Transformer iron core grounding current detection device and detection method thereof

Abstract

The invention relates to the technical field of grounding current detection devices, in particular to a transformer iron core grounding current detection device and a detection method thereof.The transformer iron core grounding current detection device comprises a detection unit, a signal processing unit, a self-checking calibration unit, a control analysis unit, a data transmission unit, a power supply unit and a display alarm unit; all the units are integrally packaged, and an integrated metal shielding structure which is grounded is adopted to restrain electromagnetic interference. According to the invention, a dual-sensor fusion acquisition structure is adopted to cooperate with a multi-stage signal processing design, so that electromagnetic interference can be effectively suppressed, the accuracy and integrity of iron core grounding current signal acquisition are ensured, and the influence of signal distortion on a detection result is avoided; the self-checking calibration unit can realize real-time self-checking and regular automatic calibration of each module without frequent manual intervention, module faults can be found in time and accurately positioned, the manual maintenance workload is reduced, and meanwhile, the detection precision of long-term operation of the device is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of current detection device technology, and in particular to a transformer core grounding current detection device and its detection method. Background Technology

[0002] Transformers are core equipment for energy conversion and transmission in power systems, and their operational reliability directly determines the safe and stable operation of the power grid. The transformer core is made of stacked silicon steel sheets, with insulating layers between them. During normal operation, the core needs to be grounded at a single point through a grounding lead to eliminate any floating potential between the core and ground, preventing excessively high floating potentials from breaking down the core insulation and causing damage. During long-term operation, factors such as vibration, insulation aging, oil deterioration, and foreign object intrusion can easily damage the insulation between the steel sheets, leading to multi-point grounding faults in the core. When the core is grounded at multiple points, a large grounding current is generated in the grounding lead. This current flows through the core, forming a closed loop, causing localized overheating of the core, accelerating insulation aging, and in severe cases, burning out the core and windings, causing transformer shutdown, and triggering power grid accidents. Therefore, real-time and accurate detection of the transformer core grounding current to promptly detect multi-point grounding faults is of significant practical importance.

[0003] Currently, existing transformer core grounding current detection devices have the following drawbacks: low detection accuracy, mostly using a single sensor to collect grounding current signals, which is affected by complex electromagnetic interference and temperature fluctuations in substations, resulting in large detection errors, and inability to capture weak grounding current signals (<10mA), making it difficult to detect early, minor multi-point grounding faults; poor fault judgment accuracy, using fixed thresholds for fault judgment without dynamically adjusting according to transformer operating conditions (ambient temperature, grid frequency, load), which easily leads to false alarms and missed alarms, and makes it impossible to distinguish the severity of the fault. Summary of the Invention

[0004] In view of the technical problems of existing transformer core grounding current detection devices mentioned in the background art, the present invention provides a transformer core grounding current detection device and its detection method.

[0005] The technical solution adopted in this invention is as follows: a transformer core grounding current detection device and its detection method. The grounding current detection device includes a detection unit, a signal processing unit, a self-test calibration unit, a control and analysis unit, a data transmission unit, a power supply unit, and a display and alarm unit. Each unit is integrated and packaged with an integrated metal shielding structure for grounding treatment to suppress electromagnetic interference. The detection unit adopts a dual-sensor fusion structure of Rogowski coil and Hall sensor, and combines an adaptive weighted fusion algorithm to achieve full-range acquisition of core grounding current. The signal processing unit adopts a multi-level processing structure of adaptive filtering, programmable gain amplification, isolation conversion, and signal conditioning to eliminate interference and complete signal conversion. The self-test calibration unit integrates a built-in standard signal source to realize real-time self-testing, periodic automatic calibration, and fault location of each module. The control and analysis unit adopts a dual-core architecture and integrates a dynamic threshold correction algorithm to distinguish different operating conditions and fault severity of core grounding. The data transmission unit adopts a dual-mode anti-interference transmission method and combines data encryption and breakpoint resume technology to achieve stable data transmission. The power supply unit adopts a dual-mode power supply structure and integrates power protection and low-power management functions. The display and alarm unit combines display, audible and visual alarms, and local operation functions to realize status feedback and local operation and maintenance. In one embodiment, the Rogowski coil adopts a hollow toroidal structure with 1000-2000 turns, an inner diameter of 50-80mm, and is wound with fine-diameter enameled copper wire; the Hall sensor is a closed-loop Hall current sensor with a range of 0-10A, a detection accuracy of not less than ±0.1%FS, and a response time of not more than 1μs; the adaptive weighted fusion algorithm dynamically adjusts the weighting coefficients of the two signals by calculating the noise variance of the two sensor acquisition signals in real time, thereby realizing the acquisition of the full-range grounding current within the range of 1mA-10A. In one embodiment, the signal processing unit includes a signal conditioning module, an adaptive filtering module, a programmable gain amplifier, and an isolation conversion module. The signal conditioning module uses high-precision resistors to achieve current-to-voltage signal conversion and, together with a high-input-impedance operational amplifier, forms a voltage follower to achieve signal impedance matching and distortion-free transmission. The adaptive filtering module uses at least a 20th-order active filter structure, and the filtering bandwidth can be adaptively adjusted within the range of 10Hz to 1MHz to suppress power grid harmonics and electromagnetic interference. The amplification factor of the programmable gain amplifier can be adjusted within the range of 1 to 1000 times to adapt to signals of different amplitudes. The isolation conversion module uses at least a 16-bit high-precision AD converter, and the sampling frequency can be adjusted within the range of 100Hz to 10kHz, with a conversion accuracy of not less than ±0.002%FS.In one embodiment, the built-in standard signal source of the self-test calibration unit can generate a standard current signal in the range of 0~10A and 50Hz~1kHz, with a signal accuracy of not less than ±0.01%FS; it can quickly switch between normal detection and self-test calibration modes, with a switching response time of no more than 10ms; during the calibration process, the calibration coefficient is obtained by calculating the ratio of the standard signal value to the actual detection value of the device, and the detection data is corrected using the calibration coefficient to ensure that the corrected detection accuracy is maintained within ±0.2%FS, with a calibration cycle of 1 day to 1 month. In one embodiment, the dual-core architecture of the control and analysis unit consists of a main control chip and a dedicated data processing chip, wherein the main control chip is a high-performance 32-bit microcontroller with a main frequency of not less than 160MHz, and the dedicated data processing chip is a high-speed digital signal processing chip with a computing speed of not less than 500MIPS; the dynamic threshold correction algorithm combines three key operating condition parameters—ambient temperature, transformer load, and power grid frequency—and adjusts the fault judgment threshold in real time through a preset correction formula. In one embodiment, the dual-mode anti-interference transmission method of the data transmission unit adopts a combination of high-speed wireless communication and long-distance wireless communication to achieve real-time data transmission and long-distance backup transmission. Data encryption technology uses hardware encryption, coupled with a data verification mechanism to ensure the security and integrity of transmitted data. Dual-mode switching is automatically triggered based on the signal strength of the main communication mode, with a switching delay of no more than 100ms. Breakpoint resume technology can buffer uncompleted data and resume transmission from the point of interruption after transmission is restored. In another embodiment, the dual-mode power supply structure of the power supply unit adopts a combination of substation DC power supply and solar power supply. The DC power supply input voltage range is adapted to the conventional DC power supply specifications of the substation, while the solar power supply section, in conjunction with a solar charging management module and an energy storage module, achieves energy storage and stable power supply. Power protection functions include four layers of protection: overcurrent, overvoltage, short circuit, and reverse connection, effectively protecting each module from damage caused by abnormal power supply. Low-power management functions can automatically adjust the power supply according to the device's operating status, reducing the device power consumption to below 3W in sleep mode. In one embodiment, the display alarm unit includes an audible and visual alarm module and a display module. The display module uses an industrial-grade LCD screen to display detection data, working status, and fault information. The audible and visual alarm module distinguishes different alarm modes according to the severity of the fault, providing an intuitive indication of the fault level. The local operation function is implemented through dedicated buttons, which can complete parameter debugging, fault reset, and self-test triggering of local operation and maintenance.

[0006] In one embodiment, the specific steps include: After the device is powered on and initialized, it first completes a full module self-test. If there are no faults, it enters the testing state. The iron core grounding current signal is collected by dual sensor fusion, processed by the signal processing unit, and then transmitted to the control and analysis unit. The control analysis unit analyzes and calculates the signal, and determines the grounding condition and fault level by combining dynamic thresholds; Simultaneously complete the encrypted transmission of detection data and fault information, and provide real-time status feedback; Regular automatic calibration is performed to continuously and cyclically achieve real-time detection of grounding current and fault early warning.

[0007] The beneficial effects of this invention are as follows: Compared with the prior art, the dual-sensor fusion acquisition structure combined with multi-level signal processing design in this invention can effectively suppress electromagnetic interference, ensure the accuracy and integrity of the core grounding current signal acquisition, and avoid signal distortion affecting the detection results; the self-testing and calibration unit can realize real-time self-testing and periodic automatic calibration of each module without frequent manual intervention, which can promptly detect module faults and accurately locate them, reducing the workload of manual maintenance, while ensuring the detection accuracy of the device during long-term operation; the dual-core architecture of the control and analysis unit combined with dynamic threshold correction design can quickly process detection data, accurately distinguish different working conditions and fault severity of core grounding, effectively avoid false alarms and missed alarms, and provide reliable fault judgment basis for operation and maintenance personnel; the data transmission unit adopts dual-mode anti-interference transmission combined with data encryption and breakpoint resume technology to ensure the stability, security and integrity of detection data and fault information transmission, adapt to different transmission distance requirements of substations, and facilitate remote monitoring by operation and maintenance personnel. The dual-mode power supply structure of the power supply unit, combined with comprehensive power protection and low-power management functions, can adapt to different power supply scenarios in substations, ensuring continuous and stable operation of the device while reducing operating energy consumption and maintenance costs. The display and alarm unit, combining display, audible and visual alarms, and local operation functions, can provide real-time feedback on the device's operating status and detection information, intuitively indicating the fault level, facilitating maintenance personnel to quickly detect, troubleshoot, and handle faults, and improving maintenance efficiency. The device as a whole adopts integrated metal shielding encapsulation and grounding treatment, further enhancing anti-interference capabilities. Its compact structure and convenient installation make it suitable for harsh outdoor operating conditions in substations. The operation process is simple and easy to understand, requiring no professional technicians to complete daily operations and basic maintenance. Overall, it improves the intelligence and automation level of transformer core grounding current detection, enabling real-time and accurate monitoring of the transformer core grounding status, timely warning of potential faults, prevention of fault expansion, ensuring safe and stable operation of the transformer, extending the transformer's service life, and reducing substation maintenance costs and safety risks. Attached Figure Description

[0008] Figure 1 This is a system block diagram of the detection device of the present invention; Figure 2 This is a side view structural diagram of the present invention.

[0009] Reference numerals: 1. Grounding current detection device; 2. Solar panel. Detailed Implementation

[0010] In the description of this invention, it should be noted that the terms "front", "up", "down", "left", "right", "vertical", "horizontal", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0011] refer to Figures 1-2 To address the problems existing in the background technology, this application proposes the following technical solution: a transformer core grounding current detection device and its detection method. The grounding current detection device includes a detection unit, a signal processing unit, a self-test calibration unit, a control analysis unit, a data transmission unit, a power supply unit, and a display and alarm unit. Each unit is integrated and packaged, and adopts an integrated metal shielding structure with grounding treatment to suppress electromagnetic interference. The detection unit adopts a dual-sensor fusion structure of Rogowski coil and Hall sensor, and combines an adaptive weighted fusion algorithm to realize full-range acquisition of core grounding current. The signal processing unit adopts adaptive filtering, programmable gain amplification, and isolation... The system features a multi-level processing structure for signal conversion and conditioning to eliminate interference and complete signal conversion; a self-test calibration unit integrates a built-in standard signal source to enable real-time self-testing, periodic automatic calibration, and fault location for each module; a control and analysis unit adopts a dual-core architecture and integrates a dynamic threshold correction algorithm to distinguish different operating conditions and fault severity of the iron core grounding; a data transmission unit uses a dual-mode anti-interference transmission method and combines data encryption and breakpoint resume technology to achieve stable data transmission; a power supply unit adopts a dual-mode power supply structure and integrates power protection and low-power management functions; and a display and alarm unit combines display, audible and visual alarms, and local operation functions to achieve status feedback and local operation and maintenance.

[0012] The units are connected by shielded cables and enclosed in an integrated metal shielded shell with grounding to effectively suppress electromagnetic interference. The device has a compact structure, is easy to install, and meets the actual installation requirements of substations. The shell is made of cold-rolled steel plate with a thickness of 1.5mm and has an anti-corrosion coating, making it suitable for the harsh outdoor environment of substations, which is humid and dusty. The overall dimensions of the device are designed to be 300mm×200mm×150mm, which facilitates installation on a bracket near the transformer core grounding lead without taking up too much space.

[0013] The core functions and connections of each module are as follows: The detection unit is used to collect the transformer core grounding current signal and output a weak analog current signal; the signal processing unit is used to filter, amplify, isolate, and perform AD conversion on the weak signal collected by the detection unit, converting the analog signal into a digital signal and eliminating the influence of interference signals; the self-test calibration unit is used to perform real-time self-tests on each module of the device, promptly prompting for faults, and completing the detection accuracy calibration through the built-in standard signal source; the control analysis unit is used to analyze and calculate the processed digital signal, including the effective value of the current, harmonic components, waveform distortion rate, etc., and determine the core grounding status and fault type in combination with dynamic thresholds; the data transmission unit is used to transmit the detection data, fault information, and self-test results to the remote monitoring platform, and simultaneously receive remote control commands; the power supply unit is used to provide stable power supply to each module of the device and supports switching between multiple power supply modes; the display and alarm unit is used to display the detection data, self-test results, and fault information in real time, and issue an audible and visual alarm when an abnormality is detected. The shielded cables between all units are all RVVP type shielded cables with a cross-sectional area of ​​1.0mm². The cable length is cut according to the on-site installation distance, with a maximum length of 5m to avoid signal attenuation. Both ends of the cable are connected with aviation plugs to ensure a firm connection and good contact.

[0014] Unlike the distributed structural design of existing technologies, this design effectively improves the device's anti-interference capability and stability. Furthermore, it deeply integrates the self-test calibration unit with the core detection module, rather than simply adding a self-test module, enabling simultaneous self-testing and detection. This solves the problems of incomplete self-testing functions and cumbersome calibration in existing technologies. The modular design facilitates later maintenance; when a module fails, it can be disassembled and replaced individually without replacing the entire device, reducing maintenance costs. The integrated shielding design comprehensively suppresses external electromagnetic interference, improving anti-interference effectiveness by more than 30% compared to existing distributed shielding.

[0015] Electromagnetic interference suppression theory states that transformer sites experience significant electromagnetic interference (such as winding leakage flux, switch operation interference, and power grid harmonic interference). This interference can affect the accuracy of detection signals through spatial coupling and cable conduction. This device employs a metal shielded enclosure with a uniform grounding resistance requirement of ≤4Ω, effectively shielding external electromagnetic interference. Furthermore, shielded cables connect all modules, with each cable shielding layer using single-end grounding to reduce conducted interference. Additionally, the signal processing unit utilizes signal isolation technology to achieve electrical isolation between input and output signals, further suppressing interference signal transmission and ensuring the stability and accuracy of the detection signal. During on-site installation, the electromagnetic interference intensity of the installation area must be tested using an EMI-3000 electromagnetic interference tester to ensure it is ≤40dBμV / m. If this range is exceeded, the device's installation location must be adjusted to be away from transformer windings, switches, and other strong interference sources.

[0016] To ensure effective grounding of the casing, the grounding electrode parameters must be precisely designed using the grounding resistance calculation formula. The specific formula is as follows: This formula is the core basis for the grounding design of the device's casing. Through calculation, parameters such as the size and burial depth of the grounding electrode can be determined, ensuring a grounding resistance ≤4Ω, achieving effective electromagnetic interference shielding, and preventing external electromagnetic interference from affecting the device's detection accuracy. Among these... ρ is the grounding resistance of the casing (Ω); ρ is the soil resistivity at the substation site (Ω·m), which needs to be determined by actual on-site measurement using a soil resistivity tester (model ZD-8). The measurement depth is uniformly 1.5m, with no less than 3 measurement points, and the average value is taken as the final value, usually ranging from 10 to 1000 Ω·m; D is the burial depth of the grounding electrode (m), which is usually taken as 0.8 to 1.2m, depending on the on-site installation space. If the soil resistivity is high, the burial depth can be appropriately increased; d is the diameter of the grounding electrode (m), which uniformly uses galvanized round steel of φ16 to φ20mm as the grounding electrode, with a zinc coating thickness ≥80μm to prevent corrosion, and d is taken as 0.016 to 0.02m; h is the length of the grounding electrode (m), which is usually taken as 1.5 to 2.0m.

[0017] Implementation instructions: During on-site installation, first use a soil resistivity tester (model ZD-8) to measure the soil resistivity ρ of the substation, and then substitute it into the above formula to calculate the grounding resistance. ,like If the resistance is greater than 4Ω, then increase the length h of the grounding electrode or use multiple grounding electrodes connected in parallel. The formula for calculating the grounding resistance after parallel connection is as follows: (n is the number of grounding electrodes, usually 2 to 3 are selected), to ensure ≤4Ω; The device casing is reliably connected to the grounding electrode via a grounding wire. The grounding wire is uniformly selected as a copper core cable with a cross-sectional area ≥2.5mm² and an insulation layer thickness ≥1.0mm. The connection part adopts a crimping process, and the crimping tool is uniformly selected as a hydraulic crimping pliers (model YQK-70). The crimping pressure is uniformly 50MPa to ensure good contact and avoid excessive contact resistance affecting the shielding effect. After the grounding electrode is buried underground, the surrounding area is filled with fine sand, and then covered with soil and compacted. The burial position is ≥1.5m away from the transformer foundation to avoid affecting the transformer's own grounding system.

[0018] The detection unit is the core acquisition module of the device, used to acquire the transformer core grounding current signal. Its performance directly determines the detection accuracy of the device. In the existing technology, the detection unit mostly uses a single Rogowski coil or Hall sensor. The Rogowski coil has low accuracy when detecting low-frequency signals (such as 50Hz fundamental wave), with an error ≥5%. The Hall sensor is easily affected by temperature. For every 10°C change in temperature, the detection error increases by 0.5%, resulting in poor detection stability. Furthermore, it cannot capture weak grounding current signals (<10mA).

[0019] This invention employs a dual-sensor fusion acquisition scheme using a Rogowski coil and a Hall sensor, combined with an adaptive weighted fusion algorithm, to achieve high-precision acquisition of iron core grounding current signals. It also addresses the problems of low detection accuracy, poor stability, and inability to capture weak signals associated with single sensors. Specifically, the Rogowski coil is used to acquire high-frequency grounding current signals (1kHz~1MHz), while the Hall sensor is used to acquire low-frequency grounding current signals (50Hz~1kHz). The fusion algorithm combines the two signals to obtain a full-frequency, high-precision grounding current signal that can accurately capture both weak and large current signals within the range of 1mA~10A, meeting the dual requirements of early fault detection and severe fault detection. The detection accuracy after fusion is uniformly ≤±0.2%FS, representing an improvement of over 60% compared to single-sensor detection accuracy.

[0020] In this embodiment, the specific design of the detection unit is as follows: The Rogowski coil adopts a hollow ring structure with a coil turn count N = 1000~2000 turns, adjusted according to the detection frequency band. 2000 turns are used for the high frequency band and 1000 turns are used for the low frequency band. The inner diameter of the coil D1 = 50~80mm, which is suitable for iron core grounding leads of different diameters (usually the diameter of the grounding lead is 10~30mm). The outer diameter D2 = 70~100mm, the thickness h = 10~20mm, and it is wound with φ0.1mm enameled copper wire at a winding density of 10 turns / mm². The coil inductance L = 10~20mH, the distributed capacitance C ≤ 100pF, and the Rogowski coil is wrapped around the transformer iron core grounding lead to collect the rate of change signal of the grounding current in a non-contact manner. A 100Ω matching resistor is connected in parallel at the coil output terminal to reduce signal reflection. The Hall sensors used are all closed-loop Hall current sensors (model ACS724), with a uniform range of 0~10A, a uniform accuracy of ±0.1%FS, a response time ≤1μs, and a uniform operating temperature range of -40℃~85℃, suitable for the harsh environment of substations. The Hall sensor is connected in series between the iron core grounding lead and the ground, and collects the effective value signal of the grounding current through contact. The sensor output is connected to the signal processing unit through a signal isolation circuit with an isolation voltage ≥2500Vrms to avoid affecting the reliability of the iron core grounding. The system utilizes electromagnetic induction theory (Rogowski coil acquisition principle) and Hall effect theory (Hall sensor acquisition principle). The Rogowski coil is based on Faraday's law of electromagnetic induction. When the iron core grounding current passes through the coil, a changing magnetic field is generated, inducing an electromotive force proportional to the rate of change of current in the coil. The grounding current can be calculated by measuring the induced electromotive force, specifically following the formula... ,in M is the induced electromotive force (V) of the Rogowski coil; H is the mutual inductance (H) between the Rogowski coil and the grounding lead of the iron core, which is determined by the number of coil turns and structural dimensions, and can be expressed by the formula. Calculate (k is the coupling coefficient, ranging from 0.95 to 0.99, adjusted according to the coil winding accuracy; the higher the winding accuracy, the larger the k value; L is the Rogowski coil's own inductance (H). The inductance (H) of the iron core grounding lead is typically taken as 1~5μH. The instantaneous value (A) of the core grounding current; This represents the rate of change of the iron core grounding current (A / s); the negative sign indicates that the direction of the induced electromotive force is opposite to the direction of the current change, conforming to Lenz's law. Hall sensors are based on the Hall effect; when current passes through a Hall element, a Hall voltage is generated under the influence of a magnetic field. The Hall voltage is proportional to the current flowing through it. By measuring the Hall voltage, the effective value of the grounding current can be obtained, specifically following the formula... ,in The output voltage (V) of the Hall sensor; The sensitivity (V / (A·T)) of the Hall sensor is determined by the sensor model; for example, the ACS724 model... The value is 0.1~0.2V / (A·T), and can be adjusted to 0.15V / (A·T) through calibration; B is the instantaneous value of the iron core grounding current (A); B is the magnetic field strength (T) inside the Hall sensor, provided by the permanent magnet inside the sensor, and B takes a value of 0.1~0.2T and remains constant. Dual sensor fusion acquisition is based on data fusion theory. By weighted and fused the signals acquired by the two sensors, the defects of a single sensor can be compensated for, and the acquisition accuracy and stability can be improved.

[0021] To achieve high-precision fusion of signals from two sensors, an adaptive weighted fusion algorithm is adopted, the specific formula of which is as follows: At the same time satisfy The weighting coefficients are dynamically adjusted based on the signal-noise variance. The core function of this set of formulas is to process the current signal acquired by the Rogowski coil. and the current signal collected by the Hall sensor Adaptive weighted fusion is performed to obtain a high-precision ground current fusion signal. The weighting coefficients can be adjusted in real time based on the noise variance of the two signals. Signals with larger noise variance receive smaller weighting coefficients, and signals with smaller noise variance receive larger weighting coefficients. This solves the problems of low detection accuracy and poor stability of single sensors, improving signal acquisition accuracy and anti-interference capability. Specific explanations and standardized values ​​for each parameter in the above formula are as follows: The instantaneous value (A) of the core grounding current after dual-sensor fusion; The weighting coefficients (unitless) for the Rogowski coil-acquired signal are dynamically adjusted in real time. When high-frequency interference is low, The value is set to 0.6~0.7, and automatically adjusted to 0.3~0.4 when there is significant high-frequency interference. The weighting coefficients (unitless) of the signals acquired by the Hall sensor are dynamically adjusted in real time, and... Complementary; The instantaneous value (A) of the core grounding current collected by the Rogowski coil is derived from... The integration time constant is uniformly set to 10μs to ensure integration accuracy. The instantaneous value (A) of the iron core grounding current acquired by the Hall sensor is obtained from... Calculated; The noise variance (A²) of the Rogowski coil signal is calculated in real time and reflects the stability of the Rogowski coil signal. Under normal operating conditions, its value is ≤1×10⁻ 6 A²; The noise variance (A²) of the signal acquired by the Hall sensor is calculated in real time and reflects the stability of the Hall sensor signal. Under normal operating conditions, its value is ≤5×10⁻ 7 A².

[0022] Implementation Instructions: The Rogowski coil uses a non-contact assembly design, directly fitting onto the transformer core grounding lead without disconnecting the lead. This facilitates installation and does not affect the normal operation of the transformer. After assembly, it is secured with cable ties with a uniform tensile strength of 50N to ensure concentricity between the coil and the lead, preventing misalignment that could affect acquisition accuracy. The Hall sensor (model ACS724) is connected in series between the core grounding lead and the grounding electrode. The lead passes through the detection hole of the Hall sensor, and the connection is made using bolts. The bolts are all M8×20mm, with a uniform tightening torque of 20N·m to ensure good contact with a contact resistance ≤0.1Ω. The crimped area is wrapped with insulating tape for insulation. The signals acquired by the dual sensors are transmitted to the signal processing unit via shielded cable (RVVP type, 1.0mm² cross-sectional area). The cable length is controlled between 1 and 3m to avoid signal attenuation. The cable shield is grounded at one end, and the grounding end is connected to the device's outer casing grounding system (grounding resistance ≤4Ω). The weighted fusion algorithm is implemented through microcontroller programming of the control and analysis unit. Written in C language and embedded in the microcontroller's main program, the algorithm uses a uniform sampling period of 10μs to calculate the noise variance and weighting coefficients of the two signals in real time, completing signal fusion. The fused signal is then transmitted to the signal processing unit via the SPI interface. During on-site debugging, both sensors need to be calibrated using a standard current source (FLUKE5520A) outputting standard currents of 0.01A, 1A, and 10A. The initial values ​​of the weighting coefficients are adjusted to ensure that the detection error after fusion is ≤ ±0.2%FS.

[0023] In this embodiment, the signal processing unit includes a signal conditioning module, an adaptive filtering module, a programmable gain amplifier, and an isolation conversion module. The signal conditioning module uses high-precision resistors to achieve current-to-voltage signal conversion, and works with a high-input-impedance operational amplifier to form a voltage follower, achieving signal impedance matching and distortion-free transmission. The adaptive filtering module uses at least a 20th-order active filter structure, and the filtering bandwidth can be adaptively adjusted within the range of 10Hz to 1MHz to suppress power grid harmonics and electromagnetic interference. The amplification factor of the programmable gain amplifier can be adjusted within the range of 1 to 1000 times to adapt to signals of different amplitudes. The isolation conversion module uses at least a 16-bit high-precision AD converter, and the sampling frequency can be adjusted within the range of 100Hz to 10kHz, with a conversion accuracy of not less than ±0.002%FS.

[0024] The specific details are as follows: In this embodiment, the signal processing unit is used to further process the fused current signal collected by the detection unit, eliminate the influence of noise signals such as power grid harmonics and electromagnetic interference, and convert the analog signal into a digital signal to provide a high-quality digital signal for signal analysis by the control analysis unit. In the prior art, the signal processing unit mostly adopts a simple low-pass filter + amplification + AD conversion structure, which has poor filtering effect and cannot effectively suppress harmonic interference. The harmonic suppression ratio is ≤20dB, and the signal amplification factor is fixed, which cannot adapt to ground current signals of different amplitudes, resulting in low detection accuracy and an error ≥3%. The signal processing unit of this invention adopts a four-stage processing structure of adaptive LMS filtering, programmable gain amplification, isolated AD conversion, and signal conditioning. Unlike the simple processing structure of the prior art, it can achieve accurate suppression of interference signals, adaptive amplification of signals of different amplitudes, and high-precision conversion. The harmonic suppression ratio is uniformly ≥60dB, solving the problems of poor filtering effect, fixed amplification factor, and low conversion accuracy in the prior art. The specific design is as follows: Signal conditioning module: used to preprocess the analog current signal after dual-sensor fusion, convert the current signal into a voltage signal, and perform impedance matching to avoid signal attenuation. Using precision resistors The current signal is converted into a voltage signal, and the conversion follows the formula... This formula is the core basis for signal conditioning, achieved by selecting high-precision resistors. This ensures the accuracy of the current-to-voltage conversion and avoids conversion errors affecting the accuracy of subsequent detection. Among these, The input voltage signal (V) after signal conditioning; The instantaneous value (A) of the core grounding current after dual-sensor fusion; For current-to-voltage conversion resistors (Ω), high-precision metal film resistors ranging from 100Ω to 1kΩ are uniformly selected, with a uniform accuracy of ±0.1% and a temperature coefficient ≤10ppm / ℃ to ensure that the conversion accuracy is not affected by temperature. The specific value is adjusted according to the detection range; for example, 1kΩ is used when detecting weak signals (1mA~100mA), and 100Ω is used when detecting large currents (100mA~10A). The impedance matching circuit uses an operational amplifier to form a voltage follower. The operational amplifier is uniformly selected as OPA2277, with an input impedance ≥1MΩ and an output impedance ≤10Ω, to achieve impedance matching between the signal source and the filtering module, reduce signal attenuation, and ensure that the voltage follower has a bandwidth ≥1MHz to ensure distortion-free signal transmission. Adaptive LMS Filtering Module: Used to suppress power grid harmonic interference and on-site electromagnetic interference. Existing fixed-parameter filtering technologies cannot adapt to interference signals under different operating conditions. This module adopts an adaptive minimum mean square error (LMS) filtering algorithm, which can adjust filtering parameters in real time to accurately suppress harmonic interference beyond the 50Hz fundamental frequency (such as the 3rd, 5th, and 7th harmonics) and high-frequency interference signals, ensuring signal purity. The filtering module uses an operational amplifier OPA2277 to form an active filter circuit, combined with microcontroller programming to implement the LMS algorithm. The filtering bandwidth can be adaptively adjusted within the range of 10Hz to 1MHz, and the filtering order is uniformly set to 24th order, balancing filtering effect and computational efficiency. The LMS filtering algorithm follows the formula: Filtered output signal... Error signal Weight coefficient update formula The core function of this set of formulas is to achieve precise suppression of interference signals. This is the filtered output voltage signal. To compensate for the error between the filtered output signal and the desired signal, the filter weights are adjusted in real time using a weight update formula. To make the error This minimizes interference, thereby suppressing power grid harmonics and electromagnetic interference, and solving the problem that fixed filters in existing technologies cannot adapt to complex interference scenarios. Compared with existing fixed filters, it has stronger anti-interference capabilities and adaptability. Specific explanations and standardized values ​​for each parameter in the above LMS filter formula are as follows: The output voltage signal (V) after LMS filtering is given, n is the sampling time, and N is the filter order, which is uniformly set to 24. The higher the order, the better the filtering effect. Considering the processing capability of the microcontroller, 24 is selected to balance the filtering effect and the processing speed. The weight coefficient is the k-th filter weight (unitless), and the weight coefficient at time n. The initial values ​​are uniformly 0.01~0.05. The input voltage signal (V) at time nk is the input signal after signal conditioning. The error signal (V) is the difference between the desired signal and the filtered output signal under normal operating conditions. ; The desired signal (V) is the voltage signal corresponding to the core grounding current without interference. It is obtained through on-site calibration. A standard current source (model FLUKE5520A) is used to output standard current and convert it into a voltage signal as the desired signal. μ is the step size factor (unitless), with a value range of 0 < μ < 1, which controls the update speed of the weighting coefficient. In this embodiment, μ = 0.01 is uniformly selected to balance update speed and stability. The filter weight coefficients (unitless) are updated in real time. The programmable gain amplifier module adaptively amplifies the filtered voltage signal to accommodate ground current signals of different amplitudes (1mA~10A corresponds to a converted voltage signal of 0.1mV~1V), avoiding low conversion accuracy due to excessively small signals or saturation distortion due to excessively large signals. It uses a programmable gain amplifier PGA202; the amplification factor can be adjusted via the control signal output from the control analysis unit, and the amplification follows the formula... This formula can adjust the amplification factor in real time according to the amplitude of the input signal. This ensures that the amplified signal amplitude is within the input range of the AD conversion module (0~5V), preventing low conversion accuracy due to excessively small signals or saturation distortion due to excessively large signals, thereby improving detection accuracy. The amplified voltage signal (V); This refers to the programmable amplification factor (unitless), the amplification factor at time n, with a uniform range of 1 to 1000 times and a step size of 1, determined by the control and analysis unit based on... The amplitude is controlled in real time, and the specific control logic is unified as follows: When <5mV =1000;5mV≤ When <50mV =100;50mV≤ When <500mV =10; At ≥500mV, =1; The output voltage signal (V) is after LMS filtering. The programmable gain amplifier PGA202 has a bandwidth ≥1MHz and an input offset voltage ≤10μV, ensuring distortion-free and high-precision amplification. An isolated AD converter module converts the amplified analog voltage signal into a digital signal, while simultaneously achieving electrical isolation between the input and output signals to further suppress electromagnetic interference. A high-precision 16-bit AD converter, ADS8364, is uniformly selected. The sampling frequency is adjustable from 100Hz to 10kHz, the conversion accuracy is uniformly ±0.0015%FS, and the sampling period is ≤10μs. The AD conversion follows the formula... This formula is the core basis of AD conversion, using a high-precision reference voltage. To ensure conversion accuracy, the 16-bit conversion can achieve 65,536 discrete levels, improving the detection accuracy of weak signals. The value is a 16-bit digital signal (unitless) after AD conversion, with a value range of 0~65535; The amplified voltage signal (V); For the reference voltage (V) of the AD converter, the high-precision reference voltage source REF5050 is uniformly selected. The accuracy is uniformly ±0.01%, and the temperature coefficient is ≤5ppm / ℃; This is the maximum quantization level (65536) for the 16-bit AD converter; the subtraction of 1 is because digital signals start counting from 0. An opto-isolator TLP521 is used for isolation between the AD converter and the control analysis unit, with an isolation voltage ≥2500Vrms, ensuring the converted digital signal is stable, reliable, and unaffected by interference.

[0025] Relevant theoretical explanations: Adaptive filtering theory, signal amplification theory, and AD conversion theory. The adaptive LMS filtering algorithm, based on the minimum mean square error criterion, minimizes the mean square error between the filtered output signal and the desired signal by adjusting the filter weight coefficients in real time, thus achieving precise suppression of interference signals. It is suitable for scenarios with uncertain interference signals (such as complex electromagnetic environments in substations). Programmable gain amplification allows for adjustable amplification by changing the amplifier's feedback resistor value, adapting to input signals of different amplitudes. AD conversion discretizes continuous analog signals into digital signals, providing a foundation for digital signal processing. A 16-bit AD converter can achieve high-precision signal conversion, meeting the needs of weak signal detection. Signal isolation technology, based on the principle of opto-isolation, uses the TLP521 optocoupler to achieve electrical isolation between input and output, preventing interference signals from being transmitted to the control unit via cables. Implementation Notes: The operational amplifier used in the signal conditioning module is the OPA2277, which features low noise (input noise voltage ≤1nV / √Hz), high precision, and wide bandwidth, making it suitable for weak signal conditioning. The operational amplifier's power supply voltage is uniformly ±15V, provided by a dedicated power supply unit to ensure stable operation. The LMS filtering algorithm is implemented through programming the STM32F407 microcontroller in the control and analysis unit, written in C language and embedded in the main program. It acquires the input signal in real time, calculates the weighting coefficients and the filtered output, with a uniform operation cycle of 10μs. The programmable gain amplifier PGA202 controls the amplification factor through the microcontroller's I2C interface, with a uniform I2C communication rate of 10. The amplifier operates at 0kHz, enabling adaptive amplification. A 100Ω current-limiting resistor is connected in series at the output of the amplifier to protect the subsequent AD conversion module. The sampling frequency of the ADS8364 AD converter is controlled by the GPIO port of the microcontroller and adjusted according to the detection requirements. During normal detection, a sampling frequency of 1kHz is selected, and it is increased to 10kHz during fault detection to capture the instantaneous current change during the fault. The conversion result of the AD converter is transmitted to the microcontroller through the SPI interface, and the SPI communication rate is uniformly 1MHz. The TLP521 opto-isolator is connected in series between the AD converter and the microcontroller to ensure electrical isolation and suppress interference. The input current of the isolator is controlled at 5~10mA, and the output current is controlled at 1~5mA. During on-site debugging, the signal processing unit needs to be calibrated. A standard voltage signal of 0.1mV~5V is output from a standard signal source and input to the signal processing unit. The LMS filter parameters and amplification factor are adjusted to ensure that the output digital signal error is ≤±0.0015%FS. At the same time, the anti-interference capability of the signal processing unit is tested. An electromagnetic interference simulator (model EMC-2000) is used to generate 40dBμV / m electromagnetic interference to ensure that the output error of the signal processing unit does not change significantly (the change is ≤0.0005%FS).

[0026] In this embodiment, the built-in standard signal source of the self-test calibration unit can generate a standard current signal in the range of 0~10A and 50Hz~1kHz, with a signal accuracy of not less than ±0.01%FS; it can realize rapid switching between normal detection and self-test calibration modes, with a switching response time of no more than 10ms; during the calibration process, the calibration coefficient is obtained by calculating the ratio of the standard signal value to the actual detection value of the device, and the detection data is corrected using the calibration coefficient to ensure that the corrected detection accuracy is maintained within ±0.2%FS, and the calibration cycle is within the range of 1 day to 1 month.

[0027] The specific details are as follows: The self-test calibration unit is a key module to ensure the detection accuracy and operational reliability of the device. In the existing technology, most detection devices lack self-test functions or only have simple power-on self-tests, which cannot detect core faults such as sensor failures and signal processing module abnormalities. Moreover, calibration requires an external standard signal source, which is cumbersome to operate and cannot achieve automatic on-site calibration. The calibration cycle is long (usually 3 months), which affects the continuity of detection.

[0028] The self-test calibration unit in this embodiment adopts an integrated design of real-time self-testing, periodic calibration, and fault location, which differs from the simple self-testing schemes of existing technologies. It realizes comprehensive self-testing, automatic calibration, and accurate fault location for all modules of the device, solving the problems of incomplete self-testing, cumbersome calibration, and inability to locate faults in existing technologies. The self-test calibration unit integrates a built-in standard signal source, fault detection circuit, and calibration algorithm, and is deeply integrated with the detection unit, signal processing unit, and control analysis unit to achieve synchronous self-testing and detection without affecting the normal detection function of the device. It meets the needs of on-site operation and maintenance, and the calibration cycle can be flexibly set (1 day to 1 month) to ensure long-term stable detection accuracy.

[0029] The self-test calibration unit is designed as follows: Built-in standard signal source: used to generate a high-precision standard current signal, providing a reference for self-testing and calibration. A high-precision signal generator AD9833 is uniformly selected, capable of generating a standard current signal in the range of 0~10A, with a uniform accuracy of ±0.01%FS. The frequency is adjustable in the range of 50Hz~1kHz. The signal source output is connected to the input of the detection unit via a switching switch. The switching switch is uniformly selected as a relay K1 (model G6K-2P-Y), with a uniform control voltage of 5V. Switching is controlled by the control and analysis unit. During normal testing, the switching switch connects the detection unit and the signal processing unit; during self-testing or calibration, the switching switch connects the built-in standard signal source and the signal processing unit. The switching response time is ≤10ms. The standard current signal generated by the standard signal source follows the formula... This formula is used to generate high-precision standard current signals. As a reference signal for self-testing and calibration, the amplitude, frequency, and phase of this signal are adjustable to adapt to the self-testing and calibration requirements under different operating conditions, ensuring the accuracy of self-testing and calibration. The standard instantaneous current value (A) output by the built-in standard signal source; The effective value (A) of the standard current is adjustable in the range of 0~10A with an accuracy of ±0.0196FS. Three typical values ​​of 0.01A, 1A and 10A are often selected for calibration. The frequency (Hz) of the standard current is adjustable from 50Hz to 1kHz and is used to simulate ground current signals of different frequencies. 50Hz (fundamental), 150Hz (3rd harmonic), and 250Hz (5th harmonic) are commonly selected for calibration. t is time (s); φ is the phase (rad) of the standard current, with a value of 0 to 2π. The default value is 0 to ensure stable signal phase.

[0030] Real-time self-test module: Used for real-time fault detection of various modules in the device, covering the detection unit (Rogowski coil, Hall sensor), signal processing unit (filtering, amplification, AD conversion), power supply unit, and data transmission unit. It collects the operating status signals of each module through fault detection circuits and transmits them to the control and analysis unit, which then determines whether the module is functioning correctly. The self-test frequency is uniformly 1 time / second to ensure timely fault detection. Specifically, it includes: sensor fault detection (checking if the sensor output signal is within the normal range), filter module fault detection (checking if the noise of the filtered output signal exceeds the standard), amplification module fault detection (checking if the amplification factor meets the set value), AD conversion fault detection (checking if the converted digital signal is abnormal), power supply fault detection (checking if the power supply output voltage is stable), and data transmission fault detection (checking the communication status of the transmission module). Fault judgment for each module follows corresponding formulas to ensure accurate judgment and avoid misjudgment and omission: Sensor fault judgment: ,in The actual instantaneous current value (A) collected by the detection unit, i.e., the value after fusion of the two sensors. ; The standard instantaneous current value (A) output by the built-in standard signal source; The sensor fault judgment threshold (A) is set according to the detection accuracy, and is uniformly set to 0.001A (1mA) to ensure that even minor sensor faults can be detected and to avoid missed detections. When this formula is met, the sensor is judged to be faulty, specifically a Rogowski coil or Hall sensor fault, which will be further distinguished by the fault location module.

[0031] Amplification module fault diagnosis: ,in The actual magnification factor of the amplification module (unitless), determined by... Calculation yields ( This is the actual amplified voltage signal. (The filtered voltage signal). The amplification factor (unitless) set for the control analysis unit, i.e., the target amplification factor set according to the amplitude of the input signal; A threshold value of 1 is set for diagnosing amplification module faults (unitless) to ensure that the amplification factor deviation is within the allowable range and to avoid decreased detection accuracy due to abnormal amplification factor. When this formula is met, the amplification module is considered faulty, possibly due to a damaged PGA202 chip or an abnormal control signal. AD conversion fault diagnosis: ,in The actual digital signal after AD conversion (unitless); It is a standard digital signal (unitless), derived from a standard voltage signal. Substituting into the AD conversion formula, we can calculate ( (Standard input voltage signal); A threshold value of 5 is set for judging AD conversion faults (unitless). This is to take into account the inherent error of AD conversion and to set a reasonable threshold to avoid false judgments while ensuring that conversion faults can be identified in a timely manner. When this formula is met, an AD conversion fault is judged, which may be due to damage to the ADS8364 chip, abnormal reference voltage, or failure of the opto-isolator TLP521.

[0032] In addition, power supply unit faults are detected by detecting the power supply output voltage. When the DC5V output voltage deviates from the set value by ±5% or the DC12V output voltage deviates from the set value by ±5%, a power supply fault is determined. Data transmission faults are detected by detecting the communication response signal of the transmission module. If no response signal is received for three consecutive times, a data transmission fault is determined.

[0033] Periodic Calibration Module: Used for periodic automatic calibration of the device's detection accuracy. The calibration cycle can be set by the user (e.g., once a day, once a week) or manually triggered. Calibration time is ≤30 seconds and does not affect normal device detection. During calibration, the control and analysis unit controls the switching switch (relay K1, model G6K-2P-Y) to switch to the built-in standard signal source (AD9833), generating standard current signals of different amplitudes. After processing by the signal processing unit, the detection value is obtained, compared with the standard value, and the calibration coefficient is calculated to correct the detection value and ensure detection accuracy. The calculation of the calibration coefficient follows the formula... The corrected detection value follows the formula The core function of this set of formulas is to calculate the calibration coefficient. Furthermore, it corrects the actual detection values ​​of the device using calibration coefficients, eliminating detection errors and ensuring long-term stability of detection accuracy. It solves the problem of requiring manual calibration with an external standard signal source in existing technologies, achieving automatic calibration and improving maintenance convenience, which is one of the core innovations of the self-test calibration unit. The calibration coefficient (unitless) reflects the detection deviation of the device; under ideal conditions... =1, in actual calibration The value range is uniformly 0.998~1.002; The standard effective value of the current (A) output by the built-in standard signal source; The actual effective value (A) of the device's detection of the standard current signal; The instantaneous value of the calibrated grounding current (A); The value is the actual instantaneous value (A) before calibration. Fault location module: Used for precise fault location. When the real-time self-test module detects a fault, the control analysis unit analyzes the signals collected by the fault detection circuit to determine the faulty module (e.g., Rogowski coil fault, AD conversion fault), and transmits the fault information to the display alarm unit and remote monitoring platform, prompting maintenance personnel to perform repairs. Simultaneously, if the fault does not affect the device's basic detection functions, the device can switch to standby mode (e.g., single sensor acquisition) to ensure continuous operation. Fault location uses a fault coding mechanism. Each faulty module corresponds to a unique fault code, uniformly set as follows: Rogowski coil fault code 001, Hall sensor fault code 002, AD conversion fault code 003, power supply fault code 004, and data transmission fault code 005. The fault codes are preset through microcontroller programming; when a fault occurs, the corresponding fault code and fault description are automatically invoked.

[0034] Relevant theoretical explanations: Standard signal calibration theory and fault detection theory. Standard signal calibration theory is based on the comparative calibration method. By comparing the device's detected values ​​with the standard values ​​of a high-precision standard signal, calibration coefficients are calculated to correct detection errors and ensure detection accuracy. Fault detection theory is based on signal feature analysis. By collecting the operating status signals of each module (such as output voltage, current, and communication signals), it analyzes whether the signal characteristics conform to the normal range to determine whether a module is faulty. Different module faults will exhibit different signal characteristics; through feature analysis, precise fault location can be achieved.

[0035] Implementation Notes: The built-in standard signal source AD9833 of the self-test calibration unit is connected to the STM32F407 microcontroller of the control and analysis unit via an SPI interface. The SPI communication rate is uniformly 1MHz. The microcontroller controls the amplitude, frequency, and phase of the standard current signal output. A 10Ω current-limiting resistor is connected in series at the output of the signal source to protect subsequent modules. The switching relay K1 (model G6K-2P-Y) is a small electromagnetic relay with a control voltage of 5V. It is controlled by the GPIO port of the microcontroller, and the switching response time is ≤10ms to ensure rapid switching between self-test and normal detection without affecting the continuity of detection. The fault detection circuit uses a voltage comparator LM311 to compare the working status signals of each module with the set threshold and output high and low level signals to the microcontroller. The threshold of the voltage comparator can be adjusted by a potentiometer, with a uniform accuracy of ±1mV. The microcontroller determines whether the module is faulty by identifying the high and low levels. The calibration cycle of the periodic calibration module can be set through the remote monitoring platform or local buttons on the device. The default calibration cycle is 2:00 AM every day (transformer). During periods of low load and minimal interference, calibration generates three sets of standard current signals with different amplitudes (0.01A, 1A, and 10A), corresponding to weak, normal, and high-current signals, respectively. The calibration coefficient for each set is calculated, and the average value is used as the final calibration coefficient to ensure comprehensive calibration. The calibration coefficient is stored in the microcontroller's Flash memory and is used to correct the detected value during each test. The fault location module implements fault coding through microcontroller programming. Different module faults correspond to different fault codes (e.g., Rogowski coil fault code 001, AD conversion fault code 002). The fault code and fault information are transmitted to the display alarm unit and remote monitoring platform, allowing maintenance personnel to quickly locate the faulty module based on the fault code, improving repair efficiency. When a single sensor fault is detected, the microcontroller automatically switches to another sensor to collect the signal, ensuring continuous device operation and preventing detection interruptions due to a single sensor failure. When a core module (such as the AD converter or control chip) fault is detected, the device stops detecting, issues an emergency alarm, and prompts maintenance personnel to perform immediate repairs. During on-site commissioning, the self-test calibration function needs to be manually triggered to simulate faults in each module, check the accuracy of fault location and calibration effect, ensure that the fault location error is 0, and the detection accuracy after calibration is ≤ ±0.2%FS.

[0036] In this embodiment, the dual-core architecture of the control and analysis unit consists of a main control chip and a dedicated data processing chip. The main control chip is a high-performance 32-bit microcontroller with a main frequency of not less than 160MHz, and the dedicated data processing chip is a high-speed digital signal processing chip with a computing speed of not less than 500MIPS. The dynamic threshold correction algorithm combines three key operating condition parameters, namely ambient temperature, transformer load, and power grid frequency, and adjusts the fault judgment threshold in real time through a preset correction formula.

[0037] The control and analysis unit is the core control and data processing hub of the device, undertaking core functions such as coordinated control of various modules, in-depth analysis of detection data, fault logic judgment, dynamic threshold correction, and command issuance. It is crucial for achieving high-precision grounding current detection and accurate fault diagnosis. In existing technologies, control and analysis units mostly use 8-bit microcontrollers, which have slow processing speeds (main frequency ≤100MHz), weak data processing capabilities, and cannot achieve synchronous operation of multiple algorithms (such as weighted fusion, LMS filtering, and harmonic analysis). Furthermore, they lack dynamic threshold adjustment mechanisms, can only perform simple current RMS value calculations, have simplistic fault judgment logic, cannot adapt to the complex operating conditions of substations, and are prone to misjudgments and omissions. Additionally, poor interface compatibility with other units results in an overall device response delay ≥100ms. The control and analysis unit of this invention adopts a dual-core architecture of a high-performance microcontroller and a dedicated data processing chip, unlike the single-microcontroller design of existing technologies. This significantly improves data processing speed and control accuracy, enabling synchronous operation of multiple algorithms, coordinated control of various units, and real-time dynamic threshold correction, solving the problems of slow processing speed, weak data processing capabilities, fixed thresholds, and high response delays in existing technologies. Meanwhile, it integrates multiple interface modules to achieve seamless connection with various modules such as the detection unit, signal processing unit, and self-calibration unit, ensuring smooth command transmission and stable data interaction. The overall response delay of the device is ≤10ms, meeting the needs of real-time detection and rapid fault response.

[0038] The specific design of the control and analysis unit is as follows: Hardware selection and design: The core controller uniformly uses the STM32F407VET6 microcontroller as the main control chip, with a unified main frequency of 168MHz, built-in 512KB Flash and 192KB RAM, possessing high-speed computing power and sufficient storage capacity. It can simultaneously run multiple algorithms such as weighted fusion, LMS filtering, harmonic analysis, and fault diagnosis to meet the real-time data processing requirements. The dedicated data processing chip uses the ADSP-BF533 to assist the main controller in completing complex data calculations (such as harmonic component and waveform distortion rate calculations), with a computing speed ≥600MIPS, reducing the load on the main controller and ensuring the real-time performance and accuracy of data processing. Data interaction between the two is achieved through the SPI interface, with a unified SPI communication rate of 1MHz and a data transmission delay ≤1μs, ensuring real-time synchronization of the calculated data. The supporting hardware module design is standardized and the specific parameters are as follows: (1) Clock module: High-precision crystal oscillator (model HC-49S) is selected, with a frequency of 12MHz and an accuracy of ≤±10ppm / ℃, to provide a stable clock signal for the microcontroller and data processing chip, and to avoid data processing errors caused by clock deviation; (2) Storage module: Extended SD card storage (capacity ≥8GB) is used to store test data, self-test results, calibration coefficients and fault records, with a storage period of ≥1 year, and supports data cycle overwriting. At the same time, the FRAM chip FM24CL64 (capacity 64KB) is integrated to save key parameters (such as calibration coefficients and fault codes) when power is off, with a power-off retention time of ≥10 years to ensure that the parameters are not lost; (3) Interface module: Integrated SPI Interface (2 channels, 1MHz rate) is connected to the signal processing unit AD converter and data processing chip respectively; integrated I2C interface (1 channel, 100kHz rate) is connected to the programmable gain amplifier and built-in standard signal source; integrated UART interface (2 channels, 9600bps baud rate) is connected to the display alarm unit and data transmission unit; integrated GPIO interface (16 channels) is used to control the switching switch, fault alarm trigger and the working status detection of each module. All interfaces are photoelectrically isolated (isolation voltage ≥2500Vrms) to suppress electromagnetic interference; (4) Reset module: adopts a dual design of power-on reset + manual reset. The reset voltage is uniformly 3.3V and the reset time is ≤10ms to avoid the device from crashing due to voltage fluctuation.

[0039] Software Design: The software of the control and analysis unit adopts a modular programming approach, is developed based on C language, and is embedded in the STM32F407VET6 microcontroller. The main program flow is clear and is divided into an initialization module, a data acquisition module, a data processing module, a self-test interaction module, a fault judgment module, a threshold correction module, a data transmission module, and a display control module. Each module operates independently and works together to ensure the stability and reliability of the device. The software running cycle is uniformly 10μs and is synchronized with the sampling cycle of each unit. The core functions and design details of each software module are as follows: (1) Initialization module: After the device is powered on, the microcontroller, data processing chip, each interface module, and storage module are initialized. The calibration coefficient, fault record and threshold parameters stored in the FRAM are read, and the initial working state of each module is set (such as sampling frequency, initial value of amplification factor, self-test frequency). After initialization, a ready signal is sent to the display alarm unit to indicate that the device can work normally. The initialization time is ≤500ms; (2) Data acquisition module: The digital signal output by the signal processing unit is acquired in real time through the SPI interface and transmitted through the I2C interface. The self-test results and standard signal data of the self-test calibration unit are collected. The sampling frequency is consistent with the sampling frequency of the signal processing unit (1kHz for normal detection and 10kHz for fault detection). After the data is collected, a preliminary verification is performed to remove abnormal data (such as exceeding the range or data mutation) to ensure the validity of the collected data; (3) Data processing module: The collected digital signal is transmitted to the ADSP-BF533 data processing chip to complete the calculation of the effective value of grounding current, harmonic components and waveform distortion rate. The calculation results are transmitted back to the microcontroller. At the same time, the detection data is corrected in combination with the calibration coefficient to ensure the accuracy of the data. Data processing delay ≤ 5μs; (4) Self-test interaction module: receives self-test data from the self-test calibration unit in real time, judges the working status of each module, and if a fault is detected, triggers the fault location module, calls the corresponding fault code, and controls the device to switch to standby mode (such as single sensor acquisition) to ensure detection continuity; (5) Fault judgment module: combines data processing results and dynamic thresholds to judge the core grounding status, distinguishes between normal single-point grounding, mild multi-point grounding fault, and severe multi-point grounding fault, and issues corresponding control commands (normal command, early warning command, emergency alarm command); (6) Threshold correction module :Combining ambient temperature, power grid frequency, transformer load and other operating conditions, the fault judgment threshold is adjusted in real time through the threshold correction formula to avoid false alarms and missed alarms; (7) Data transmission module: After organizing the corrected detection data, fault information and self-test results, it is transmitted to the data transmission unit through the UART interface, and key data is stored in the SD card, supporting real-time transmission and historical data backtracking; (8) Display control module: The detection data, working status and fault information are transmitted to the display alarm unit, and the display interface is updated in real time. If a fault is detected, an audible and visual alarm is triggered to prompt the maintenance personnel to handle it in time.Core data processing algorithms and formulas: The core of the control analysis unit is data processing and fault diagnosis. Combining the detection data and operating parameters mentioned above, we supplement key calculation formulas and algorithms to ensure consistency with the overall parameters, as follows: (1) Calculation of the effective value of grounding current: Based on the collected instantaneous value of grounding current, the effective value is calculated using the root mean square method, and the formula is as follows: ,in The effective value of the grounding current (A); N is the number of sampling points, which corresponds to the sampling frequency (N=1000 for 1kHz sampling, N=10000 for 10kHz sampling). The instantaneous grounding current value (A) after calibration at the nth sampling point; this formula can accurately calculate the effective value of the grounding current, avoiding judgment errors caused by instantaneous value fluctuations, with a calculation accuracy ≤ ±0.2%FS, consistent with the overall accuracy of the detection unit. (2) Harmonic component calculation: The Fast Fourier Transform (FFT) algorithm is used to perform harmonic analysis on the calibrated grounding current signal, extracting the fundamental wave (50Hz) and the 3rd, 5th, and 7th harmonic components, with the formula as follows. ,in Let be the effective value (A) of the h-th harmonic, where h = 3, 5, 7; The active component (A) of the h-th harmonic; The reactive component (A) of the h-th harmonic is given; the total harmonic distortion (THD) is calculated using the following formula: ,in The effective value of the fundamental current (A); under normal operating conditions, THD ≤ 5%, under fault conditions, THD will increase significantly, which can be used as an auxiliary basis for fault judgment. The calculation accuracy of the FFT algorithm is ≤ ±0.1%, ensuring accurate detection of harmonic components. (3) Dynamic threshold correction formula: Combining the three key operating condition parameters of ambient temperature (T), transformer load (P), and power grid frequency (f), the fault judgment threshold is corrected in real time. The correction formula is as follows: This formula is the core basis for dynamic threshold judgment. It can adjust the threshold in real time according to the on-site working conditions, avoiding false alarms and missed alarms caused by changes in working conditions, echoing the innovative point of dynamically adjustable detection threshold mentioned earlier. Among them, Let A be the dynamic fault judgment threshold at time t. The baseline threshold (A) is uniformly set as follows: 0.1A for minor faults and 1A for severe faults; T is the ambient temperature (°C), which is collected by a temperature sensor (model DS18B20, accuracy ±0.5°C, operating temperature range -40°C to 85°C). The standard ambient temperature (°C) is uniformly set to 25°C; P is the actual transformer load (kVA), which is collected through the current transformer. is the rated load of the transformer (kVA); f is the actual frequency of the power grid (Hz), which is collected by a frequency acquisition module (model LMV831, accuracy ±0.1 Hz); is the standard frequency of the power grid (Hz), which is uniformly set to 50 Hz; is the temperature correction coefficient (°C⁻¹), and the unified value is 0.002; is the load correction coefficient, and the unified value is 0.05; is the frequency correction coefficient (Hz⁻¹), and the unified value is 0.01; all correction coefficients are determined through on-site calibration, stored in the FRAM, and can be adjusted through the remote monitoring platform. (4) Fault judgment logic: Combining the effective value of the grounding current , the total harmonic distortion rate THD and the dynamic threshold , a clear fault judgment logic is formulated, and the unified judgment criteria are as follows: ① Single-point grounding normal condition: and THD ≤ 5%, where is the dynamic threshold for mild faults, and the device detects normally and shows normal grounding; ② Mild multi-point grounding fault: or 5% < THD ≤ 10%, where is the dynamic threshold for severe faults, and the device issues a warning signal, shows a mild grounding fault, and records the fault data at the same time; ③ Severe multi-point grounding fault: or THD > 10%, and the device issues an emergency audible and visual alarm, shows a severe grounding fault, and transmits the fault information to the remote monitoring platform preferentially at the same time, triggering an operation and maintenance warning.

[0040] Relevant theoretical explanations: The design of the control and analysis unit is based on the embedded system theory, digital signal processing theory, and dynamic threshold judgment theory. The embedded system theory realizes the collaborative control of each module and real-time data processing. The STM32F407VET6 single-chip microcomputer and the ADSP-BF533 dual-core architecture are selected, taking into account the control flexibility and operation speed, and are suitable for real-time control scenarios in industrial fields; the digital signal processing theory realizes harmonic component analysis through the FFT algorithm and calculates the effective value of the current through the root mean square method, ensuring the accuracy and real-time nature of data processing, and solving the problems of single data processing and low accuracy in the existing technology; the dynamic threshold judgment theory corrects the fault threshold in combination with working condition parameters, based on the principle of adaptive control, making the fault judgment more in line with the actual situation on site, avoiding misjudgment and missed judgment caused by fixed thresholds, and improving the accuracy of fault judgment.

[0041] Implementation Instructions: Hardware soldering of the control and analysis unit must adhere to industrial-grade soldering standards. The STM32F407VET6 microcontroller and ADSP-BF533 data processing chip are surface-mount soldered, with the soldering temperature controlled at 260±5℃ and the soldering time ≤3s to avoid chip damage. Connections between the clock module, storage module, interface module, and microcontroller require shortened lead lengths, not exceeding 5cm, to reduce signal interference. All interfaces must be soldered with transient suppression diodes (SMBJ6.5CA) to prevent surge voltage damage. The DS18B20 temperature sensor is installed inside the device housing, close to the detection unit, ensuring the ambient temperature collected matches the operating temperature of the detection unit. The sensor leads use RVVP shielded cable (0.5mm² cross-section), with the shield grounded at one end. The LMV831 frequency acquisition module connects to the power grid transformer; the connection points are insulated to avoid the risk of electric shock. The software programming utilizes the Keil 5 development environment, written in C language. Each module is compiled independently and debugged jointly to ensure stable program operation without crashes or freezes. A watchdog timer (STM32 built-in watchdog, 1s overflow time) is added to the program to automatically trigger a reset if an exception occurs, ensuring the device returns to normal operation. Data processing algorithms (FFT, root mean square calculation) need to be optimized to reduce computational complexity and ensure data processing is completed within a 10μs cycle. The dynamic threshold correction parameter needs to be determined through on-site debugging, adjusting the correction coefficient k according to different operating conditions (high temperature, low temperature, full load, light load). T k P k f To ensure accurate threshold correction, the false alarm rate should be ≤0.1%, and the missed alarm rate should be 0. During on-site commissioning, different core grounding states (normal, minor fault, and severe fault) should be simulated, and grounding current signals of different amplitudes should be input to check the data processing accuracy and fault judgment accuracy of the control analysis unit, ensuring that the fault judgment error is 0. At the same time, simulate ambient temperature changes (-40℃~85℃), load changes (0~100% rated load), and frequency fluctuations (49.5~50.5Hz) to check the effect of dynamic threshold correction, ensuring that the threshold adjustment is timely and accurate. In addition, the interaction performance between the control analysis unit and other units should be tested to ensure smooth command transmission, response delay ≤10ms, and no data loss or error in data transmission. The dual-core architecture improves data processing speed and accuracy, incorporates a multi-condition dynamic threshold correction algorithm to achieve accurate fault diagnosis, and combines self-test results to achieve adaptive fault switching. This solves the problems of slow processing speed, fixed thresholds, inaccurate fault diagnosis, and poor interaction with various units in existing technologies. At the same time, the modular software design facilitates later upgrades and maintenance, and the standardized hardware interfaces and unified parameters improve the compatibility and scalability of the device, meeting the actual needs of substation on-site operation and maintenance.

[0042] In this embodiment, the dual-mode anti-interference transmission mode of the data transmission unit adopts a combination of high-speed wireless communication and long-distance wireless communication to realize real-time data transmission and long-distance backup transmission; the data encryption technology adopts hardware encryption and is combined with a data verification mechanism to ensure the security and integrity of the transmitted data; the dual-mode switching is automatically triggered according to the signal strength of the main communication mode, and the switching delay does not exceed 100ms; the breakpoint resume technology can cache data that has not been transmitted and continue transmission from the breakpoint after the transmission is restored.

[0043] In this embodiment, the data transmission unit serves as a data interaction bridge between the device and the remote monitoring platform and the display alarm unit. Its core function is to transmit the detection data, fault information, and self-test results output by the control and analysis unit to the remote monitoring platform in real time. Simultaneously, it receives control commands (such as calibration triggering, parameter adjustment, and fault reset) issued by the remote monitoring platform, ensuring that maintenance personnel can remotely monitor the device's operating status and the transformer core grounding condition in real time. This solves the problems of unstable data transmission, weak anti-interference capability, single transmission method, and poor data security in existing technologies. In existing technologies, data transmission for transformer core grounding current detection devices often uses a single 4G or wired transmission method. Wired transmission is cumbersome to install, limited by substation wiring, and has poor adaptability. Single wireless transmission is susceptible to strong electromagnetic interference, easily leading to transmission interruptions and data loss. It also has a limited transmission distance (≤1km) and lacks a data encryption mechanism, making the detection data easily tampered with, failing to meet the long-distance, high-reliability, and high-security data transmission requirements of substations. The data transmission unit of this invention adopts a 5G+LoRa dual-mode anti-interference transmission design, combined with data encryption and breakpoint resume technology, to meet both real-time and long-distance transmission requirements, adapt to the complex electromagnetic environment of substations, and simultaneously support local data interaction and remote operation and maintenance. The specific design is as follows: Hardware Selection and Design: The core hardware of the data transmission unit includes a 5G module, a LoRa module, an encryption module, an interface conversion module, and an antenna module. All hardware uses industrial-grade products, adapted to the harsh outdoor environment of substations (operating temperature -40℃~85℃, humidity 0~95%RH, non-condensing). Parameters are standardized. Specific selection and design are as follows: 5G Module: An industrial-grade 5G module (Quectel EC200S) is uniformly selected, supporting SA / NSA dual-mode networking, communication rate ≥100Mbps, transmission latency ≤10ms, supporting full network compatibility, and adapting to the existing 5G network of the substation. It is mainly used for real-time transmission of detection data and fault information, meeting the requirements of short-distance, high-bandwidth real-time transmission. The module power supply voltage is uniformly DC5V, power consumption ≤2W, integrated SIM card slot, supports industrial-grade IoT SIM cards, and has an anti-drop design. The module and the control and analysis unit are connected via a UART interface, with a uniform baud rate of 9600bps and data transmission format of JSON to ensure data interaction compatibility. LoRa Module: Industrial-grade LoRa modules (model SX1278) are uniformly selected, with a communication frequency of 433MHz (industrial unlicensed band). The transmission distance is ≥3km (open environment), with strong anti-interference capabilities, receiving sensitivity ≤-148dBm, and power consumption ≤500mW. Primarily used for long-distance transmission and as backup transmission during 5G network outages, ensuring uninterrupted data transmission. The module supports star topology, allowing simultaneous connection of multiple detection devices for centralized monitoring of multiple transformers in substations. The module connects to the control and analysis unit via an SPI interface, with a unified communication rate of 1MHz and a transmission format consistent with the 5G module, enabling seamless dual-mode switching. Encryption Module: Utilizes a hardware encryption chip (AES128), supporting the AES-128 encryption algorithm for real-time encryption and decryption of transmitted data. The encryption key is stored internally, making it unreadable and tamper-proof, ensuring the security of detection data and fault information and preventing data theft or alteration. The encryption module connects to the control and analysis unit via an I2C interface, with an encryption latency ≤1μs, ensuring uninterrupted data transmission. The encryption key can be periodically updated via a remote monitoring platform, with a flexible update cycle (1-3 months) further enhancing data security. Interface Conversion Module: Employs an industrial-grade interface conversion chip (MAX3232) to convert between UART and RS232 interfaces, connecting the display and alarm unit to the local maintenance terminal. It supports local data reading, parameter debugging, and fault diagnosis. The interface conversion module has an input voltage of DC5V, a conversion rate ≤115200bps, and features anti-static and surge protection. The interface uses a DB9 connector for secure and reliable connections.Antenna Modules: The 5G module is equipped with an omnidirectional fiberglass antenna with a gain ≥5dBi, a uniform impedance of 50Ω, and an operating frequency of 700~2700MHz. It is installed on the top of the device housing and features a waterproof design (IP67 protection rating) to prevent damage from rain and dust. The LoRa module is equipped with an omnidirectional silicone antenna with a gain ≥3dBi, a uniform impedance of 50Ω, and an operating frequency of 433MHz. It is installed on the side of the device housing and kept at a distance of ≥10cm from the 5G antenna to avoid signal interference between the antennas. Both antennas use SMA connectors and are detachable for easy maintenance and replacement.

[0044] The hardware circuit of the data transmission unit adopts a shielded design, with a double-layer shielded structure on the printed circuit board (PCB). The grounding resistance is ≤4Ω, and it is connected to the overall grounding system of the device to effectively suppress electromagnetic interference. A series fuse (model 0452005.MRL, rated current 500mA) is used in the circuit to prevent overcurrent damage to the module. A parallel capacitor (100μF + 0.1μF) is connected at the power supply end to filter out power supply noise and ensure stable power supply to the module. Software design and transmission mechanism: The software of the data transmission unit is based on embedded programming and works in conjunction with the control and analysis unit and the remote monitoring platform. The core functions include dual-mode switching logic, encrypted data transmission, breakpoint resume, and command interaction. The software operation cycle is consistent with that of the control and analysis unit (10μs) to ensure data synchronization.

[0045] Dual-mode switching logic: The software monitors the 5G network signal strength in real time (via the signal strength detection interface of the 5G module). When the 5G signal strength is ≥-85dBm, 5G transmission mode is prioritized to ensure real-time performance. When the 5G signal strength is <-85dBm or the 5G network is interrupted, it automatically switches to LoRa transmission mode for backup transmission, with a switching latency of ≤100ms and no data loss during the switching process. When the 5G network returns to normal (signal strength ≥-85dBm, lasting 3 seconds), it automatically switches back to 5G transmission mode to ensure data transmission stability and real-time performance. The switching logic is implemented through microcontroller programming and embedded in the control chip of the data transmission unit, requiring no manual intervention. Encrypted Data Transmission: Data transmitted by the control and analysis unit is first sent to the encryption module, where it is encrypted using the AES-128 encryption algorithm. A data checksum (CRC32 checksum) is added to the encrypted data. Then, depending on the current transmission method, it is sent to the 5G module or LoRa module, which transmits it to the remote monitoring platform. Upon receiving the data, the remote monitoring platform performs a CRC32 checksum. If the checksum passes, the data is decrypted to obtain the original detection data and fault information, ensuring the security and integrity of data transmission and preventing data tampering or loss. The CRC32 checksum formula is... Where data(i) is the i-th data byte, n is the number of data bytes, and the verification accuracy is ≥99.99%, which can effectively detect bit errors during data transmission. Resume interrupted transmission mechanism: The data transmission unit has a built-in data cache module (capacity ≥64KB) to cache data to be transmitted. When transmission is interrupted (e.g., network interruption, signal loss), the untransmitted data is stored in the cache module and will not be lost. When transmission resumes, the software automatically reads the untransmitted data from the cache module and continues transmission from the point of interruption until all data transmission is complete. If the cache module is full, it automatically overwrites the oldest untransmitted data (prioritizing fault information and key detection data) to ensure that important data is not lost. The trigger condition for resume interrupted transmission is: failure to receive a response signal from the remote monitoring platform for 3 consecutive times is considered a transmission interruption, triggering resume interrupted transmission. The condition for transmission recovery is: receiving a response signal from the remote monitoring platform for 2 seconds is considered transmission recovery.

[0046] Command interaction mechanism: Control commands issued by the remote monitoring platform (such as calibration trigger, parameter adjustment, fault reset, and self-test trigger) are transmitted to the data transmission unit via 5G or LoRa network. After receiving the command, the unit performs encryption verification. If the verification is successful, the command is converted into a command format that the control analysis unit can recognize and transmitted to the control analysis unit via UART interface. After executing the command, the control analysis unit feeds back the execution result to the data transmission unit, which then encrypts it and transmits it back to the remote monitoring platform, forming a closed loop of command interaction. This ensures the effective execution of remote control commands, with a command interaction latency of ≤50ms, meeting the needs of remote operation and maintenance.

[0047] Unified Data Transmission Format: To ensure compatibility and standardization of data transmission, a unified data transmission format (JSON format) is set, which includes detection data format, fault information format, self-test result format, and command format, as detailed below: Detection data format: {device id Device number, time: detection time, I rms I1: RMS grounding current (A), I3: RMS 3rd harmonic current (A), I5: RMS 5th harmonic current (A), THD: Total harmonic distortion (%), T: Ambient temperature (°C), P: Transformer load (kVA), f: Grid frequency (Hz), cal coeff Calibration coefficient, status: working status (0=normal, 1=warning, 2=alarm)}; Fault information format: {device id Device number, fault time Time of fault occurrence code Fault codes escFault Description, I rms : Effective value of grounding current during fault (A), THD: Total harmonic distortion rate during fault (%) level Fault severity (1 = minor, 2 = severe), handle status Processing status (0 = not processed, 1 = processing, 2 = processed); Self-test result format: {device id Device number, self checktime Self-check time, module status : {detection: detection unit status (0=normal, 1=fault), signal process : Signal processing unit status (0=normal, 1=fault), self cal : Self-test calibration unit status (0=normal, 1=fault), control: Control analysis unit status (0=normal, 1=fault), power: Power supply unit status (0=normal, 1=fault)}}; Command format: {cmd type Command type (1=calibration trigger, 2=parameter adjustment, 3=fault reset, 4=self-test trigger), cmd param Command parameters, cmd time : Instruction issuance time, sign: Instruction signature (encrypted verification code)}.

[0048] Relevant theoretical explanations: The design of the data transmission unit is based on wireless communication theory, data encryption theory, and breakpoint resumption theory. In wireless communication theory, 5G communication is based on Orthogonal Frequency Division Multiplexing (OFDM) technology, which features high bandwidth, low latency, and strong anti-interference capabilities, making it suitable for real-time transmission; LoRa communication is based on spread spectrum communication technology, which features long-distance, low power consumption, and strong anti-interference capabilities, making it suitable for backup transmission and long-distance transmission, and the combination of dual modes can meet different transmission needs; in data encryption theory, the AES-128 encryption algorithm is a symmetric encryption algorithm with fast encryption and decryption speeds and high security, making it suitable for data transmission in industrial fields; CRC32 checksum is based on cyclic redundancy check theory, which can effectively detect bit errors during data transmission and ensure data integrity; breakpoint resumption theory is based on data caching and status recognition, which achieves breakpoint resumption by caching untransmitted data and recognizing the transmission status, thus avoiding data loss. Implementation Instructions: The data transmission unit must be installed synchronously with the overall device. The 5G antenna and LoRa antenna should be installed in designated locations on the device casing. Shielded cables should be used for the antenna leads, with a length controlled between 1 and 2 meters to avoid signal attenuation. The antennas should be installed facing an open area, away from strong interference sources such as transformer windings and switches to ensure stable communication signals. The SIM card for the 5G module must be activated with a dedicated IoT data plan to support high-speed data transmission. Unnecessary functions (such as voice and SMS) should be disabled to reduce power consumption. The communication frequency of the LoRa module must be adjusted to 433MHz to avoid frequency interference with other wireless devices in the substation. The communication address should be uniformly set to the device number to ensure no confusion during network setup.

[0049] During on-site debugging, the stability and switching effect of dual-mode transmission need to be tested. A 5G network interruption (blocking the 5G antenna) should be simulated to check whether automatic switching to LoRa transmission occurs, whether the switching latency is ≤100ms, and whether data is transmitted normally. The data encryption effect should also be tested by capturing transmitted data using a network packet capture tool to check if it is encrypted and whether the decrypted data matches the original. The breakpoint resumption function should be tested by manually interrupting transmission and checking whether data resumes transmission from the breakpoint and whether any data is lost. The command interaction function should be tested by issuing calibration and parameter adjustment commands from the remote monitoring platform and checking whether the control and analysis unit executes them correctly and whether the feedback results are accurate. Simultaneously, the anti-interference capability of the data transmission unit needs to be tested. An electromagnetic interference simulator (EMC-2000) should be used to generate 40dBμV / m electromagnetic interference to test the bit error rate of 5G and LoRa transmissions, ensuring the bit error rate is ≤0.01%. The transmission distance should be tested in an open environment to check whether the LoRa transmission distance is ≥3km and whether the 5G transmission is stable. Finally, power consumption should be tested to ensure that the power consumption of the data transmission unit is ≤2.5W, meeting the overall low power consumption requirements of the device. During later maintenance, it is necessary to regularly check whether the antenna connection is secure, the SIM card's data usage, and whether the encryption key needs to be updated to ensure the long-term stable operation of the data transmission unit. Utilizing 5G and LoRa dual-mode transmission, combined with data encryption and breakpoint resume technology, it solves the problems of single transmission mode, instability, poor security, and limited transmission distance in existing technologies. Unified data format and standardized interfaces improve the device's compatibility and scalability. Industrial-grade hardware is selected, adapted to the harsh environment of substations, with low power consumption to ensure long-term stable operation. It closely matches actual on-site maintenance, enabling remote real-time monitoring and maintenance, and reducing maintenance costs.

[0050] In this embodiment, the power supply unit and the display and alarm unit are the guarantee for the normal operation of the device and the carrier of status feedback. The power supply unit provides stable power to each module, and the display and alarm unit provides real-time feedback on the device's working status, detection data, and fault information. The two work together to ensure reliable operation and convenient maintenance of the device, solving the problems of single power supply method, high power consumption, unstable power supply, unclear display alarms, and unclear fault prompts in the prior art. The dual-mode power supply structure of the power supply unit adopts a combination of substation DC power supply and solar power supply. The DC power supply input voltage range is adapted to the conventional DC power supply specifications of the substation, and the solar power supply part works with the solar charging management module and energy storage module to realize energy storage and stable power supply. The power protection function includes four protections: overcurrent, overvoltage, short circuit, and reverse connection, which can effectively protect each module from damage caused by abnormal power supply. The low power management function can automatically adjust the power supply according to the device's working status, and the device power consumption is reduced to less than 3W in sleep mode. The alarm display unit includes an audible and visual alarm module and a display module. The display module uses an industrial-grade LCD screen to display detection data, operating status, and fault information. The audible and visual alarm module distinguishes different alarm modes according to the severity of the fault, providing intuitive prompts for the fault level. Local operation functions are implemented through dedicated buttons, which can complete parameter debugging, fault reset, and self-test triggering of local maintenance operations.

[0051] The power supply unit is designed to provide stable and reliable power to the detection unit, signal processing unit, self-test calibration unit, control analysis unit, data transmission unit, and display alarm unit. It adapts to different power supply scenarios in substations and features low power consumption, overcurrent, overvoltage, and short-circuit protection to ensure long-term stable operation of each module. Existing technologies often use a single DC220V power supply, which is insufficient for scenarios without DC220V power. Some power supply units lack comprehensive protection functions, making them susceptible to damage from overcurrent and overvoltage. Furthermore, they consume a lot of power, resulting in high long-term operating costs and failing to meet the energy-saving operation and maintenance requirements of substations. This invention's power supply unit adopts a dual-mode power supply design using both DC220V and solar power (solar power supply is a mature technology widely used in various fields). Combined with low-power power management strategies and a comprehensive protection mechanism, it adapts to different power supply scenarios in substations, offering low power consumption and stable power supply. The specific design is as follows: Component Selection and Design: The core hardware of the power supply unit includes a power input module, a power conversion module, a solar charging module, an energy storage module, a power management module, and a protection module. All hardware uses industrial-grade products with standardized parameters. Specific selection and design are as follows: Power Input Module: Supports two input methods: ① DC220V input (existing DC power supply in the substation), input voltage range DC180V~260V, input current ≤1A; ② Solar input, compatible with 12V / 20W solar panels, input voltage range DC10V~15V, input current ≤2A; The input module uses a dual-channel switch (model G6K-2P-Y, the same model as the self-test calibration unit switch), controlled by the power management module. DC220V input is prioritized. When DC220V is interrupted, it automatically switches to solar input with a switching delay ≤10ms to ensure uninterrupted power supply. A fuse (model 0452010.MRL, rated current 10A) is connected in series at the input terminal of the input module to prevent damage from overcurrent. Power Conversion Module: Converts the input power to the standard voltage required by each module, uniformly converting it to three output voltages: ① DC5V, output current ≤3A, used to power the control and analysis unit, data transmission unit, self-test calibration unit, and display and alarm unit; ② DC12V, output current ≤2A, used to power the detection unit and signal processing unit; ③ ±15V, output current ≤1A, used to power the operational amplifier of the signal processing unit. The power conversion module uses an industrial-grade DC-DC converter (model LM2596, DC5V and DC12V output) and a positive and negative power converter (model LM7815 / LM7915, ±15V output), with a conversion efficiency ≥85%, output voltage accuracy ≤±1%, and ripple voltage ≤50mV, ensuring stable power supply and reducing the impact of power supply noise on each module. Solar charging module: Utilizes an industrial-grade solar charging management chip (MPPT7606), supporting Maximum Power Point Tracking (MPPT) with a charging efficiency ≥90%, maximizing solar energy utilization for charging the energy storage module. The module supports both constant current and constant voltage charging modes. The constant current charging current is uniformly 1A, and the constant voltage charging voltage is uniformly 13.8V to prevent overcharging and damage to the energy storage module. The charging module connects to the solar panel 2 and the energy storage module, with the charging status controlled by the power management module. Charging stops when the energy storage module's charge level is ≥90% and starts when the charge level is ≤30%.Energy storage module: A 12V / 10Ah lithium battery (industrial grade, capacity ≥10Ah) is selected as the energy storage carrier when powered by solar energy, ensuring that the device can work continuously for ≥24 hours (full load operation) after a DC220V interruption; the lithium battery has a built-in protection board with overcharge, over-discharge, overcurrent, and short circuit protection functions. The protection thresholds are uniformly set as follows: overcharge protection voltage 14.2V, over-discharge protection voltage 10.8V, overcurrent protection current 5A, and short circuit protection response time ≤1ms; the energy storage module is connected to the power conversion module to provide it with a stable DC input, and at the same time, the power management module detects the power status and feeds it back to the control and analysis unit in real time. Power Management Module: An industrial-grade power management chip (STM32L431, low-power model) is selected. Its core functions are to detect the status of each input power source and the power level of the energy storage module, control power switching and solar charging, monitor the supply current and voltage of each module, and achieve low-power management. The power management module is connected to the control and analysis unit through an I2C interface to transmit power status data (such as input mode, energy storage capacity, output voltage, and output current) in real time. When a power supply abnormality is detected, the protection module is triggered, and fault information is sent to the control and analysis unit to issue a power fault alarm. Protection Module: Integrates four functions: overcurrent protection, overvoltage protection, short circuit protection, and reverse connection protection. ① Overcurrent Protection: The output current is detected by a current sensor (ACS712, range 0~5A, accuracy ±1%). When the output current exceeds the set threshold (DC5V output ≥3.5A, DC12V output ≥2.5A, ±15V output ≥1.5A), overcurrent protection is triggered, cutting off the corresponding output circuit and protecting the module. ② Overvoltage Protection: The output current is detected by a voltage comparator (LM311). ① Overvoltage protection: When the output voltage exceeds the set threshold (DC5V≥5.5V, DC12V≥13.2V, ±15V≥±16.5V), overvoltage protection is triggered, cutting off the input power supply; ② Short circuit protection: Dual protection is adopted with fuse and short circuit detection circuit. When the output circuit is short-circuited, the fuse blows and the short circuit detection circuit is triggered at the same time, cutting off the input power supply. The response time is ≤1ms; ③ Reverse connection protection: A reverse connection diode (model 1N4007) is connected in series at the input terminal of the input module to prevent the module from being damaged by reverse connection of the positive and negative terminals of the power supply.

[0052] The hardware circuit of the power supply unit adopts a shielded design and is connected to the overall grounding system of the device. The grounding resistance is ≤4Ω to suppress electromagnetic interference. Parallel filter capacitors (1000μF+0.1μF) are connected in the circuit to filter out power ripple and ensure stable power supply. All power leads are made of RVVP shielded cables with a cross-sectional area ≥1.5mm². The connection parts adopt crimping technology to ensure good contact and avoid excessive contact resistance that could lead to overheating. Low-power management strategy: To reduce the overall power consumption of the device, the power management module adopts a low-power control strategy, adjusting the power supply mode in real time according to the working status of each module, as follows: ① Normal detection mode: All modules are working normally, the power supply unit is fully powered, and the power consumption is ≤10W; ② Sleep mode: When the device does not detect any abnormality for 10 consecutive minutes and the remote monitoring platform does not issue any instructions, it automatically enters sleep mode, shuts down the 5G module of the data transmission unit, the backlight of the display alarm unit, reduces the main frequency of the control analysis unit (from 168MHz to 84MHz), and reduces the sampling frequency of the detection unit to 100Hz, reducing the power consumption to ≤3W; ③ Wake-up mode: When an abnormal grounding current is detected, an instruction is received from the remote monitoring platform, or the DC220V power supply is restored, the device is automatically woken up and switched to normal detection mode, with a wake-up delay of ≤100ms; ④ Fault mode: When a module fault is detected, the power supply to the faulty module is shut down, and only the core modules (control analysis unit, self-test calibration unit, display alarm unit) are powered to reduce power consumption. At the same time, a fault alarm is issued to prompt maintenance personnel to perform maintenance. Relevant theoretical explanations: The design of the power supply unit is based on power conversion theory, solar charging theory, low-power management theory, and power protection theory. In power conversion theory, the DC-DC converter utilizes pulse width modulation (PWM) technology to convert the input voltage into the required output voltage by adjusting the pulse width, resulting in high conversion efficiency and good stability. In solar charging theory, maximum power point tracking (MPPT) technology tracks the maximum power point of the solar panel in real time, maximizing solar energy utilization and improving charging efficiency. Low-power management theory reduces overall device power consumption by adjusting module operating status, main frequency, and sampling frequency, achieving energy-saving operation and maintenance. Power protection theory prevents module damage and ensures long-term stable operation of the power supply unit through overcurrent, overvoltage, short circuit, and reverse connection protection. Implementation Instructions: The installation of the power supply unit must be synchronized with the overall device. The DC220V input lead is connected to the existing DC power supply cabinet in the substation. The connection points are insulated, and the positive and negative terminals are marked to avoid reverse connection. Solar panel 2 is installed in an open area near the device, facing south, with the tilt angle corresponding to the local latitude (to ensure maximum solar energy reception). The solar lead uses RVVP shielded cable, with a length controlled between 5 and 10 meters to avoid signal attenuation and interference. The energy storage module is installed inside the device and is securely fixed to prevent damage from vibration. The lithium battery charge is checked regularly, and a charge-discharge calibration is performed every 6 months to ensure that the capacity is not less than 80% of the rated capacity.During on-site commissioning, it is necessary to test the dual-mode power supply switching effect, simulate a DC220V power outage, and check whether it automatically switches to solar power, whether the switching delay is ≤10ms, and whether the power supply is stable; test the power conversion accuracy, measure the actual values ​​of each output voltage, and ensure that the error is ≤±1% and the ripple voltage is ≤50mV; test the protection function, simulate overcurrent, overvoltage, short circuit, and reverse connection faults, and check whether the protection module triggers in time, whether it cuts off the corresponding circuit, and whether the module is intact; test the low-power mode, check the trigger conditions and power consumption of the sleep mode, and the response delay of the wake-up mode, to ensure that it meets the design requirements; test the solar charging function, and check whether the charging module charges normally in a well-lit environment, whether the charging efficiency is ≥90%, and whether the energy storage module's capacity increases normally. In summary, the adoption of DC220V and solar dual-mode power supply adapts to different power supply scenarios in substations, overcoming the limitations of a single power supply method; it integrates a comprehensive protection mechanism to prevent module damage; it adopts a low-power management strategy to reduce operating costs; all hardware parameters are unified and industrial-grade, ensuring stable and reliable power supply, and conforming to actual on-site operation and maintenance.

[0053] Among them, the display alarm unit is designed to display the device's working status, detection data, self-test results and fault information in real time. When an abnormality or fault is detected, it will issue an audible and visual alarm to prompt maintenance personnel to handle it in a timely manner. It also supports local parameter debugging and fault diagnosis, solving the problems of unclear display interface, incomplete data display, unclear fault prompts and single alarm methods in the existing technology. The specific design of the display alarm unit of this invention is as follows: Software selection and design: The core hardware of the display alarm unit includes an LCD display module, an audible and visual alarm module, a local button module, and an interface module. All hardware uses industrial-grade products, adapted to harsh outdoor environments, and with standardized parameters. The specific selection and design are as follows: LCD display module: An industrial-grade LCD screen (model LCD12864, 128×64 dot matrix) is uniformly selected. The screen size is 3.5 inches, and a backlight design is adopted (backlight brightness is adjustable, with 5 levels), adapted to strong outdoor light environments, and can be clearly displayed even under strong light. The display screen supports Chinese display and can simultaneously display detection data, working status, fault information, and self-test results. The display refresh rate is consistent with the operating cycle of the control and analysis unit (10μs), with no afterimages or flicker. The power supply voltage of the display screen is DC5V, and the power consumption is ≤500mW. It is connected to the control and analysis unit through a UART interface to receive display data. The display screen shell adopts a waterproof design (IP67 protection level) to avoid damage from rain and dust. Audible and visual alarm module: Integrates LED alarm light and buzzer to achieve dual audible and visual alarm. Alarm modes are differentiated according to fault level: ① Minor fault (early warning): Green LED flashes (1 flash / second), buzzer sounds intermittently (1 second on, 1 second off); ② Severe fault (emergency alarm): Red LED is constantly on, buzzer sounds continuously; ③ Power failure: Yellow LED flashes (2 flashes / second), buzzer sounds intermittently (0.5 seconds on, 0.5 seconds off); ④ Self-test fault: Blue LED flashes (3 flashes / second), buzzer does not sound (only a warning, no alarm). The LED alarm light uses high-brightness industrial-grade LEDs (5mm diameter), with a brightness ≥500cd / m². The buzzer uses an industrial-grade active buzzer (model HRB-PB1206P), with a volume ≥85dB. The power supply voltage is DC5V, powered separately by the power supply unit. The alarm trigger signal is sent by the control and analysis unit through the GPIO interface, and the alarm response time is ≤10ms.Local Button Module: Features four local buttons with unified functions: ① Confirm Button: Confirm parameter adjustment, fault reset, and self-test trigger; ② Up / Down Buttons: Adjust display backlight brightness, switch display interfaces, and adjust parameter values; ③ Menu Button: Access parameter setting interface, self-test interface, and fault record interface. The buttons are industrial-grade tactile buttons (model TS-1102), waterproof, dustproof, and wear-resistant, with a press life of ≥1 million cycles. The buttons connect to the control and analysis unit via a GPIO interface. Pressing a button sends a control signal to the control and analysis unit to execute the corresponding function. The buttons provide clear feedback, and the pressing pressure is uniformly 500g. Interface Module: Uses an industrial-grade UART interface to connect to the control and analysis unit, transmitting display data and control signals. It also integrates an RS232 interface (implemented via an interface conversion module MAX3232) for connecting to a local maintenance terminal to read test data, fault records, and debug parameters. The interface uses an aviation plug to ensure a secure connection and good contact, and the interface area is waterproofed. The hardware circuit of the display alarm unit is connected to the overall grounding system of the device, with a grounding resistance ≤4Ω to suppress electromagnetic interference. A current-limiting resistor is connected in series in the circuit to protect the LED alarm light and buttons, ensuring long-term stable operation of the module. Display interface design: The display interface adopts a modular design, divided into 4 interfaces, which can be switched via local buttons. The interface is clear, the data is comprehensive, and the display format is unified, as follows: Main display interface (default interface): Displays core detection data and working status, including device number, detection time, effective value of grounding current (I_rms), ambient temperature (T), power grid frequency (f), working status (normal / warning / alarm), power supply method (DC220V / solar), and energy storage capacity (%). The data display accuracy is consistent with the detection accuracy (current retains 3 decimal places, temperature retains 1 decimal place, and frequency retains 1 decimal place). The working status is marked with different colors (normal = green, warning = yellow, alarm = red). Data Details Interface: Displays detailed test data, including the fundamental current RMS value (I1), the 3rd harmonic current RMS value (I3), the 5th harmonic current RMS value (I5), the total harmonic distortion (THD), the transformer load (P), and the calibration coefficient (cal). coeffThe interface includes a sampling frequency display, standardized data display, and clearly labeled units, facilitating detailed data viewing and core grounding status analysis for maintenance personnel. The fault record interface displays historical fault information, including fault occurrence time, fault code, fault description, fault level, and handling status. It can store the most recent 100 fault records, supporting up and down arrow keys for switching between views. Fault records can be exported from the local maintenance terminal for easy fault tracing and analysis. The parameter setting interface supports local parameter debugging, including calibration coefficient adjustment, fault threshold adjustment, self-test cycle adjustment, backlight brightness adjustment, and sampling frequency adjustment. After parameter adjustment, the confirmation key must be pressed to save. Saved parameters are automatically synchronized to the FRAM of the control and analysis unit and are not lost upon power failure. The parameter adjustment range is consistent with the previous description (e.g., self-test cycle 1 day to 1 month, backlight brightness level 5). Alarm Control Logic: The alarm logic of the display alarm unit is controlled by the control analysis unit. When the control analysis unit determines a fault or abnormality, it sends an alarm trigger signal. After receiving the signal, the display alarm unit triggers the corresponding audible and visual alarm according to the fault level, and simultaneously displays the fault information (fault code, fault description, and fault occurrence time) on the LCD screen. After the maintenance personnel discover the alarm, they can press the confirmation key on the local button to stop the buzzer from sounding (the LED light continues to flash / remains on, indicating that the fault has not been handled). After the fault is handled, pressing the fault reset key will stop the alarm, and the display interface will return to normal. At the same time, the fault handling status will be fed back to the control analysis unit to update the fault record. If the fault remains unhandled, the audible and visual alarm will continue (severe fault) or intermittently (minor fault, power failure) until the fault is handled or manually stopped. Self-test faults only display fault information and do not trigger the buzzer alarm, prompting maintenance personnel to perform self-test troubleshooting.

[0054] Relevant theoretical explanation: The design of the display alarm unit is based on human-computer interaction theory and alarm control theory. Human-computer interaction theory enhances the user experience for maintenance personnel through a clear display interface and convenient button operation, facilitating quick data viewing and troubleshooting. Alarm control theory differentiates fault levels and employs different audible and visual alarm methods to clearly define the urgency of the fault, prompting maintenance personnel to prioritize urgent faults, preventing escalation, and improving maintenance efficiency. Implementation Instructions: The display alarm unit is installed on the front of the device casing for easy viewing and operation by maintenance personnel. The display screen faces an easily observable direction, and the buttons are located below the screen in a logical and easy-to-use layout. During on-site commissioning, the refresh rate of the display interface needs to be tested to ensure accurate data display, no ghosting, and no flickering. The backlight brightness should be adjustable to adapt to strong outdoor lighting conditions. The audible and visual alarm functions should be tested, simulating different levels of faults to check if the alarm method matches the settings, if the alarm response time is ≤10ms, and if the volume and brightness meet the standards. The local button functions should be tested to check the button feedback, accuracy of function execution, and whether parameter adjustments can be saved correctly. The interface functions should be tested by connecting to the local maintenance terminal to check if data can be read normally, fault records can be exported, and debugging parameters can be adjusted. During subsequent maintenance, the display screen surface should be cleaned regularly to prevent dust from affecting the display effect. The buttons should be checked for smooth operation; if any are stuck, they should be replaced promptly. The LED alarm lights and buzzers should be checked for proper functioning; if any are damaged, they should be replaced promptly to ensure the long-term stable operation of the display alarm unit. It adopts a backlit LCD screen, which is suitable for outdoor strong light environments; it has dual sound and light alarms, and the fault level is clearly distinguished, which is convenient for maintenance personnel to quickly identify; the local buttons support parameter debugging and fault handling, making it easy to operate; the display interface is modular, the data is comprehensive, and the format is uniform, which meets the actual needs of on-site maintenance and solves various defects in the display alarms of existing technologies.

[0055] The specific usage method of the above technical solution is as follows: Step 1: Device Installation. Install the detection device in a suitable location near the transformer core grounding lead, ensuring it is securely installed and free from looseness to avoid interference from transformer vibration. After installation, connect the cables of each unit sequentially, ensuring correct connections and good contact. Simultaneously connect the antenna required for data transmission and reliably connect the device casing to the grounding system to suppress electromagnetic interference and ensure detection accuracy. Protect all interfaces and cables during installation to avoid damage. Step 2: Device Power-On Initialization. Based on the site power supply conditions, select either substation DC power or solar power to connect the device. After power is connected, the device will automatically start and enter the initialization state. During initialization, it will automatically complete the startup checks and parameter readings of each unit. Maintenance personnel should observe the status prompts on the display units and wait for initialization to complete. If alarm messages are displayed during initialization, check the cable connections and power supply for proper functioning. After confirming everything is correct, re-power on and initialize.

[0056] Step 3: Self-test Calibration Operation. After the device initialization is complete, it will automatically initiate a full-module self-test to comprehensively check the working status of all modules, including the detection unit, signal processing unit, and control analysis unit. Maintenance personnel need to check the self-test result prompts on the display unit. If the self-test shows no abnormalities, the device will automatically enter the detection state; if the self-test indicates a fault, the corresponding module's connection and working status must be checked according to the displayed fault information. After troubleshooting, the self-test will be restarted until it passes. In daily use, the device will automatically complete the calibration operation according to a preset cycle without manual intervention. If manual calibration is required, the calibration process can be triggered via local button operation. During calibration, the device maintains normal detection status, and automatically updates calibration parameters and displays the calibration result after calibration is completed. Step 4: Normal Detection and Monitoring. After the self-test calibration passes, the device enters normal detection status, automatically acquiring the transformer core grounding current signal. After signal processing and data analysis, the detection results and device working status are displayed on the display unit in real time. Maintenance personnel can view relevant detection information through the display unit and can also receive detection data and status information transmitted by the device through a remote monitoring platform to achieve remote monitoring. During the testing process, if the device detects an abnormal core grounding current or abnormal operating condition, it will automatically trigger an audible and visual alarm and transmit the fault information to the remote monitoring platform. Maintenance personnel must promptly check the alarm prompts, distinguish the fault level, and make a preliminary judgment on the fault type and severity based on the fault information. Step 5: Fault Handling Operation. When the device issues a fault alarm, maintenance personnel must first confirm the alarm information via local buttons, stop the alarm sound, and then troubleshoot the cause of the fault based on the fault description displayed on the display unit. If it is a minor abnormality, it can be troubleshooted by adjusting the device's installation position and checking cable connections; if it is a serious fault, the device testing must be suspended, the power turned off, and a comprehensive inspection and repair of the corresponding module must be performed. After the fault is completely eliminated, the device should be powered on again, and normal testing should resume after completing the self-test calibration. After the fault handling is completed, the fault record must be reset via local buttons or remote commands to update the device's operating status and ensure normal device operation. Step 6: Daily Maintenance Operation: Regularly perform daily inspections of the device, cleaning dust and debris from the display unit surface and the device casing to ensure clear display and no damage to the casing. Check that all cable connections are secure, the antenna is intact, and the grounding connection is reliable. If any loose cables or damaged antennas are found, tighten or replace them immediately. Regularly check the power supply. If solar power is used, check the cleanliness and operational status of the solar energy collection components to ensure proper energy storage. If DC power is used, check the stability of the power connection. Simultaneously, view the device parameters using the local buttons and adjust the relevant parameter settings as needed based on actual site requirements to ensure the device's detection accuracy and operational stability.Before the device is to be shut down for an extended period, the power must be turned off, all cables and antenna connections must be disconnected, and the device must be stored properly to avoid damage. When restarting, the above steps must be followed to complete the installation, power-on, and self-calibration before it can be put into normal testing.

[0057] In summary, this invention employs a dual-sensor fusion acquisition structure combined with a multi-level signal processing design, which effectively suppresses electromagnetic interference, ensures the accuracy and integrity of the core grounding current signal acquisition, and avoids signal distortion affecting the detection results. The self-testing and calibration unit can realize real-time self-testing and periodic automatic calibration of each module without frequent manual intervention, enabling timely detection and precise location of module faults, reducing manual maintenance workload, and ensuring the detection accuracy of the device during long-term operation. The dual-core architecture of the control and analysis unit, combined with a dynamic threshold correction design, can quickly process detection data, accurately distinguish different operating conditions and fault severity of core grounding, effectively avoid false alarms and missed alarms, and provide maintenance personnel with reliable fault judgment basis. The data transmission unit adopts dual-mode anti-interference transmission combined with data encryption and breakpoint resume technology to ensure the stability, security, and integrity of detection data and fault information transmission, adapting to different transmission distance requirements of substations, and facilitating remote monitoring by maintenance personnel. The power supply unit... The dual-mode power supply structure, coupled with comprehensive power protection and low-power management functions, can adapt to different power supply scenarios in substations, ensuring continuous and stable operation of the device while reducing operating energy consumption and maintenance costs. The display and alarm unit, combining display, audible and visual alarms, and local operation functions, can provide real-time feedback on the device's operating status and detection information, intuitively indicating the fault level, facilitating maintenance personnel to quickly detect, troubleshoot, and handle faults, and improving maintenance efficiency. The device adopts an integrated metal shielding enclosure and grounding treatment, further enhancing its anti-interference capability. Its compact structure and convenient installation make it suitable for harsh outdoor operating conditions in substations. The operation process is simple and easy to understand, requiring no professional technicians to complete daily operations and basic maintenance. Overall, it improves the intelligence and automation level of transformer core grounding current detection, enabling real-time and accurate monitoring of the transformer core grounding status, timely warning of potential faults, prevention of fault escalation, ensuring safe and stable operation of the transformer, extending the transformer's service life, and reducing substation maintenance costs and safety risks.

[0058] Although embodiments of the invention have been shown and described, the scope of the invention will be defined by the appended claims and their equivalents by those skilled in the art.

Claims

1. A transformer core grounding current detection device, characterized in that, The ground current detection device (1) includes a detection unit, a signal processing unit, a self-test calibration unit, a control analysis unit, a data transmission unit, a power supply unit, and a display alarm unit. Each unit is integrated and packaged and uses an integrated metal shielding structure with grounding treatment to suppress electromagnetic interference. The detection unit adopts a dual-sensor fusion structure of Rogowski coil and Hall sensor, and combines an adaptive weighted fusion algorithm to realize full-range acquisition of iron core grounding current; The signal processing unit adopts a multi-stage processing structure of adaptive filtering, programmable gain amplification, isolation conversion and signal conditioning to eliminate interference and complete signal conversion; The self-test calibration unit integrates a built-in standard signal source to enable real-time self-testing, periodic automatic calibration, and fault location for each module. The control and analysis unit adopts a dual-core architecture and integrates a dynamic threshold correction algorithm to distinguish different working conditions and fault severity of iron core grounding. The data transmission unit adopts a dual-mode anti-interference transmission method and combines data encryption and breakpoint resume technology to achieve data transmission; the power supply unit adopts a dual-mode power supply structure and integrates power protection and low power management functions. The display alarm unit combines display, audible and visual alarms, and local operation functions to achieve status feedback and local operation and maintenance.

2. The transformer core grounding current detection device according to claim 1, characterized in that, The Rogowski coil adopts a hollow toroidal structure with 1000~2000 turns and an inner diameter of 50~80mm, and is wound with fine-diameter enameled copper wire; the Hall sensor is a closed-loop Hall current sensor. The adaptive weighted fusion algorithm dynamically adjusts the weighting coefficients of the two signals by calculating the noise variance of the signals acquired by the two sensors in real time, and acquires the grounding current across the entire range.

3. The transformer core grounding current detection device according to claim 2, characterized in that, The signal processing unit includes a signal conditioning module, an adaptive filtering module, a programmable gain amplifier, and an isolation conversion module. The signal conditioning module uses resistors to convert current and voltage signals, and works with a high input impedance operational amplifier to form a voltage follower, achieving signal impedance matching and distortion-free transmission. The adaptive filtering module uses at least a 20th-order active filter structure, and the amplification factor of the programmable gain amplifier is adjustable in the range of 1 to 1000 times to adapt to signals of different amplitudes. The isolation conversion module uses at least a 16-bit AD converter.

4. The transformer core grounding current detection device according to claim 3, characterized in that, The built-in standard signal source of the self-test calibration unit can generate a standard current signal in the range of 0~10A and 50Hz~1kHz. During the calibration process, the calibration coefficient is obtained by calculating the ratio of the standard signal value to the actual detection value of the device, and the detection data is corrected by using the calibration coefficient.

5. A transformer core grounding current detection device according to claim 4, characterized in that, The dual-core architecture of the control and analysis unit consists of a main control chip and a dedicated data processing chip. The main control chip is a 32-bit microcontroller, and the dedicated data processing chip is a high-speed digital signal processing chip. The dynamic threshold correction algorithm combines three key operating parameters—ambient temperature, transformer load, and power grid frequency—and adjusts the fault judgment threshold in real time through a preset correction formula.

6. A transformer core grounding current detection device according to claim 5, characterized in that, The dual-mode anti-interference transmission method of the data transmission unit adopts a combination of high-speed wireless communication and long-distance wireless communication to realize real-time data transmission and long-distance backup transmission; the data encryption technology adopts hardware encryption; the dual-mode switching is automatically triggered according to the signal strength of the main communication mode.

7. A transformer core grounding current detection device according to claim 6, characterized in that, The dual-mode power supply structure of the power supply unit adopts a combination of substation DC power supply and solar power supply. The DC power supply input voltage range is adapted to the conventional DC power supply specifications of the substation. The solar power supply part works with the solar charging management module and the energy storage module to realize energy storage and power supply. The power protection function includes four protections: overcurrent, overvoltage, short circuit, and reverse connection. The low power management function automatically adjusts the power supply according to the working status of the device.

8. A transformer core grounding current detection device according to claim 7, characterized in that, The alarm display unit includes an audible and visual alarm module and a display module. The display module uses an LCD screen to display detection data, operating status, and fault information. The audible and visual alarm module distinguishes different alarm modes according to the severity of the fault. Local operation functions are realized through dedicated buttons to complete parameter debugging, fault reset, and self-test triggering of local maintenance operations.

9. A method for detecting transformer core grounding current, characterized in that, Specifically, the following steps are included: After the device is powered on and initialized, it first completes a full module self-test. If there are no faults, it enters the testing state. The iron core grounding current signal is collected by dual sensor fusion, processed by the signal processing unit, and then transmitted to the control and analysis unit. The control analysis unit analyzes and calculates the signal, and determines the grounding condition and fault level by combining dynamic thresholds; Simultaneously complete the encrypted transmission of detection data and fault information, and provide real-time status feedback; Regular automatic calibration is performed to continuously and cyclically achieve real-time detection of grounding current and fault early warning.

Images