Intelligent post insulator with leakage current self-sensing function and on-line monitoring system thereof

Abstract

The application discloses a smart post insulator with leakage current self-sensing function and an online monitoring system thereof, and relates to the technical field of equipment monitoring. The smart post insulator comprises an insulator body and at least one self-sensing sensitive unit embedded in the insulator body. The sensitive unit adopts a laminated structure and is composed of piezoelectric composite material or electrostrictive material, and generates inverse piezoelectric strain under the action of a polarization electric field generated by the leakage current. A deformation reading module collects strain signals and inverts the leakage current. The online monitoring system comprises a signal acquisition and preprocessing unit, an on-site data collection gateway and a background comprehensive monitoring system. The background system adopts fast Fourier transform to identify the discharge type in the frequency domain, establishes a piezoelectric coupling stress model to predict the remaining life of the insulator, and configures a mixed deep neural network and a federal incremental learning strategy to adapt to different environments. The application has the advantages of being passive, resisting electromagnetic interference, multi-dimensional monitoring and early warning, etc.

Description

Technical Field

[0001] This invention relates to the field of equipment monitoring technology, and more specifically, to an intelligent post insulator with leakage current self-sensing function and its online monitoring system. Background Technology

[0002] Post insulators are indispensable supporting and insulating components in power systems, widely used in substation GIS, switchgear, and high-voltage transmission lines. Their operational reliability directly affects the safety and stability of the power grid. Leakage current is a key characteristic parameter reflecting the degree of surface contamination, moisture status, and insulation degradation trend of the insulator. Online monitoring of leakage current is an important means of achieving early warning of insulator faults.

[0003] Currently, various leakage current detection schemes have been disclosed in existing technologies, but the following prominent problems still exist: First, there is the issue of coupling interference from external sensors. Existing technologies typically require the external installation of TMR sensors, Hall elements, or Rogowski coils on the outside of the insulator. These external sensors not only disrupt the original electric field distribution of the insulator but may also lead to poor contact due to loose installation, resulting in unstable measurement results. Furthermore, external sensors are prone to displacement or damage in harsh environments such as strong winds and vibrations, leading to insufficient long-term operational reliability.

[0004] Second, there is the challenge of power supply. Some solutions attempt to use self-powered technology, utilizing the leakage current to power the main control chip. However, when the insulator is clean, the leakage current is only at the microampere level, and the energy conversion efficiency is usually less than 30%, making it difficult to stably start the acquisition circuit. In cases of severe pollution, although the leakage current increases, it is close to the flashover threshold, and the self-powered solution cannot meet the real-time requirements.

[0005] Third, electromagnetic compatibility is poor. Substations contain strong electromagnetic interference, especially high-frequency ripple and harmonic pollution, which severely interferes with the time-domain waveform of leakage current signals, significantly reducing the accuracy of traditional time-domain amplitude analysis methods. Using metal wires to transmit signals also introduces additional antenna effects, further worsening the signal-to-noise ratio.

[0006] Fourth, the monitoring dimensions are limited. Most existing technologies can only sense the amplitude of leakage current, and cannot obtain deeper state information such as charge accumulation, dielectric loss, and stress fatigue caused by leakage current to the insulator body, making it difficult to quantitatively assess the remaining life of the insulator.

[0007] Fifth, the sensing unit is separate from the insulator body. In existing smart insulator products, the monitoring sensing unit is a component "attached" to the outside of the insulator. The lifespan of the two is not synchronized, and it is difficult to replace the sensing unit after it fails, which leads to a decrease in the reliability of the entire monitoring system.

[0008] In conclusion, there is an urgent need for a new type of intelligent post insulator and its online monitoring system that can fundamentally solve the above problems. Summary of the Invention

[0009] In order to overcome the problems of poor coupling of external sensors, susceptibility to electromagnetic interference, power supply difficulties and limited monitoring dimensions in existing leakage current monitoring technologies, this invention discloses an intelligent post insulator with leakage current self-sensing function and its online monitoring system, which can effectively solve the above-mentioned technical problems.

[0010] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: A smart post insulator with leakage current self-sensing function includes an insulator body and at least one self-sensing sensitive unit embedded inside the insulator body. The insulator body includes a central cylinder, a skirt covering the outside of the central cylinder, and a first fitting and a second fitting respectively disposed at both ends of the insulator body; The self-sensing sensitive unit adopts a stacked structure, including a grounding layer, a sensitive layer, and a high-voltage sensing layer stacked sequentially; the grounding layer is electrically connected to the second fitting, and the high-voltage sensing layer is coupled with the first fitting to form an electric field excitation circuit; the sensitive layer is made of piezoelectric composite material or electrostrictive material, and the sensitive layer is configured such that when the insulator body is subjected to voltage and generates leakage current, the sensitive layer undergoes inverse piezoelectric strain proportional to the instantaneous amplitude of the leakage current under the action of the polarized electric field; The intelligent post insulator also includes a deformation reading module, which is located adjacent to the self-sensing sensitive unit and is used to collect the deformation signal generated by the self-sensing sensitive unit.

[0011] Furthermore, the deformation reading module includes a fiber optic Fabry-Perot cavity micro-displacement sensor; the Fabry-Perot interferometer cavity of the fiber optic Fabry-Perot cavity micro-displacement sensor is fixed below the self-sensing sensitive unit, and the cavity length changes with the deformation of the self-sensing sensitive unit; The smart post insulator also includes a strain gauge bridge reading module as a hot backup. The strain gauge bridge reading module includes a MEMS strain gauge attached to the back of the self-sensing sensitive unit. The MEMS strain gauge forms a Wheatstone half-bridge, and its output signal is used to calibrate the temperature drift of the fiber optic Fabry-Perot cavity micro-displacement sensor.

[0012] Furthermore, the number of self-sensing sensitive units is multiple, and the multiple self-sensing sensitive units are arranged at intervals along the axial direction of the insulator body to form a ring-shaped micro-unit array; the ring-shaped micro-unit array is configured to monitor the distribution changes of leakage current in segments along the axial direction and determine the local flashover initiation position.

[0013] An online monitoring system for intelligent post insulators with self-sensing leakage current includes: At least one of the aforementioned smart post insulators; The signal acquisition and preprocessing unit is electrically or optically connected to the deformation reading module of each of the smart post insulators, and is used to convert and preprocess the deformation signals output by each of the self-sensing sensitive units. The local data aggregation gateway is communicatively connected to the signal acquisition and preprocessing unit, and is used to aggregate the preprocessed deformation signal data and transmit it remotely. The background integrated monitoring system is communicatively connected to the local data aggregation gateway, including a Web cloud integrated monitoring platform and distributed edge computing nodes; The background integrated monitoring system is configured to: invert and calculate the leakage current characteristic value from the received deformation signal data, and establish a state database based on the multi-dimensional characteristics of the leakage current; use fast Fourier transform to convert the time-domain signal of the leakage current to the frequency domain to identify the discharge type; establish a piezoelectric coupling stress model to predict the remaining life of the insulator body and output an early warning signal.

[0014] Furthermore, the signal acquisition and preprocessing unit includes a transimpedance amplifier circuit, a bandpass filter, a power frequency notch filter, and a local FPGA module; the local FPGA module has a built-in finite impulse response digital filter for extracting the fundamental and harmonic components of the leakage current.

[0015] Furthermore, the background integrated monitoring system is configured to divide the 0-2kHz frequency band into three characteristic segments when converting the leakage current time-domain signal to the frequency domain using Fast Fourier Transform: The first segment consists of the 50Hz fundamental frequency and lower harmonics below 250Hz, corresponding to the leakage current dominated by the conductivity of the dirt layer. The threshold for judgment is to issue a wet dirt warning when the amplitude is greater than 5mA. The second segment is 250Hz-1kHz, corresponding to weak partial discharge pulses; The third segment is 1kHz-2kHz, which corresponds to the high-frequency components in the arc pre-ignition process. The threshold for determining the flashover warning is when the amplitude is greater than 2mA.

[0016] Furthermore, the background integrated monitoring system is also equipped with a discharge type identification model. The discharge type identification model adopts a support vector machine or gradient boosting decision tree to automatically identify different fault precursors, including pollution flashover, wet flashover, ice flashover and bird droppings flashover, based on the multi-dimensional characteristics of the leakage current.

[0017] Furthermore, the piezoelectric coupling stress model is as follows: , Where Y represents the stress accumulation factor, Indicates the effective value of leakage current. Indicates the runtime factor. Represents the material constant of the insulating medium; The background integrated monitoring system is configured to monitor the micro-stress drift rate dσ / dt output by each of the self-sensing sensitive units. When the micro-stress drift rate exceeds a preset threshold and the leakage current amplitude does not exceed the standard, it is judged as early hidden degradation and an early warning is issued.

[0018] Furthermore, the background integrated monitoring system is also equipped with a hybrid deep neural network, which adopts a combination architecture of one-dimensional convolutional neural network and long short-term memory network to model leakage current time series data; the background integrated monitoring system is configured to adopt a federated incremental learning strategy to gradually update the local diagnostic model of different geographical locations to adapt to the discharge characteristics under different environmental conditions.

[0019] Furthermore, it also includes a self-powered module, which includes a rectifier circuit and a supercapacitor connected in parallel to the self-sensing sensitive unit; when the leakage current is greater than 3mA, the self-powered module collects and stores the leakage current and supplies power to at least the signal acquisition and preprocessing unit and the local data aggregation gateway.

[0020] Compared with existing technologies, the advantages of this invention are as follows: This invention directly embeds the self-sensing sensitive unit inside the insulator body (such as the interface layer between the FRP center cylinder and the silicone rubber shed), adopting a stacked structure. The sensitive unit itself does not introduce exposed conductors and does not change the creepage distance and electric field distribution of the insulator, realizing an integrated design where the insulator body is the sensor. The grounding layer is connected to the bottom fitting, and the high-voltage sensing layer is coupled to the top fitting, without damaging the original insulation structure, resulting in high reliability in long-term operation. Unlike existing TMR and Hall sensors that require external power supply, the self-sensing sensitive unit of this invention is based on the inverse piezoelectric effect of piezoelectric / electrostrictive materials, directly converting the electric field energy generated by leakage current into mechanical strain, generating measurable deformation signals without any external power supply. Simultaneously, the preferred fiber optic Fabry-Perot cavity micro-displacement sensor uses all-dielectric optical path transmission, is completely non-conductive, and is naturally immune to strong electromagnetic fields, completely solving the signal pollution problem in the harsh electromagnetic environment of substations. The MEMS strain gauge bridge serves as a hot backup and can be used to calibrate temperature drift, further improving measurement accuracy. This invention not only retrieves conventional parameters such as leakage current amplitude and waveform distortion rate through inversion, but also achieves segmented monitoring along the insulator surface using an axial ring-shaped micro-unit array. This allows for precise location of the most heavily polluted sections and the starting point of local flashovers. More importantly, this invention innovatively establishes a piezoelectric coupling stress model. By monitoring the micro-stress drift rate dσ / dt output by the sensitive unit, it can quantitatively assess the aging rate and remaining life of the insulating medium. Even when the leakage current amplitude has not exceeded the limit, once the stress drift abnormally intensifies, the system can identify early latent degradation and issue an early warning, upgrading traditional post-event alarms to pre-event predictions. The integrated monitoring system uses Fast Fourier Transform to convert the leakage current time-domain signal to the frequency domain, dividing the 0-2kHz frequency band into three characteristic segments. Combined with support vector machines or gradient boosting decision tree models, it can automatically identify different fault precursors such as pollution flashover, wet flashover, ice flashover, and bird droppings flashover. Furthermore, based on a hybrid architecture of one-dimensional convolutional neural networks and long short-term memory networks, combined with a federated incremental learning strategy, the system can continuously optimize the local diagnostic model according to different environmental conditions such as salinity, humidity, and pollution type in various locations, exhibiting strong generalization ability and adaptability. Under conditions where the leakage current exceeds 3mA, by connecting a parallel rectifier circuit and a supercapacitor, the leakage current can be collected to provide temporary operating power for the local pre-processing unit and gateway, further reducing the system's dependence on external power sources and making it suitable for deployment on transmission lines in remote areas. In summary, this invention significantly outperforms existing technologies in terms of monitoring accuracy, reliability, intelligence, and service life, providing a novel technical solution for condition-based maintenance of power equipment. Attached Figure Description

[0021] To more clearly illustrate the embodiments of the present invention or the technical solutions in 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 merely exemplary. For those skilled in the art, other embodiments can be derived from the provided drawings without creative effort.

[0022] Figure 1 A structural diagram of the intelligent post insulator online monitoring system provided in this application embodiment; Figure 2 A diagram of a smart post insulator with leakage current self-sensing function provided in an embodiment of this application. Detailed Implementation

[0023] The accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this patent. To better illustrate this embodiment, some parts in the accompanying drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It will be understood by those skilled in the art that certain well-known structures and their descriptions may be omitted in the accompanying drawings.

[0024] The embodiments of this application will be further described in detail below with reference to the accompanying drawings and examples.

[0025] It is understood that the specific embodiments described herein are merely illustrative of the embodiments of this application and are not intended to limit the embodiments of this application. Furthermore, it should be noted that, for ease of description, the accompanying drawings only show the parts related to the embodiments of this application, not all structures. Those skilled in the art, after reading this specification, should be able to realize that any combination of technical features can constitute an optional implementation method, provided that the technical features do not contradict each other.

[0026] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship. In the description of this application, "multiple" means two or more, and "several" means one or more.

[0027] This embodiment provides an intelligent post insulator with leakage current self-sensing function. While maintaining the electrical insulation and mechanical support functions of traditional post insulators, the insulator achieves real-time, in-situ, and highly sensitive monitoring of leakage current by embedding a self-sensing sensitive unit inside the insulator body.

[0028] Please see Figure 2 The insulator body adopts a composite insulator structure, including a central cylinder, sheds covering the outside of the central cylinder, and first and second fittings respectively located at both ends of the insulator body. The central cylinder is constructed of high-strength epoxy resin fiberglass plucked rods, with its outer diameter designed according to voltage level and mechanical load requirements, typically ranging from 40mm to 120mm. The sheds are made of high-temperature vulcanized silicone rubber (HTV) or liquid silicone rubber (LSR) and are injection molded onto the outside of the central cylinder. The shed structure uses alternating large and small sheds to increase creepage distance and suppress rain flashover. The first and second fittings are located at the upper and lower ends of the insulator body, respectively. The first fitting is used for connection to the high-voltage busbar or electrical equipment, and the second fitting is used for grounding installation or fixing to the bracket. Both fittings are connected to the ends of the central cylinder by crimping or adhesive bonding.

[0029] Please see Figure 1 At least one self-sensing sensitive unit is embedded inside the insulator body. The self-sensing sensitive unit adopts a sandwich structure, with a grounding layer, a sensitive layer, and a high-voltage induction layer stacked sequentially from the inside to the outside along the radial direction of the insulator. The grounding layer is made of copper or aluminum foil, with a thickness of 0.1 mm to 0.3 mm, and is electrically connected to a second fitting via conductive adhesive or welding to ensure equipotential connection with ground potential. The high-voltage induction layer is made of a metal thin film or flexible conductive electrodes, forming an electric field excitation circuit with the first fitting through capacitive or electric field coupling. An electric field signal related to the operating voltage can be induced on the high-voltage induction layer without direct electrical connection. The sensitive layer is located between the grounding layer and the high-voltage induction layer and is made of piezoelectric composite material or electrostrictive material. In this embodiment, the piezoelectric composite material is preferably lead zirconate titanate (PZT) ceramic particles and a polymer matrix, such as polyvinylidene fluoride (PVDF) or epoxy resin, forming a 0-3 or 1-3 type piezoelectric composite material with a thickness of 0.5 mm to 2 mm; the electrostrictive material is preferably lanthanum-doped lead zirconate titanate (PLZT) relaxor ferroelectric ceramic, which has a high sensitivity characteristic that is proportional to the square of the electric field.

[0030] When an insulator experiences leakage current on its surface or inside under operating voltage, this leakage current flows through the insulator body and generates a polarized electric field within it. This polarized electric field acts on the sensitive layer, causing mechanical deformation under the inverse piezoelectric effect or electrostriction effect, and the instantaneous amplitude of this deformation is proportional to the instantaneous amplitude of the leakage current. Specifically, the larger the leakage current, the stronger the electric field distortion inside the insulator, the stronger the polarized electric field on the sensitive layer, and the greater the resulting inverse piezoelectric strain. By accurately calibrating the correspondence between the leakage current amplitude and the deformation of the sensitive layer, the indirectly observable leakage current can be converted into a measurable mechanical displacement signal.

[0031] The intelligent post insulator also includes a deformation reading module, which is located adjacent to the self-sensing sensitive unit and is used to collect the deformation signal generated by the self-sensing sensitive unit. In this embodiment, the deformation reading module includes a fiber optic Fabry-Perot cavity micro-displacement sensor. The fiber optic Fabry-Perot cavity micro-displacement sensor is composed of a single-mode fiber, a quartz capillary, and a reflective film. Its Fabry-Perot interference cavity is fixed below or to the side of the self-sensing sensitive unit, and one reflective surface of the interference cavity is correspondingly attached to the back of the self-sensing sensitive unit. When the self-sensing sensitive unit deforms, the deformation is transmitted to the Fabry-Perot interference cavity, causing a change in cavity length, which in turn causes a drift in the interference spectrum. By reading the phase or wavelength change of the interference spectrum using a demodulator, the deformation of the self-sensing sensitive unit can be reconstructed with high precision, and its displacement resolution can reach the nanometer level.

[0032] To improve system reliability, the intelligent post insulator is also equipped with a strain gauge bridge reading module as a hot backup. The strain gauge bridge reading module includes MEMS strain gauges attached to the back of the self-sensing sensitive unit. These MEMS strain gauges are made of silicon-based piezoresistive material using microelectromechanical systems (MEMS) technology, with dimensions on the micrometer scale, and feature low power consumption and high integration. Four MEMS strain gauges form a Wheatstone half-bridge circuit, with two serving as working strain gauges attached to the area of ​​maximum deformation in the sensitive layer, and the other two serving as temperature compensation gauges attached to strain-free areas or structural components made of the same material but not subject to principal strain. The output signal of the Wheatstone half-bridge can not only serve as a backup measurement method in case of fiber optic sensor failure, but can also be used for real-time calibration of the temperature drift of the fiber optic Fabry-Perot cavity micro-displacement sensor. The specific calibration method is as follows: during temperature changes, the temperature-dependent strain measured by the MEMS strain gauges and the total strain measured by the fiber optic sensor are differentially calculated to separate the thermal expansion strain component caused by temperature, thereby performing temperature compensation correction on the output of the fiber optic sensor.

[0033] In a preferred embodiment, multiple self-sensing sensitive units are arranged at intervals along the axial direction of the insulator body, forming a ring-shaped micro-unit array. Each self-sensing sensitive unit is an independent ring or fan-shaped structure, arranged around the central cylinder, with an axial spacing of 5cm to 20cm between units to cover the main leakage current path of the insulator body. The ring-shaped micro-unit array is configured to monitor the distribution changes of leakage current segmentally along the axial direction. Since flashover or partial discharge on the insulator surface often starts at a specific location, by comparing the amplitude and phase of the deformation signal output by the self-sensing sensitive units at each axial location, the section with the largest leakage current anomaly can be located, thereby determining the starting location of the local flashover. For example, when the output signal of the third sensitive unit near the high-voltage end is significantly higher than that of the other units, it can be determined that there is a local electric field concentration or pollution accumulation at that location, providing precise spatial positioning information for targeted maintenance.

[0034] To achieve effective extraction and long-distance transmission of the output signals from the self-sensing sensitive units, this embodiment sets up a signal acquisition and preprocessing unit. This unit is connected to the deformation reading module of each smart post insulator via electrical or optical connection, and is used to convert and preprocess the deformation signals output by each sensing sensitive unit.

[0035] The signal acquisition and preprocessing unit includes a transimpedance amplifier circuit, a bandpass filter, a power frequency notch filter, and a local FPGA module. Since the deformation signal generated by the self-sensing sensing unit under leakage current is extremely weak, the corresponding photocurrent or voltage signal is typically in the nanoamp to microamp range. Therefore, the gain of the transimpedance amplifier circuit is designed to be 10. 6 The voltage is V / A to convert weak current signals into processable voltage signals. The transimpedance amplifier circuit uses a low-noise precision operational amplifier (such as ADA4530-1 or OPA656), and a high-stability metal film resistor is selected for the feedback resistor. A small-capacity compensation capacitor is connected in parallel in the feedback loop to suppress high-frequency self-oscillation and ensure that the circuit operates stably within the target frequency band.

[0036] A bandpass filter is placed after the transimpedance amplifier circuit to filter out high-frequency electromagnetic interference and low-frequency temperature drift noise in the signal. In this embodiment, the bandpass filter adopts a fourth-order Butterworth active filter structure, with the passband initially set to 5Hz to 5kHz to cover the main frequency components of the leakage current signal. The power frequency notch filter adopts a double-T active notch network, with the center frequency precisely tuned to 50Hz (or 60Hz, depending on the power grid frequency), and a notch depth greater than 40dB to effectively suppress strong power frequency electromagnetic interference from the high-voltage transmission environment and prevent the power frequency components from drowning out the weak leakage current harmonic signal.

[0037] The local FPGA module, serving as the core of signal preprocessing, incorporates a finite impulse response (FIR) digital filter with a bandwidth precisely configured from 5Hz to 5kHz. This local FPGA module (such as the Xilinx Artix-7 series or Intel Cyclone series) performs analog-to-digital conversion on the analog signal at a sampling rate of at least 10kS / s, followed by real-time filtering of the digital signal using a parallel FIR filter architecture. This FIR digital filter is designed using the window function method, has an order of at least 128, and exhibits linear phase characteristics. It can accurately extract the fundamental component (50Hz) and various harmonic components (especially 250Hz, 350Hz, 550Hz, etc.) of the leakage current, while suppressing out-of-band noise. After FPGA preprocessing, the original waveform data and characteristic parameters (such as fundamental amplitude, harmonic content, pulse count, etc.) are encapsulated into standard data frames and transmitted to the next-level device via SPI or Ethernet interface.

[0038] For branches employing fiber optic Fabry-Perot cavity micro-displacement sensors, the signal acquisition and preprocessing unit is also equipped with a fiber optic demodulation module. This module uses white light interferometry demodulation or tunable laser scanning demodulation to convert the optical signal into an electrical signal, which is then fed into the transimpedance amplifier circuit and subsequent processing links. Fiber optic channels possess inherent electromagnetic interference immunity, making them particularly suitable for strong electromagnetic environments such as substations.

[0039] The deformation signal data, after on-site preprocessing, is remotely transmitted through a local data aggregation gateway. This local data aggregation gateway is an industrial-grade IoT gateway, supporting multiple transmission methods such as 4G / 5G wireless communication, fiber optic Ethernet, or power line carrier communication, and features data caching, breakpoint resumption, and protocol conversion. The gateway aggregates, compresses, and encrypts data streams from multiple smart post insulators before uploading them to the backend integrated monitoring system.

[0040] The integrated backend monitoring system includes a Web-cloud integrated monitoring platform and distributed edge computing nodes. The Web-cloud integrated monitoring platform is deployed on a cloud server or in the power company's data center, providing human-computer interaction functions such as data storage, visualization, historical query, and remote configuration; the distributed edge computing nodes are deployed in substations or along transmission lines, undertaking some real-time computing tasks to reduce transmission latency and cloud load.

[0041] One of the core functions of the integrated monitoring system is to invert and calculate the leakage current characteristic value from the received deformation signal data. Since there is a calibrated linear or quadratic correspondence between the deformation of the self-sensing sensitive unit and the leakage current amplitude, the system first converts the deformation signal into an equivalent leakage current time-domain waveform based on the factory calibration coefficients of each insulator. Subsequently, the system establishes a state database based on the multi-dimensional characteristics of the leakage current. This database stores multi-dimensional feature vectors such as the effective value, peak value, harmonic spectrum, pulse frequency, and phase distribution of the leakage current at each monitoring point, and is indexed by time series to support long-term trend analysis and horizontal comparison.

[0042] In terms of frequency domain analysis, the integrated monitoring system uses Fast Fourier Transform (FFT) to convert the leakage current time-domain signal to the frequency domain, and divides the 0 to 2kHz frequency band into three characteristic segments for targeted identification: The first band comprises the 50Hz fundamental frequency and lower harmonics below 250Hz (mainly the 3rd and 5th harmonics). This band corresponds to leakage current dominated by contamination conductivity. When the insulator surface is contaminated and damp, surface conductivity increases, and the leakage current is dominated by the power frequency and its lower harmonics. The system sets the threshold for this band to be an amplitude greater than 5mA. When the energy proportion of this band exceeds the preset ratio and the amplitude is greater than 5mA, a wet contamination warning is issued, prompting maintenance personnel to clean or apply anti-flashover coating.

[0043] The second band, from 250 Hz to 1 kHz, corresponds to weak partial discharge pulses. The pulse current generated by partial discharge has a broad spectrum, exhibiting either discrete spectral lines or a continuous noise floor within the 250 Hz to 1 kHz range. The system assesses the development level of partial discharge by calculating the number of spectral peaks and the cumulative energy value in this band.

[0044] The third band is from 1kHz to 2kHz, corresponding to the high-frequency components during the arc pre-ignition process. Before a flashover occurs, a pre-discharge arc will appear on the insulator surface, generating abundant high-frequency harmonics. The system sets the judgment threshold for this band to be an amplitude greater than 2mA. When the amplitude of the high-frequency components exceeds this threshold, a flashover warning is issued, providing a basis for decision-making regarding emergency shutdown or voltage reduction.

[0045] To further improve the accuracy of fault identification, the integrated monitoring system is also equipped with a discharge type identification model. This model employs a Support Vector Machine (SVM) or Gradient Boosting Decision Tree (GBDT) algorithm. During model training, the system collects a large number of historical samples, including multi-dimensional feature vectors of leakage current corresponding to different fault precursors such as pollution flashover, wet flashover, ice flashover, and bird droppings flashover. Features such as frequency domain energy distribution, time domain pulse shape, and phase-resolved spectrum are extracted as input, and supervised learning is performed using the fault type as a label. During online diagnosis, the model automatically identifies the type of fault risk currently faced by the insulator based on the real-time extracted multi-dimensional features of leakage current. For example, pollution flashover typically manifests as a slow increase in fundamental amplitude accompanied by an increase in low-order harmonics; wet flashover is characterized by a sudden increase in leakage current during rainfall and a broad spectrum; ice flashover is characterized by intermittent pulses accompanied by a DC component; and bird droppings flashover is characterized by sudden, large-amplitude pulses and random phase distribution. By accurately identifying the discharge type, the system can provide a basis for differentiated operation and maintenance strategies.

[0046] Regarding the assessment of the remaining life of insulators, the integrated monitoring system establishes a piezoelectric coupling stress model, the expression of which is: Wherein, Y represents the stress accumulation factor, which reflects the cumulative electromechanical stress of the insulating medium under the action of long-term leakage current; This represents the effective value of the leakage current, obtained through real-time monitoring. This represents the operating time factor, which is related to the cumulative operating hours of the insulator and the number of environmental temperature and humidity cycles. This represents the material constant of the insulating medium, determined by the dielectric properties, tracking resistance, and mechanical fatigue characteristics of the center tube and skirt materials, and is calibrated at the factory. The system continuously calculates the stress accumulation factor Y at each monitoring point and fits the growth curve of the Y value using historical data. When the Y value approaches the material's critical failure threshold, the system outputs a warning about the remaining life.

[0047] Furthermore, the system is configured to monitor the micro-stress drift rate dσ / dt output by each sensing unit. In certain early stages of degradation, such as the appearance of microcracks or interface debonding within the insulator, the leakage current amplitude may not yet be significantly exceeded, but the local stress field in the sensitive layer will slowly change, manifesting as micro-stress drift. When the system detects that the micro-stress drift rate exceeds a preset threshold (e.g., 0.1 με / day), while the leakage current amplitude remains within the normal range, it determines this as early-stage latent degradation and issues an early warning. This early warning mechanism based on micro-stress drift can capture latent defects that traditional current threshold methods cannot detect, enabling proactive condition-based maintenance.

[0048] To adapt to complex and ever-changing operating environments and continuously improve diagnostic accuracy, the integrated monitoring system is also equipped with a hybrid deep neural network. This hybrid deep neural network employs a combined architecture of a one-dimensional convolutional neural network (1D-CNN) and a long short-term memory network (LSTM). The 1D-CNN is used to automatically extract local features from leakage current time-series data, such as pulse edges and harmonic envelopes; the LSTM is used to model the long-term temporal dependencies of the leakage current and capture the evolutionary trend of discharge development. The network input is a fixed-length leakage current waveform segment or a multi-dimensional feature sequence, and the output is a fault probability distribution or a predicted remaining lifetime.

[0049] In terms of model update strategy, the back-end integrated monitoring system adopts a federated incremental learning strategy. Because the environmental conditions (such as pollution levels, altitude, and climate type) of insulators in different geographical locations vary significantly, a single global model is insufficient to adapt to all scenarios. The system deploys local diagnostic models at various local monitoring sites (such as different substations). Each local model is trained based on local data, and only model parameter updates (not the original data) are uploaded to the cloud aggregation server. The cloud server uses a federated averaging (FedAvg) algorithm to weight and aggregate the parameters of each local model, generating a globally optimized model, which is then distributed to each local node. As operational data accumulates, the local models gradually adapt to local environmental characteristics through incremental learning, while the global model integrates knowledge from multiple regions, achieving continuous evolution of diagnostic capabilities. This strategy effectively solves the problem of adaptability to discharge characteristics under different environmental conditions while protecting data privacy.

[0050] To ensure the long-term autonomous operation of the intelligent post insulator and its monitoring system in remote areas or scenarios where external power is inconvenient, this embodiment also includes a self-powered module. The self-powered module comprises a rectifier circuit and a supercapacitor connected in parallel to the self-sensing sensing unit. The rectifier circuit uses a bridge rectifier constructed with low forward voltage drop Schottky diodes to convert the alternating micro-current or induced charge generated by the self-sensing sensing unit under leakage current into direct current. The supercapacitor, as the energy storage element, is an electrically double-layer capacitor (EDLC) with a capacitance range of 1F to 10F and a rated voltage of 5V to 16V, featuring a long charge-discharge cycle life (up to hundreds of thousands of times) and good adaptability to a wide temperature range.

[0051] The self-powered module operates as follows: When the insulator is running normally and the leakage current is greater than 3mA, the self-sensing sensing unit generates continuous charge migration or induced current under the action of the polarization electric field. The rectifier circuit collects and stores this energy in the supercapacitor. The supercapacitor's energy storage prioritizes powering the signal acquisition and preprocessing unit and the local data aggregation gateway, ensuring the continuous operation of front-end sensing, amplification, filtering, analog-to-digital conversion, and data transmission functions without an external power supply. When the leakage current is less than 3mA (e.g., in sunny and dry weather), the supercapacitor relies on its previously stored energy to maintain the system in a low-power standby state, with standby power consumption controlled below the milliwatt level. In extreme cases, if the supercapacitor voltage falls below the minimum operating threshold, the system automatically enters sleep mode, retaining only the timed wake-up sampling function, and fully restarts after the leakage current recovers.

[0052] The overall system operation logic can be summarized as follows: Under the action of operating voltage, the leakage current of the insulator body causes the internal self-sensing sensitive unit to generate inverse piezoelectric strain proportional to the current amplitude; the deformation reading module (mainly fiber optic Fabry-Perot cavity micro-displacement sensor, with MEMS strain gauge bridge as hot backup) converts this deformation into an electrical signal or optical signal; after on-site acquisition, amplification, filtering, and FPGA preprocessing, the signal is uploaded to the backend through the gateway; the backend system, based on frequency domain analysis, machine learning models, and piezoelectric coupling stress models, realizes discharge type identification, fault early warning, and life assessment; simultaneously, the system utilizes the leakage current itself to achieve energy harvesting and self-powering, forming a complete closed loop of sensing, transmission, diagnosis, and power supply. This intelligent post insulator and its online monitoring system, through integrated structural and sensing design, upgrades traditional passive insulators into active intelligent devices, providing a reliable technical means for the status perception and intelligent operation and maintenance of power grid equipment.

[0053] The above implementation details the structural design, sensing unit working mechanism, signal processing flow, intelligent diagnostic algorithm, and self-powered scheme of the intelligent post insulator. The various technical features are interconnected and work together to form a complete leakage current self-sensing and online monitoring technology system.

[0054] The same or similar labels correspond to the same or similar parts; The terms used to describe positional relationships in the accompanying drawings are for illustrative purposes only and should not be construed as limiting this patent. Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. For those skilled in the art, other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all implementation methods here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the claims of the present invention.

Claims

1. An intelligent post insulator with leakage current self-sensing function, characterized in that, It includes an insulator body and at least one self-sensing sensitive unit embedded inside the insulator body; The insulator body includes a central cylinder, a skirt covering the outside of the central cylinder, and a first fitting and a second fitting respectively disposed at both ends of the insulator body; The self-sensing sensitive unit adopts a stacked structure, including a grounding layer, a sensitive layer, and a high-voltage sensing layer stacked sequentially; the grounding layer is electrically connected to the second fitting, and the high-voltage sensing layer is coupled with the first fitting to form an electric field excitation circuit; the sensitive layer is made of piezoelectric composite material or electrostrictive material, and the sensitive layer is configured such that when the insulator body is subjected to voltage and generates leakage current, the sensitive layer undergoes inverse piezoelectric strain under the action of a polarized electric field, which is proportional to the instantaneous amplitude of the leakage current; The intelligent post insulator also includes a deformation reading module, which is located adjacent to the self-sensing sensitive unit and is used to collect the deformation signal generated by the self-sensing sensitive unit.

2. The intelligent post insulator with leakage current self-sensing function according to claim 1, characterized in that, The deformation reading module includes a fiber optic Fabry-Perot cavity micro-displacement sensor; the Fabry-Perot interferometer cavity of the fiber optic Fabry-Perot cavity micro-displacement sensor is fixed below the self-sensing sensitive unit, and the cavity length changes with the deformation of the self-sensing sensitive unit; The smart post insulator also includes a strain gauge bridge reading module as a hot backup. The strain gauge bridge reading module includes a MEMS strain gauge attached to the back of the self-sensing sensitive unit. The MEMS strain gauge forms a Wheatstone half-bridge, and its output signal is used to calibrate the temperature drift of the fiber optic Fabry-Perot cavity micro-displacement sensor.

3. The intelligent post insulator with leakage current self-sensing function according to claim 1, characterized in that, The number of self-sensing sensitive units is multiple, and the multiple self-sensing sensitive units are arranged at intervals along the axial direction of the insulator body to form a ring-shaped micro-unit array; the ring-shaped micro-unit array is configured to monitor the distribution changes of leakage current in segments along the axial direction and determine the local flashover initiation position.

4. An intelligent online monitoring system for post insulators with self-sensing leakage current function, characterized in that, include: At least one smart post insulator according to any one of claims 1 to 3; The signal acquisition and preprocessing unit is electrically or optically connected to the deformation reading module of each of the smart post insulators, and is used to convert and preprocess the deformation signals output by each of the self-sensing sensitive units. The local data aggregation gateway is communicatively connected to the signal acquisition and preprocessing unit, and is used to aggregate the preprocessed deformation signal data and transmit it remotely. The background integrated monitoring system is communicatively connected to the local data aggregation gateway, including a Web cloud integrated monitoring platform and distributed edge computing nodes; The background integrated monitoring system is configured to: invert and calculate the leakage current characteristic value from the received deformation signal data, and establish a state database based on the multi-dimensional characteristics of the leakage current; use fast Fourier transform to convert the time-domain signal of the leakage current to the frequency domain to identify the discharge type; establish a piezoelectric coupling stress model to predict the remaining life of the insulator body and output an early warning signal.

5. The intelligent post insulator online monitoring system with leakage current self-sensing function according to claim 4, characterized in that, The signal acquisition and preprocessing unit includes a transimpedance amplifier circuit, a bandpass filter, a power frequency notch filter, and a local FPGA module; the local FPGA module has a built-in finite impulse response digital filter for extracting the fundamental and harmonic components of the leakage current.

6. The intelligent post insulator online monitoring system with leakage current self-sensing function according to claim 4, characterized in that, The aforementioned integrated monitoring system is configured to divide the 0-2kHz frequency band into three characteristic segments when converting the leakage current time-domain signal to the frequency domain using Fast Fourier Transform: The first segment consists of the 50Hz fundamental frequency and lower harmonics below 250Hz, corresponding to the leakage current dominated by the conductivity of the dirt layer. The threshold for judgment is to issue a wet dirt warning when the amplitude is greater than 5mA. The second segment is 250Hz-1kHz, corresponding to weak partial discharge pulses; The third segment is 1kHz-2kHz, which corresponds to the high-frequency components in the arc pre-ignition process. The threshold for determining the flashover warning is when the amplitude is greater than 2mA.

7. The intelligent post insulator online monitoring system with leakage current self-sensing function according to claim 4, characterized in that, The background integrated monitoring system is also equipped with a discharge type identification model. The discharge type identification model adopts support vector machine or gradient boosting decision tree to automatically identify different fault precursors, including pollution flashover, wet flashover, ice flashover and bird droppings flashover, based on the multi-dimensional characteristics of the leakage current.

8. The intelligent post insulator online monitoring system with leakage current self-sensing function according to claim 4, characterized in that, The piezoelectric coupling stress model is as follows: , Where Y represents the stress accumulation factor, Indicates the effective value of leakage current. Indicates the runtime factor. Represents the material constant of the insulating medium; The background integrated monitoring system is configured to monitor the micro-stress drift rate dσ / dt output by each of the self-sensing sensitive units. When the micro-stress drift rate exceeds a preset threshold and the leakage current amplitude does not exceed the standard, it is judged as early hidden degradation and an early warning is issued.

9. The intelligent post insulator online monitoring system with leakage current self-sensing function according to claim 4, characterized in that, The background integrated monitoring system is also equipped with a hybrid deep neural network, which adopts a combination architecture of one-dimensional convolutional neural network and long short-term memory network to model leakage current time series data. The background integrated monitoring system is configured to adopt a federated incremental learning strategy to gradually update the local diagnostic model in different geographical locations to adapt to the discharge characteristics under different environmental conditions.

10. The intelligent post insulator online monitoring system with leakage current self-sensing function according to any one of claims 4 to 9, characterized in that, It also includes a self-powered module, which includes a rectifier circuit and a supercapacitor connected in parallel to the self-sensing sensitive unit; when the leakage current is greater than 3mA, the self-powered module collects and stores the leakage current and supplies power to at least the signal acquisition and preprocessing unit and the local data aggregation gateway.

Images