when a new sensor is connected to the controller's local area network bus, the processing module automatically identifies its device type. The system then configures the corresponding sampling frequency and communication parameters based on the identification result. This plug-and-play mechanism simplifies the on-site installation and commissioning process.

[0042] In this invention, by incorporating environmental humidity data, the system can distinguish between changes in leakage current caused by external pollution and internal faults. Internal pressure monitoring provides direct evidence for determining the sealing status of the equipment. Modular design makes system function expansion more convenient. New sensors can be integrated without changing the main control hardware. The controller LAN bus has strong anti-interference capabilities, ensuring the stability of signal transmission. The entire expansion module is seamlessly integrated with the core system, improving the comprehensiveness and intelligence of monitoring.

[0043] In one embodiment, the expansion module is used to access ambient humidity signals and surge arrester internal pressure signals via a controller area network bus interface.

[0044] The expansion module connects to external sensor signals via the Controller Area Network (CLAN) bus interface. Its extended functions include acquiring ambient humidity and surge arrester internal pressure signals. The CLAN bus is a serial communication network supporting multi-node connections, offering high reliability and real-time communication capabilities. Employing a multi-master operating mode, the CLAN bus allows nodes to actively send information, achieving a communication rate of up to 1 Mbit per second, meeting the real-time transmission requirements of monitoring data. The CLAN bus structure facilitates quick and easy integration of new devices without interrupting the main system program. The ambient humidity sensor connects to the system via the CLAN bus interface. Deployed near the surge arrester, it monitors the relative humidity of the surrounding air. Humidity data helps determine the impact of the external environment on the equipment's insulation performance. Excessive humidity may exacerbate surface leakage current. The surge arrester internal pressure sensor, also connected via the CLAN bus interface, is installed within the sealed cavity of the surge arrester body. It senses changes in internal gas pressure; abnormal pressure may indicate sealing failure or internal discharge.

[0045] For example, after the expansion module is powered on, the system automatically scans new nodes on the bus, identifies the sensor type, and configures the corresponding sampling parameters. The processing module periodically reads sensor data and stores and analyzes it. In this invention, by accessing the ambient humidity signal, the system can distinguish between state changes caused by internal defects and external contamination, and internal pressure monitoring provides an effective basis for judging the equipment's sealing performance. The modular design makes system function expansion simple and easy, allowing the integration of new monitoring parameters without hardware changes. The controller area network bus has strong anti-electromagnetic interference capabilities, ensuring the stability of signal transmission. The entire expansion module works seamlessly with the system, improving the comprehensiveness and intelligence of monitoring.

[0046] Furthermore, the expansion module adopts a plug-and-play structure, and the processing module automatically configures the sampling frequency and alarm threshold according to the type of sensor connected.

[0047] The expansion module features a flexible hardware and software architecture with a plug-and-play structure, enabling easy integration of external sensors into the system. The processing module possesses automatic identification and configuration capabilities. When a new sensor is connected, the system can operate without manual intervention. The plug-and-play structure relies on standardized hardware interfaces and communication protocols. The Controller Area Network (CLAN) bus serves as the underlying communication network, supporting multi-node mounting. Each expansion sensor has a unique electronic tag storing sensor type, measurement range, and manufacturer information. When a sensor connects to the CLAN bus, the processing module first reads its electronic tag information. Based on the identified sensor type, the processing module automatically loads the corresponding driver. Subsequently, the system retrieves the default configuration parameters for that type of sensor from its built-in database. These parameters include sampling frequency and alarm thresholds. For example, when an ambient humidity sensor is connected, the system automatically configures a lower sampling frequency because ambient humidity changes relatively slowly. Simultaneously, the system loads an alarm threshold range specific to humidity. If an internal pressure sensor for a surge arrester is connected, the processing module configures a higher sampling frequency, as the pressure signal may fluctuate rapidly in the early stages of a fault, requiring high-frequency monitoring; the corresponding alarm threshold is also set according to the characteristics of the pressure sensor. This automatic configuration mechanism ensures the rationality and effectiveness of data collection.

[0048] For example, maintenance personnel connect a new sensor to the interface terminal of the expansion module. After the system is powered on, the processing module sends a broadcast query command. The newly connected sensor responds with its identity information. After the processing module parses the information, it confirms that the sensor is an "internal pressure sensor". Subsequently, the system automatically sets the sampling frequency to 1 Hz and sets the upper limit of the pressure alarm threshold to 0.6 MPa and the lower limit to 0.4 MPa.

[0049] In this invention, the plug-and-play structure greatly simplifies on-site installation and commissioning, reducing the professional skills required of maintenance personnel. Automatic configuration of the sampling frequency avoids data loss or redundancy caused by improper parameter settings. Matching alarm thresholds based on sensor characteristics improves the accuracy of fault warnings. The modular design makes system function expansion very convenient without modifying the core program code. The entire system has good compatibility and maintainability, adapting to the monitoring needs of different application scenarios.

[0050] In one embodiment, the power supply module is used to extract energy from the induced current in the grounding lead of the surge arrester to provide the electrical energy required for system operation.

[0051] The power supply module utilizes the principle of electromagnetic induction to extract energy from the grounding lead of the surge arrester, solving the problem of traditional monitoring devices relying on external power supplies. The power supply module mainly consists of a micro-current energy harvesting unit and a supercapacitor energy storage unit.

[0052] The micro-current energy harvesting unit is an energy extraction component that uses a high-permeability magnetic core material and is wound around the grounding down conductor of the surge arrester. When the surge arrester is in operation, a weak leakage current flows through the grounding down conductor. This leakage current generates an alternating magnetic field in the surrounding space. The micro-current energy harvesting unit converts the magnetic field energy into an induced electromotive force by cutting the magnetic field lines through an induction coil.

[0053] Because the leakage current of a surge arrester is extremely weak during normal operation, the induced voltage is often low and unstable. Therefore, the micro-current energy harvesting unit integrates a high-efficiency rectifier and filter circuit. The rectifier and filter circuit converts the unstable AC induced current into pulsating DC current, which is then regulated before being used as input for the subsequent energy storage module.

[0054] The supercapacitor energy storage unit is electrically connected to the microcurrent harvesting unit, and its main function is to store electrical energy. Because the current in the grounding lead fluctuates with the grid load, direct power supply could lead to system instability. The supercapacitor energy storage unit acts as an energy buffer, storing the energy collected by the microcurrent harvesting unit and releasing a stable voltage when the system needs it.

[0055] For example, the microcurrent energy harvesting unit adopts an open-type structure design. The open-type structure design makes it easy to install on the grounding lead without power interruption. When the grounding current reaches the microampere level, the microcurrent energy harvesting unit can start working. The generated electrical energy is regulated by the power management chip to power the processing module and the communication module. The excess electrical energy is charged into the supercapacitor energy storage unit for use when the current is too low.

[0056] In this invention, the power supply module completely eliminates reliance on external mains power or disposable batteries. This significantly reduces system maintenance costs, eliminating the need for periodic battery replacements. Energy is extracted locally using the induced current in the grounding conductor. This method does not alter the original wiring structure, making installation convenient and safe. The micro-current energy harvesting unit can start with extremely low current, ensuring continuous power supply during the normal operation of the surge arrester. The introduction of the supercapacitor energy storage unit ensures the continuity of power supply during current fluctuations or transient processes. The entire power supply solution is green and environmentally friendly, making it particularly suitable for complex environments where laying power lines is difficult, such as in the field or at high altitudes.

[0057] Furthermore, the power supply module includes a micro-current energy harvesting unit and a supercapacitor energy storage unit. The micro-current energy harvesting unit is used to convert the current in the grounding lead into DC power, and the supercapacitor energy storage unit is used to store power and maintain system operation when the main power supply is disconnected.

[0058] The microcurrent energy harvesting unit is used for energy capture and conversion. During normal operation of the surge arrester, a weak leakage current flows through the grounding lead. This leakage current contains fundamental and harmonic components. The microcurrent energy harvesting unit utilizes the principle of electromagnetic induction, inducing an alternating electromotive force (EMF) through a high-permeability magnetic core surrounding the grounding lead. This induced EMF is converted into pulsating direct current by a rectifier circuit. Subsequently, a filter circuit smooths the pulsation components, outputting stable DC power. The design of the microcurrent energy harvesting unit must balance energy harvesting efficiency and load influence. Amorphous alloys or nanocrystalline alloys are selected for the magnetic core material to improve permeability and reduce losses. The number of coil turns is optimized to ensure sufficient induced voltage is generated even under microampere-level currents. The rectifier circuit uses Schottky diodes with low forward voltage drop to reduce energy loss during rectification. The supercapacitor energy storage unit undertakes the tasks of energy storage and power compensation. Supercapacitors possess Faraday quasi-capacitor characteristics, with higher energy density than traditional capacitors and a longer cycle life than chemical batteries. The supercapacitor energy storage unit consists of multiple supercapacitor cells connected in series. The series structure improves the energy storage voltage level, meeting the system's power supply requirements. Voltage equalization circuits are configured between individual units to ensure balanced voltage distribution and prevent overvoltage damage. The supercapacitor energy storage unit plays a dual role in system operation. Under normal operating conditions, it stores the electrical energy converted by the micro-current harvesting unit. When the system starts up or data is transmitted, and the instantaneous power demand increases, the supercapacitor energy storage unit releases the stored electrical energy to supplement the instantaneous power gap. When the main power supply is disconnected due to line faults or maintenance, the supercapacitor energy storage unit serves as a backup power source, maintaining the operation of the system's core modules.

[0059] For example, the magnetic core of the micro-current energy harvesting unit surrounds the grounding lead of the surge arrester. The AC voltage induced by the coil is converted into DC voltage by a bridge rectifier circuit. The DC voltage is then regulated by a DC-DC converter to charge the supercapacitor energy storage unit. The output of the supercapacitor energy storage unit is connected to the system power management module. The power management module dynamically allocates power according to the system load demand. When the grounding lead current is lower than the energy harvesting threshold, the supercapacitor energy storage unit is powered independently. When the main power supply is disconnected, the supercapacitor energy storage unit maintains the operation of the data processing module and the communication module, ensuring that fault data is not lost. The system utilizes the high power density characteristics of supercapacitors to support the communication module in completing the final data upload task.

[0060] In this invention, the micro-current energy harvesting unit efficiently extracts energy from the grounding lead, eliminating dependence on an external power source. The supercapacitor energy storage unit provides highly reliable energy buffering and backup, solving the problems of short lifespan and difficult maintenance associated with traditional batteries. The combination of these two components ensures continuous power supply to the monitoring system under various operating conditions, improving the system's independence and stability. This power supply scheme adapts to the complex electromagnetic environment at the surge arrester site, providing reliable energy security for long-term online monitoring.

[0061] In one embodiment, the communication module is used to send monitoring data to a remote operation and maintenance server.

[0062] The communication module is responsible for uploading processed monitoring data and fault diagnosis results to the cloud. It adopts a dual-mode communication architecture, integrating a long-range radio communication unit and a mobile communication technology unit. This design takes into account the communication needs of different scenarios, ensuring the stability and real-time performance of data transmission.

[0063] Long-range radio communication units are primarily used for short-range, interference-resistant data transmission. Operating in unlicensed frequency bands, they feature strong diffraction capabilities and long transmission distances. In areas with complex electromagnetic environments, such as substations, long-range radio communication units can penetrate metal shielding to transmit data to the on-site aggregation gateway. Their low-power characteristics help reduce overall system energy consumption.

[0064] The mobile communication technology unit is used for remote data transmission. When the monitoring point is located in an area with good public network signal coverage, the mobile communication technology unit establishes a connection through the mobile base station network and directly sends the monitoring data to the remote operation and maintenance server, without the need for additional gateway equipment. This plug-and-play approach greatly simplifies on-site construction.

[0065] The communication module supports the International Electrotechnical Commission (IEC) 61850 standard communication protocol. The IEC 61850 standard communication protocol is the common language for power system automation. By adopting this protocol, monitoring data can be encapsulated in a standardized format, enabling the system to seamlessly connect to existing power monitoring networks and achieve interoperability between equipment from different manufacturers.

[0066] For example, the communication module first attempts to establish a connection using a mobile communication technology unit. If a weak or interrupted public network signal is detected, the system automatically switches to a long-distance radio communication unit. Data is forwarded to the remote maintenance server via the on-site aggregation node. The processing module packages the resistive current characteristic value, temperature data, and fault alarm information and sends them in the format specified by the International Electrotechnical Commission (IEC) 61850 protocol.

[0067] In this invention, the dual-mode communication architecture serves as a backup for each other, greatly improving the reliability of the communication link. The long-distance radio communication unit solves the problems of signal penetration and interference resistance in complex environments. The mobile communication technology unit provides a convenient wide area network access method. Support for the International Electrotechnical Commission (IEC) 61850 standard protocol breaks down information silos and achieves standardized access for the monitoring system. Flexible data transmission strategies ensure that maintenance personnel can obtain the operating status of surge arresters in a timely manner, providing strong support for the safe dispatching of the power grid.

[0068] Furthermore, the communication module includes a long-range radio communication unit and a mobile communication technology unit. The long-range radio communication unit is used for short-range anti-interference data transmission, and the mobile communication technology unit is used for remote data backhaul. The communication module supports the International Electrotechnical Commission 61850 standard communication protocol.

[0069] The communication module is a key component for long-distance data transmission in this system. This module adopts a dual-mode communication architecture, integrating a long-distance radio communication unit and a mobile communication technology unit, thus solving the problem of insufficient reliability of a single communication method in complex environments. The communication module fully supports the International Electrotechnical Commission (IEC) 61850 standard communication protocol.

[0070] Long-range radio communication units are primarily responsible for short-range, interference-resistant data transmission. They operate in the unlicensed sub-gigahertz frequency band. Unlicensed band signals have longer wavelengths, possessing strong diffraction and penetration capabilities. In environments such as substations with extensive metal shielding and complex electromagnetic interference, long-range radio communication units can effectively suppress noise using spread spectrum technology, typically transmitting data to a nearby aggregation gateway. This transmission method has extremely low power consumption, making it suitable for battery-powered or low-energy-harvesting monitoring equipment.

[0071] The mobile communication technology unit is primarily used for remote data transmission. It utilizes existing public mobile communication networks for data transmission. When the monitoring point is located in an area with good public network signal coverage, the mobile communication technology unit can establish a high-speed data link, directly sending monitoring data to the remote operation and maintenance server. This eliminates the need for users to build their own communication backbone network, significantly reducing networking costs.

[0072] The communication module integrates a protocol stack compliant with the International Electrotechnical Commission (IEC) 61850 standard. The IEC 61850 standard protocol is a universal standard in the field of power system automation. The communication module models the collected monitoring data, such as total current, temperature, and partial discharge, according to the common data classes defined in this standard. The data is encapsulated into a manufacturing message specification format, enabling the monitoring device to be directly connected to the power monitoring system like an intelligent electronic device.

[0073] For example, the communication module has an automatic link switching function. The system first attempts to connect to the server via the mobile communication technology unit. If the public network signal strength is detected to be lower than a set threshold, the communication module automatically switches to the long-distance radio communication unit. Data is converted into a protocol by the on-site aggregation node and then uploaded via fiber optic or wireless network. During the data transmission phase, the communication module reads the status information generated by the processing module, maps it to a logical node object in the IEC 61850 standard, and sends it to the master station through the manufacturing message specification service.

[0074] In this invention, the dual-mode communication architecture serves as a backup for each other, greatly improving the reliability of the communication link. The long-distance radio communication unit solves the problems of signal penetration and anti-interference in complex electromagnetic environments. The mobile communication technology unit provides a convenient wide area network access method. Supporting the International Electrotechnical Commission (IEC) 61850 standard protocol breaks down information silos and achieves standardized access for the monitoring system. Flexible data transmission strategies ensure that maintenance personnel can obtain the operating status of surge arresters in a timely manner, providing strong support for the safe dispatching of the power grid.

[0075] The processing module is used to receive signals from various sensors, extract resistive current components using the Fast Fourier Transform algorithm, construct a three-dimensional feature vector containing the fundamental component of resistive current, the number of partial discharge pulses, and the rate of temperature change, and output the fault type identification result through an adaptive network fuzzy inference system classifier optimized by the Bee Algorithm.

[0076] The processing module is the computational and analytical unit of this system. Electrically connected to each sensor, it receives full-current signals, partial discharge signals, and surface temperature signals. Internally, it runs high-precision signal processing algorithms and intelligent diagnostic models. The processing module mainly consists of a signal preprocessing unit, a feature extraction unit, and an intelligent classifier unit.

[0077] The signal preprocessing unit first performs analog-to-digital conversion on the received raw analog signal. The converted digital signal then undergoes digital filtering to remove high-frequency noise and power frequency interference. For the full-current signal, the preprocessing unit uses the Fast Fourier Transform (FFT) algorithm for spectral analysis. The FFT algorithm decomposes the time-domain signal into sinusoidal components of different frequencies. Through calculation, the system can accurately separate the resistive current component that is in phase with the grid voltage. The magnitude of the resistive current component directly reflects the degree of insulation aging of the surge arrester varistors.

[0078] The feature extraction unit constructs a multi-dimensional state space based on the preprocessed data. The first dimension is the fundamental component of the resistive current, which reflects the leakage characteristics of the surge arrester under power frequency voltage. The second dimension is the number of partial discharge pulses, characterizing the presence of air gaps or cracks within the insulating medium. The third dimension is the temperature change rate, reflecting the dynamic changes in the internal thermal equilibrium of the equipment. These three parameters together constitute a three-dimensional feature vector. This three-dimensional feature vector comprehensively describes the electrical, physical, and thermal state of the surge arrester.

[0079] The intelligent classifier unit adopts an adaptive network fuzzy inference system architecture. This system combines the learning capabilities of neural networks with the reasoning abilities of fuzzy logic. To overcome the tendency of traditional neural networks to get trapped in local optima, this embodiment introduces a bee algorithm for optimization. The bee algorithm simulates the foraging behavior of a bee colony, using a search mechanism involving hired bees, observation bees, and scout bees to globally optimize the membership function parameters and conclusion parameters of the adaptive network fuzzy inference system.

[0080] For example, the processing module periodically collects data. The Fast Fourier Transform algorithm calculates the fundamental amplitude of the resistive current, simultaneously counting the number of partial discharge pulses per unit time, and calculating the temperature difference between the current and previous cycles to obtain the temperature change rate. These three values ​​are normalized and then input into a three-dimensional feature vector space. The optimized adaptive network fuzzy inference system classifier receives this vector, and after fuzzification, inference, and defuzzification processes, finally outputs the fault type identification result, which includes states such as normal, damp, aging, or internal discharge.

[0081] In this invention, the Fast Fourier Transform algorithm significantly improves the accuracy of resistive current measurement and effectively filters out interference from capacitive current. A three-dimensional feature vector is constructed to achieve multi-parameter fusion diagnosis, avoiding the limitations of single-index judgment. The adaptive network fuzzy inference system classifier, optimized by the Bee Algorithm, significantly improves the accuracy and convergence speed of fault identification. The processing module can adaptively learn the characteristics of different surge arresters and has a strong ability to identify nonlinear fault characteristics. The processing of this invention is highly automated, requiring no manual intervention, and provides a scientific basis for condition-based maintenance of surge arresters.

[0082] Furthermore, the expression for the three-dimensional feature vector constructed by the processing module is:

[0083] ;

[0084] Where F is a three-dimensional feature vector. To represent the fundamental component of resistive current, To represent the fundamental reference value of resistive current, To represent the number of partial discharge pulses, To indicate the maximum permissible number of partial discharges, To represent the amount of temperature change, To show the change over time, This represents the threshold for the rate of temperature change.

[0085] The processing module transforms multi-source heterogeneous data into a unified feature space using a specific mathematical model. This module constructs a three-dimensional feature vector F. The construction of this vector is based on three key physical quantities: resistive current, partial discharge, and temperature change. By normalizing the original data, the influence of different dimensions on fault diagnosis is eliminated.

[0086] The mathematical expression for the three-dimensional eigenvector is rigorously defined in the system. This expression consists of three components. The first component relates to the fundamental component of the resistive current. Reference value of resistive current fundamental frequency The second component relates to the number of partial discharge pulses. With the maximum permissible number of partial discharges The third component involves the amount of temperature change. Change over time and temperature change rate threshold .

[0087] The fundamental component of the resistive current reflects the resistive characteristics of the surge arrester varistor. The reference value for the fundamental resistive current is usually set to the initial value at the time of equipment delivery or the average value during historical normal operation. The ratio of the two values ​​characterizes the degree of aging or moisture absorption of the insulation material. When this ratio increases significantly, it indicates increased resistive current leakage and decreased insulation performance.

[0088] The number of partial discharge pulses is used to measure the frequency of discharges within the insulating medium. The maximum permissible number of partial discharges is a safe upper limit set according to the equipment's insulation class. The ratio of the two reflects the severity of the discharge phenomenon. The closer this ratio is to or exceeds a unit value, the more serious the insulation defects or air gap breakdown risk within the equipment.

[0089] The ratio of temperature change to time change constitutes the transient temperature change rate. The temperature change rate threshold defines the permissible thermal shock limit of the equipment. The ratio of the temperature change rate to the threshold reflects the dynamic changes in the equipment's thermal stability. Abnormal temperature rise rates usually indicate poor internal contact or deterioration of valve plate nonlinear characteristics.

[0090] For example, the processing module acquires raw data from each sensor in real time. The module first calculates the fundamental component of the current resistive current. Then, it counts the number of partial discharge pulses within a preset time window. Simultaneously, the processing module records the temperature sensor readings and calculates the temperature change per unit time. The system divides these three measured values ​​by their corresponding reference values ​​or thresholds. After calculation, a standardized three-dimensional feature vector is generated, which is then input into a subsequent classifier for pattern recognition.

[0091] In this invention, normalization is achieved by introducing benchmark values ​​and thresholds, thus unifying different physical dimensions. This processing method allows for weighted comparison of current, pulse, and temperature data on the same dimension. The construction of the three-dimensional feature vector fully considers the coupling relationship between electrical characteristics, insulation strength, and thermal characteristics. The ratio-based feature extraction method improves the system's adaptability to different types of surge arresters. The three-dimensional feature vector can sensitively capture subtle changes in equipment status, providing a high-quality data foundation for subsequent intelligent diagnostic algorithms, thereby enhancing the robustness of fault identification.

[0092] Furthermore, the processing module is also used to execute a self-diagnostic mechanism, which includes calculating the characteristic vector imbalance of the three-phase surge arrester. When the characteristic vector imbalance exceeds a preset threshold and continues for a preset duration, a three-phase abnormal alarm signal is generated.

[0093] The processing module possesses advanced self-diagnostic capabilities, identifying abnormal states of surge arrester groups through horizontal comparative analysis. In power systems, surge arresters are typically arranged in a three-phase configuration. Three-phase surge arresters operating under the same conditions should exhibit a high degree of consistency in their operational status. The self-diagnostic mechanism utilizes this characteristic to calculate the eigenvector imbalance of the three-phase surge arresters.

[0094] Eigenvector imbalance is a key indicator for measuring the differences in the condition of three-phase equipment. The processing module acquires the three-dimensional eigenvectors of phases A, B, and C in real time. These three-dimensional eigenvectors contain resistive current, partial discharge, and temperature information. The processing module calculates the current imbalance by comparing the magnitude difference or Euclidean distance of the three-phase eigenvectors. The imbalance can reflect whether a surge arrester in a certain phase shows signs of deviating from its normal state.

[0095] To eliminate false positives caused by transient interference, the self-diagnostic mechanism incorporates time-based judgment logic. The system sets preset thresholds and preset durations; it only determines an anomaly when the feature vector imbalance consistently exceeds the preset threshold and the time span reaches the preset duration. This dual-judgment mechanism effectively filters out transient fluctuations caused by lightning overvoltage or system operation.

[0096] For example, the processing module periodically and synchronously collects three-phase data. Assume that the characteristic vectors of phases A and C remain stable, while the resistive current component of phase B gradually increases due to moisture. At this time, the characteristic vector of phase B deviates from that of the other two phases, causing the overall imbalance to increase. When this imbalance exceeds a preset threshold, for example, set to 20%, and remains there for more than five seconds, the processing module immediately generates a three-phase abnormality alarm signal. The three-phase abnormality alarm signal is sent to the remote operation and maintenance server via the communication module.

[0097] This invention utilizes three-phase imbalance for self-diagnosis, significantly improving the accuracy of fault identification. This method eliminates the influence of common-mode interference such as ambient temperature, humidity, and grid voltage fluctuations. By introducing preset time-based judgment conditions, it effectively avoids false alarms caused by transient interference. The generation of three-phase abnormal alarm signals helps maintenance personnel quickly locate the faulty phase. This invention represents a leap from individual monitoring to group correlation analysis, enhancing the intelligence level of power grid equipment management.

[0098] Furthermore, the processing module is used to compare the relationship between the resistive current growth rate and the temperature change rate. When the resistive current growth rate exceeds a preset ratio and the temperature change rate is lower than a preset value, it determines that the current sensor is faulty and generates a device self-test alarm signal.

[0099] The processing module possesses an intelligent self-test function. This self-test function performs logical judgments based on the coupling mechanism of physical fields. During the operation of a surge arrester, the increase in resistive current is usually accompanied by a thermal effect. According to Joule's law, current flowing through a resistor generates heat. Therefore, a true increase in resistive current should cause an increase in the temperature of the surge arrester itself. The processing module utilizes this physical law to compare the relationship between the rate of increase of resistive current and the rate of change of temperature in real time.

[0100] The system has pre-defined logical judgment conditions. These conditions include a preset ratio and a preset value. The preset ratio defines the significance of the increase in resistive current. The preset value defines the sensitivity to temperature changes. When the processing module detects that the rate of increase in resistive current exceeds the preset ratio, a significant temperature rise should theoretically be observed. If, at this time, the rate of temperature change is lower than the preset value, or even remains stable, it indicates a logical paradox between the electrical signal and the thermal signal.

[0101] This logical paradox typically points to a sensor malfunction. After ruling out extreme heat dissipation anomalies, an abnormal surge in the current signal without a thermal response is highly likely due to zero-point drift, poor wiring contact, or measurement distortion caused by electromagnetic interference in the current sensor. Based on this judgment, the processing module determines that the current sensor is faulty. Subsequently, the module generates a device self-test alarm signal. This device self-test alarm signal is independent of the surge arrester fault signal and is specifically used to prompt maintenance personnel to monitor the device's own health status.

[0102] For example, the processing module calculates the slope of the resistive current change in the most recent sampling period. Assume the resistive current growth rate reaches 15% per hour. Simultaneously, the system reads the temperature sensor reading and calculates the temperature change rate as 0.5 degrees Celsius per hour. If the preset ratio is set to 10% and the preset value is set to 2 degrees Celsius per hour, then the resistive current growth rate exceeds the preset ratio, while the temperature change rate falls below the preset value. The processing module then triggers a logic check, determines that the current sensor data is abnormal, and outputs a device self-test alarm signal.

[0103] In this invention, multi-physical quantity correlation analysis effectively solves the problem of false alarms caused by data distortion from a single sensor. Temperature is used as a physical reference to verify the authenticity of the current data. The introduction of a self-test alarm signal enables the monitoring system to self-monitor, avoiding blind power outages for maintenance due to monitoring device malfunctions. This mechanism improves the reliability of the entire online monitoring system and ensures the scientific and accurate nature of operation and maintenance decisions.

[0104] Furthermore, the adaptive network fuzzy inference system classifier optimized by the bee algorithm optimizes the initial parameters and is trained using a training dataset that includes normal operating conditions, internal damp operating conditions, valve plate aging operating conditions, and external dirty operating conditions, so as to output fault type identification results.

[0105] The adaptive network fuzzy inference system classifier optimized by the bee algorithm serves as the intelligent diagnostic engine of this system. This classifier integrates the reasoning capabilities of fuzzy logic with the self-learning capabilities of neural networks. To overcome the shortcomings of traditional neural networks, such as easily getting trapped in local optima and slow convergence during training, this invention introduces the bee algorithm to globally optimize the initial parameters of the adaptive network fuzzy inference system.

[0106] The performance of adaptive fuzzy inference systems is highly dependent on the initial settings of their premise and conclusion parameters. Premise parameters determine the shape and distribution of the membership functions of the input variables, while conclusion parameters determine the output characteristics of the fuzzy rules. Initializing these parameters randomly often leads to excessively long training times or insufficient diagnostic accuracy. The bee algorithm, by simulating the foraging behavior of bee colonies—including the local development of hired bees, the probabilistic selection of observer bees, and the global search of scout bees—finds the optimal combination of parameters in the solution space.

[0107] During the training phase, the classifier is loaded with a pre-built training dataset. This training dataset covers feature samples of surge arresters under different operating conditions, specifically including normal operating condition samples, internal moisture-affected operating condition samples, valve plate aging operating condition samples, and external pollution-affected operating condition samples. Each set of samples contains feature data such as the fundamental component of resistive current, the number of partial discharge pulses, and the rate of temperature change, and corresponds to a unique fault type label.

[0108] The optimized process of the bee algorithm is as follows: First, the bee algorithm randomly generates a set of initial parameters as nectar source locations. Next, hired bees perform neighborhood searches on these locations and calculate the fitness function value. The fitness function is defined by the root mean square error of the classifier on the training dataset. The smaller the error, the higher the fitness. Observer bees, based on feedback from hired bees, probabilistically select high-quality nectar sources for further development. If a nectar source is not updated within a preset number of iterations, the scout bee abandons that location and generates a new random solution. Through iterative optimization, the optimal initial parameters for the adaptive network fuzzy inference system are finally determined.

[0109] For example, the system first uses the Bee Algorithm to train offline on a dataset containing four operating conditions: normal, internal dampness, valve plate aging, and external contamination. After several iterations, the Bee Algorithm outputs a set of optimal membership function parameters. These parameters are then assigned to the adaptive network fuzzy inference system classifier. During the online monitoring phase, the classifier receives real-time collected 3D feature vectors, performs fuzzy inference calculations, and outputs the membership degree for each fault type. Finally, the system selects the category with the highest membership degree as the fault type identification result.

[0110] In this invention, the introduction of the bee algorithm significantly improves the convergence speed and global optimization capability of the classifier. By optimizing the initial parameters, the classifier avoids getting trapped in local minima, thereby improving the accuracy of fault diagnosis. The training dataset covering various typical operating conditions gives the classifier broad adaptability and generalization ability. This classifier can effectively distinguish between easily confused faults such as internal moisture and external contamination, providing a reliable decision-making basis for the condition-based maintenance of surge arresters.

[0111] Furthermore, the processing module executes the optimized Fast Fourier Transform algorithm with a transformation length of 4096 points, and calculates the resistive current value in conjunction with the reference voltage phase.

[0112] When performing signal analysis tasks, the processing module employs a high-precision spectrum analysis strategy. This strategy involves the application of an optimized Fast Fourier Transform (FFT) algorithm. To meet the high standards of harmonic analysis required by power systems, the processing module sets the transform length to 4096 points. This setting significantly improves the frequency domain resolution, enabling the system to more precisely distinguish between the fundamental frequency and each harmonic component.

[0113] The choice of a 4096-point transform length is based on a comprehensive consideration of the sampling theorem and computational efficiency. In a 50 Hz power system, a longer sampling sequence means a narrower frequency bandwidth. By acquiring 4096 consecutive current sampling points, the processing module constructs a high-density time-domain data window. This time-domain data window, after windowing processing, is input into the Fast Fourier Transform algorithm for computation.

[0114] The Fast Fourier Transform (FFT) algorithm in this invention is an optimization of the traditional algorithm, introducing a spectrum correction mechanism to address spectral leakage and the picket fence effect caused by asynchronous sampling. The optimized FFT algorithm first performs coarse spectral analysis on 4096 data points to locate the fundamental peak. Then, it uses spectral information near the peak for interpolation to accurately correct the fundamental frequency, amplitude, and phase, significantly reducing measurement errors.

[0115] The calculation of resistive current values ​​relies on accurate phase information. The processing module simultaneously acquires a reference voltage signal. This reference voltage signal is typically derived from a voltage transformer and represents the ideal phase reference of the power grid. An optimized Fast Fourier Transform algorithm calculates the phase difference between the fundamental component of the total current and the reference voltage phase. Since the resistive current is in phase with the voltage, while the capacitive current has a 90-degree phase difference with the voltage, the system projects the fundamental vector of the total current onto the direction of the reference voltage vector using a projection principle.

[0116] For example, the processing module acquires signals at a sampling rate of 8 kHz. After approximately 51 milliseconds of accumulation, 4096 data points are collected. The processing module simultaneously executes an optimized Fast Fourier Transform (FFT) algorithm on the voltage and current signals. The optimized FFT algorithm outputs the phase angle of the fundamental current wave and the phase angle of the reference voltage. The system calculates the difference between the two to obtain the phase cosine value of the resistive current. Finally, the amplitude of the fundamental current wave is multiplied by this cosine value to obtain the accurate resistive current value.

[0117] In this invention, the use of a 4096-point transform length significantly improves frequency resolution, effectively distinguishing the fundamental frequency from adjacent interharmonics. The optimized Fast Fourier Transform algorithm, through spectral correction, eliminates the impact of spectral leakage on measurement accuracy. Combined with the reference voltage phase calculation, it achieves precise separation of resistive and capacitive components. This method is unaffected by grid frequency fluctuations, ensuring the authenticity and reliability of resistive current monitoring data under complex operating conditions, and providing accurate quantitative indicators for assessing the insulation status of surge arresters.

[0118] Each module in the aforementioned self-diagnostic modular surge arrester condition monitoring system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0119] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A self-diagnostic modular surge arrester condition monitoring system, characterized in that, include: Sensing module, expansion module, power supply module, communication module and processing module; The sensing module is used to collect the full current signal, surface temperature signal and partial discharge signal of the surge arrester; The expansion module is used to access ambient humidity signals and surge arrester internal pressure signals through the controller local area network bus interface; The power supply module is used to extract energy from the induced current of the grounding lead of the surge arrester to provide the electrical energy required for system operation. The communication module is used to send monitoring data to a remote operation and maintenance server; The processing module is used to receive signals from various sensors, extract resistive current components using the Fast Fourier Transform algorithm, construct a three-dimensional feature vector containing the fundamental component of resistive current, the number of partial discharge pulses, and the rate of temperature change, and output the fault type identification result through an adaptive network fuzzy inference system classifier optimized by the Bee Algorithm.

2. The self-diagnostic modular surge arrester condition monitoring system according to claim 1, characterized in that, The expression for the three-dimensional feature vector constructed by the processing module is: ； Where F is a three-dimensional feature vector. To represent the fundamental component of resistive current, To represent the fundamental reference value of resistive current, To represent the number of partial discharge pulses, To indicate the maximum permissible number of partial discharges, To represent the amount of temperature change, Indicates the change over time. This represents the threshold for the rate of temperature change.

3. The self-diagnostic modular surge arrester condition monitoring system according to claim 1, characterized in that, The processing module is also used to execute a self-diagnosis mechanism, which includes calculating the characteristic vector imbalance of the three-phase surge arrester. When the characteristic vector imbalance exceeds a preset threshold and continues for a preset duration, a three-phase abnormal alarm signal is generated.

4. The self-diagnostic modular surge arrester condition monitoring system according to claim 3, characterized in that, The processing module is used to compare the relationship between the resistive current growth rate and the temperature change rate. When the resistive current growth rate exceeds a preset ratio and the temperature change rate is lower than a preset value, it determines that the current sensor is faulty and generates a device self-test alarm signal.

5. The self-diagnostic modular surge arrester condition monitoring system according to claim 1, characterized in that, The communication module includes a long-range radio communication unit and a mobile communication technology unit. The long-range radio communication unit is used for short-range anti-interference data transmission, and the mobile communication technology unit is used for remote data backhaul. The communication module supports the International Electrotechnical Commission 61850 standard communication protocol.

6. The self-diagnostic modular surge arrester condition monitoring system according to claim 1, characterized in that, The power supply module includes a micro-current energy harvesting unit and a supercapacitor energy storage unit. The micro-current energy harvesting unit is used to convert the current in the grounding lead into DC power, and the supercapacitor energy storage unit is used to store power and maintain system operation when the main power supply is disconnected.

7. The self-diagnostic modular surge arrester condition monitoring system according to claim 1, characterized in that, The expansion module adopts a plug-and-play structure, and the processing module automatically configures the sampling frequency and alarm threshold according to the type of sensor connected.

8. The self-diagnostic modular surge arrester condition monitoring system according to claim 1, characterized in that, The adaptive network fuzzy inference system classifier optimized by the bee algorithm optimizes the initial parameters and is trained using a training dataset that includes normal operating conditions, internal damp operating conditions, valve plate aging operating conditions, and external contamination operating conditions, so as to output fault type identification results.

9. The self-diagnostic modular surge arrester condition monitoring system according to claim 1, characterized in that, The processing module executes the Fast Fourier Transform algorithm with a transformation length of 4096 points and calculates the resistive current value in conjunction with the reference voltage phase.

10. The self-diagnostic modular surge arrester condition monitoring system according to claim 1, characterized in that, The sensing module includes a zero flux current sensor, an infrared temperature sensor, and a partial discharge sensor. The zero flux current sensor is used to acquire the full current signal, the infrared temperature sensor is used to acquire the surface temperature signal, and the partial discharge sensor is used to acquire the partial discharge signal.

Images