Power secondary circuit multi-parameter intelligent detection system and method based on edge calculation

Abstract

The invention provides an electric power secondary circuit multi-parameter intelligent detection system and method based on edge computing, the system comprises a composite sensing array, a self-adaptive probe system and an edge computing system, the composite sensing array comprises a plurality of sensors, the self-adaptive probe system comprises a multi-direction adjustable probe array, and the multi-direction adjustable probe array is connected with the edge computing system. The tail end of each probe in the multidirectional adjustable probe array is embedded into the permanent magnet, and the multidirectional adjustable probe array transmits the acquired electrical signals of the terminal strip to be tested to one or more of the plurality of sensors through the multiplexer; and the composite sensing array and the adaptive probe system are connected with the edge computing system, so that the detection efficiency is effectively improved.

Description

Technical Field

[0001] This application relates to the field of power system secondary equipment testing technology, and in particular to a multi-parameter intelligent testing system and method for power secondary circuits based on edge computing, specifically to a live-line testing device and method for secondary circuit terminal blocks in hydropower plants. Background Technology

[0002] Traditional testing methods for secondary circuit terminal blocks in hydropower plants have numerous problems. For example, in one hydropower plant, the use of separate testing tools (such as multimeters, insulation meters, and phase meters, which had to be used alternately) resulted in a cumbersome testing process with each tool having multiple functions, averaging over 2 minutes per point and severely impacting testing efficiency. Furthermore, live-line testing carries the risk of accidentally touching terminals not being tested. A 2023 accident report from a provincial power grid showed that traditional testing methods accounted for 17% of electric shock accidents. In addition, the results of traditional testing methods cannot be exchanged with the equipment management system in real time, creating data silos and hindering intelligent equipment management. Summary of the Invention

[0003] This application aims to at least partially address one of the technical problems in the related art.

[0004] The first aspect of this application proposes a multi-parameter intelligent detection system for power secondary circuits based on edge computing, comprising: a composite sensor array, an adaptive probe system, and an edge computing system. The composite sensor array includes multiple sensors, and the adaptive probe system includes a multi-directional adjustable probe array. The ends of each probe in the multi-directional adjustable probe array are embedded with permanent magnets. The multi-directional adjustable probe array transmits the electrical signals of the terminal block under test collected by the probe array to one or more of the multiple sensors through a multiplexer. Both the composite sensor array and the adaptive probe system are connected to the edge computing system.

[0005] To achieve the above objectives, a second aspect of this application proposes a multi-parameter intelligent detection method for power secondary circuits, the method comprising the following steps:

[0006] The adaptive probe system obtains the contact impedance between each probe and the terminal of the terminal block to be tested, and adjusts the impedance matching between each probe of the adaptive probe system and the terminal of the terminal block to be tested when the contact impedance does not meet the preset impedance threshold.

[0007] The composite sensor array simultaneously acquires multiple electrical parameters of the terminal block under test;

[0008] Feature extraction is performed on the multiple electrical parameters to obtain feature values ​​that reflect the operating status of the power equipment related to the terminal block;

[0009] The feature values ​​are input into a deep learning model to identify whether there are abnormal patterns in power equipment and the identification results are presented in a visual manner.

[0010] The intelligent multi-parameter detection system and method for power secondary circuits based on edge computing provided in this application simultaneously acquires multiple electrical parameters of terminal blocks through a composite sensor array, and provides physical connection and impedance matching with the terminal blocks through an adaptive probe system to ensure the stability of signal transmission. The composite sensor array can simultaneously acquire multiple electrical parameters of the terminal blocks, and the edge computing system can ensure that the detection data can be processed and fed back in a timely manner, improving the real-time performance and effectiveness of the detection, and effectively improving the detection efficiency.

[0011] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0012] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0013] Figure 1 A block diagram of a multi-parameter intelligent detection system for power secondary circuits based on edge computing, provided in an embodiment of this application;

[0014] Figure 2 This is a schematic diagram of the structure of a multidirectional adjustable probe array provided in an embodiment of this application;

[0015] Figure 3 An example diagram of an AR interface provided in an embodiment of this application;

[0016] Figure 4 A flowchart illustrating a multi-parameter intelligent detection method for power secondary circuits provided in this application embodiment;

[0017] Figure 5 This is a flowchart illustrating an abnormal pattern recognition component provided in an embodiment of this application. Detailed Implementation

[0018] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0019] The following description, with reference to the accompanying drawings, describes an embodiment of the intelligent multi-parameter detection system and method for power secondary circuits based on edge computing.

[0020] Figure 1 A block diagram of a multi-parameter intelligent detection system for power secondary circuits based on edge computing, as provided in an embodiment of this application, is shown below. Figure 1 As shown, the detection system includes a composite sensor array, an adaptive probe system, an edge computing system, a power supply module, and an AR display module. All four components—the composite sensor array, the adaptive probe system, the power supply module, and the AR display module—are connected to the edge computing system. The composite sensor array includes multiple sensors, and the adaptive probe system includes a multi-directional adjustable probe array. The ends of each probe in the multi-directional adjustable probe array are embedded with permanent magnets. The multi-directional adjustable probe array transmits the electrical signals collected from the terminal block under test to one or more of the multiple sensors via a multiplexer. The composite sensor array and the adaptive probe system are connected via a flexible printed circuit board (FPC).

[0021] This embodiment achieves magnetic contact between the probe and the terminal by embedding a permanent magnet at the probe tip. The permanent magnet forms a magnetic attraction with the magnetic material (such as ferrous terminals) in the terminal block. The probe transmits the electrical signals of the terminal block to the back-end sensor. The sensor receives the raw electrical signals transmitted by the probe and converts them into processable digital or analog signals. All raw electrical signals collected by the probes are centrally transmitted to different sensors in the composite sensor array via a multiplexer (MUX). The edge computing system controls the channel switching of the multiplexer (MUX) according to the detection requirements, routing the signals of specific probes to the corresponding sensors. The probes and sensors are connected via an FPC to ensure low-loss signal transmission. The composite sensor array simultaneously collects multiple electrical parameters of the terminal block, and the adaptive probe system provides physical connection and impedance matching with the terminal block to ensure the stability of signal transmission. The power module supplies power to the entire system, and the AR display module is used to visualize the data analysis results of the edge computing system. The composite sensor array can simultaneously collect multiple electrical parameters of the terminal block, and the edge computing system can ensure that the detection data can be processed and fed back in a timely manner, improving the real-time performance and effectiveness of the detection, and effectively improving the detection efficiency.

[0022] In some embodiments, the composite sensor array includes a temperature sensor, a Hall sensor, a voltage divider insulation detection module, and a high-frequency sampling module. The temperature sensor is used to acquire temperature parameters, the Hall sensor is used to acquire current parameters, the voltage divider insulation detection module is used to acquire voltage parameters and insulation resistance to ground, and the high-frequency sampling module is used to acquire open-circuit characteristic signals of the secondary circuit of the current transformer (CT).

[0023] The voltage divider insulation detection module (e.g., 0-2000MΩ range) in this embodiment has dual functions. One function is to connect a high-precision voltage sensor (differential amplifier + precision resistor network) in parallel in the voltage divider circuit to directly measure the voltage value of the terminal block and obtain voltage data. The other function is to accurately detect the insulation resistance of the terminal block to ground (0-2000MΩ) over a wide range through the voltage divider principle, ensuring effective monitoring of the equipment's insulation performance.

[0024] In this embodiment, the correspondence between the multiple sensors and the multiple probes in the multidirectional adjustable probe array is not a fixed one-to-one correspondence. Instead, it is achieved through signal distribution by a multiplexer (MUX) to realize centralized signal management and dynamic switching. Specifically, the raw signals collected by all probes are centrally transmitted to different sensors in the composite sensor array via the multiplexer (MUX). The edge computing system controls the channel switching of the multiplexer (MUX) according to the detection requirements, routing the signal of a specific probe to the corresponding sensor. For example: when probes 1 and 2 contact the voltage circuit of the terminal block, the MUX switches the signal to the voltage divider insulation detection module; when probes 3 and 4 contact the current circuit, the MUX switches the signal to the Hall sensor. The cooperation method between the probes and multiple sensors in this solution needs to be distinguished according to the sensor type. Not all sensors contact the terminal block through the probes. Only temperature sensors are attached to the ends or sides of the probes, while other sensors are not.

[0025] Therefore, by attaching a temperature sensor to the surface of the terminal block, the temperature data of the terminal block is collected; a Hall sensor is connected in series in the current loop of the terminal block, and the current data of the terminal block is collected; a voltage divider insulation detection module is connected in parallel to the terminal block, and the voltage and insulation resistance data of the terminal block are collected simultaneously by the voltage divider insulation detection module; through the synchronous acquisition (time domain alignment) of the Hall sensor acquisition (current signal) and the voltage divider insulation detection module (voltage signal), the edge computing system uses the time difference between the two to calculate the phase angle and obtain the phase data. The high-frequency sampling module (1MHz ADC) captures the open-circuit characteristic signal of the CT (current transformer) secondary circuit through non-contact coupling (such as Rogowski coil), because the CT secondary open circuit generates a high-frequency transient voltage signal (usually in the kHz to MHz range). This embodiment uses the high-frequency sampling module to capture the CT secondary open-circuit characteristic; the auxiliary signal of the high-frequency sampling module can also provide higher-precision time domain data support for phase calculation. The edge computing system synchronously processes multiple electrical parameters acquired by the composite sensor array, including calculating the phase difference based on time-domain aligned current and voltage signals, jointly assessing the aging status of hydroelectric equipment based on the acquired insulation resistance and temperature, and detecting transient characteristics of CT secondary open circuit based on high-frequency signals acquired by the high-frequency sampling module.

[0026] As an example, the temperature sensor selected is the PT1000 temperature sensor (±0.5℃ accuracy), which utilizes the temperature characteristics of platinum resistance thermometers, has high accuracy and stability, and can monitor the temperature changes of the terminal block in real time.

[0027] As an example, the Hall sensor selected is a high-precision Hall sensor (±0.1% accuracy), which uses advanced magnetoelectric conversion technology and high-precision signal processing circuitry to accurately measure current and provide accurate current data.

[0028] In summary, this embodiment solves the problem of fragmented functions in traditional detection tools by using a composite sensor array. It can simultaneously detect five electrical parameters: voltage, current, insulation resistance, temperature, and phase. At the same time, it can acquire multiple electrical parameters in a single detection, thereby providing more comprehensive data for the health assessment of hydroelectric equipment.

[0029] In some embodiments, the adaptive probe system further includes a contact state self-test circuit and an electromagnetic drive unit. The contact state self-test circuit is used in conjunction with the edge computing system to obtain the contact impedance between each probe and the terminals of the terminal block to be tested. The electromagnetic drive unit includes a PWM voltage regulation module and an electromagnet independently configured for each probe. The electromagnet includes an iron core and an electromagnet coil. The edge computing system is also used to control the energizing direction and current intensity of the electromagnet through the PWM voltage regulation module based on the obtained contact impedance between each probe and the terminals of the terminal block to be tested, so as to drive the permanent magnet at the end of the probe to realize the axial extension and radial deflection of the probe. In this embodiment, the electromagnet is a miniature electromagnet.

[0030] The contact state self-test circuit in this embodiment is used to monitor the contact impedance between the probe and the terminal block in real time to ensure signal transmission stability; such as Figure 2 As shown, the electromagnetic drive unit equips each probe with an independent electromagnet (operating voltage 5VDC), and each probe's electromagnet is independently controlled by the edge computing system. Based on feedback signals from the contact status self-test circuit, the edge computing system adjusts the coil current intensity through the electromagnetic drive unit, dynamically strengthening or weakening the magnetic attraction. This, in turn, drives the probe to extend, retract, or deflect via electromagnetic force. In low-impedance environments (such as clean terminals), the coil current is reduced, decreasing the magnetic attraction and allowing for light probe contact, avoiding excessive pressure. In high-impedance environments (such as oxidized surfaces), the coil current is increased, increasing the magnetic attraction and pressing the probe firmly into contact, reducing contact impedance; thus achieving impedance matching. This solution dynamically adjusts the contact pressure between the probe and the terminal block by real-time monitoring of the contact impedance signal value, ensuring stable contact within a safe range while preventing poor contact or equipment damage due to excessive or insufficient pressure.

[0031] Furthermore, such as Figure 2 As shown, the multidirectional adjustable probe array includes eight probes arranged in a star shape. Each probe is made of tungsten steel and has a gold-plated surface.

[0032] This embodiment features a star-shaped probe array, with two probes in each of the top, bottom, left, right, and diagonal directions, achieving an 8-way adjustable probe array. This array can cover multi-angle contact requirements, adapt to terminal blocks of different specifications and shapes, and improve the contact accuracy between the probes and terminals. The probes are made of tungsten carbide (e.g., 0.8mm in diameter), which has high hardness and wear resistance. The tungsten carbide probes are gold-plated (e.g., with a thickness ≥2μm) to reduce contact resistance.

[0033] In this embodiment, the multi-directional adjustable probe array means that the orientation of each probe can be independently adjusted by its corresponding electromagnetic drive unit. The electromagnetic drive unit can achieve axial extension and radial deflection of the probes. Axial extension means that each probe can be driven by an electromagnet drive unit to extend or retract along its axial direction (e.g., vertically), adjusting the contact depth between the probe and the terminal block. Radial deflection means that a permanent magnet is embedded at the end of each probe; by controlling the direction and intensity of the current in the electromagnet coil, the probe can deflect in the up / down, left / right, or diagonal directions, achieving multi-angle contact. The eight probes are arranged in a star-shaped pattern, forming an omnidirectional contact capability covering 360°. By independently adjusting the orientation and position of each probe, it can adapt to different shapes (e.g., tilted, irregularly shaped terminals) and sizes (3-40mm). 2 Terminal blocks. Example 1: If there are tilted terminals in a certain area of ​​the terminal block, the probes in the corresponding direction of the multi-directional adjustable probe array can be deflected at a certain angle to ensure close contact. Example 2: When inspecting densely packed small-sized terminal blocks, the probes avoid adjacent terminals by radial deflection to prevent accidental contact.

[0034] The probe functions corresponding to the various sensor types in this embodiment will be explained below.

[0035] Voltage divider insulation testing module: Probes 1 and 2 (up and down direction) measure voltage and insulation resistance to ground.

[0036] Hall sensor: Probes 3 and 4 (left and right direction) are connected in series in the current loop to measure the current intensity.

[0037] Temperature sensor: Probes 5 and 6 (diagonally) are attached to the surface of the terminal block to monitor temperature changes in real time.

[0038] High-frequency sampling module: Non-contact coupling, capturing high-frequency transient signals of the CT secondary circuit through Rogowski coil.

[0039] Probes 7 and 8 are designed as redundancy: when probes 1-4 fail, they temporarily replace probes 1-4 for measuring voltage or current.

[0040] In summary, this embodiment solves the problem of poor adaptability of traditional testing tools to terminal blocks of different specifications by using an adaptive probe system. It develops an omnidirectional adaptive probe system that is compatible with terminal blocks of different sizes, thereby improving the versatility and accuracy of the testing system.

[0041] Furthermore, the contact status self-test circuit includes multiple probe voltage acquisition circuits and a multiplexer. The multiple probe voltage acquisition circuits transmit the voltage difference signal to the edge computing system through the multiplexer. The probe voltage acquisition circuit includes a constant current source, a differential amplifier, and an analog-to-digital converter. The constant current source is connected to the distal end A and distal end B of the probe through two current excitation lines, respectively. The differential amplifier is connected to the contact point C and contact point D of the probe through two voltage detection lines, respectively, to acquire and amplify the analog voltage difference between the two contact points of the probe. The analog-to-digital converter converts the analog voltage difference into a digital voltage difference. The edge computing system is used to obtain the contact impedance between each probe and the terminal of the terminal block to be tested based on the voltage difference signal and the current parameters of the constant current source.

[0042] The contact status self-test circuit of this scheme uses a four-wire measurement method (Kelvin connection) to eliminate wire resistance interference. A constant current source (1mA) is applied to each probe, and the voltage difference across the probe is measured by a differential amplifier. The contact impedance is obtained through an edge computing system. If an abnormal contact impedance (such as poor contact) is detected, the electromagnetic drive unit is triggered to adjust the probe pressure or position. The core of the four-wire measurement method is to eliminate wire resistance interference by separating the current excitation path and the voltage measurement path.

[0043] As an example, a constant current source is used to provide a stable current I (e.g., 1mA), connected to the distal end (A, B) of the probe via two wires (current excitation lines). The probe contact points (C, D) are the actual locations where the probe contacts the terminal block, with a contact impedance of R_contact. The differential voltage measurement circuit, i.e., a differential amplifier, is directly connected to the probe contact points (C, D) via two additional wires (voltage detection lines) to measure the voltage difference V. diff The current excitation lines (A, B) and voltage detection lines (C, D) are physically separated at the probe ends to ensure that the voltage measurement reflects only the contact impedance R_contact, ignoring the wire resistance R_wire. The constant current source is implemented using an operational amplifier (Howland current pump) with an output accuracy of ±0.1%. A differential amplifier (e.g., AD620, adjustable amplification factor) is used to amplify small voltage differences. The analog-to-digital converter (ADC) is a 24-bit Σ-Δ ADC (e.g., ADS1248) that converts analog voltages into digital signals. A multiplexer (MUX) is used to switch the measurement channels of multiple probes (e.g., ADG708). The constant current source applies a 1mA current to the probes through the current excitation lines, and the voltage V at probe contact points C and D is... C and VD The voltage signal V is input to a differential amplifier via a voltage detection line and outputs amplified voltage signal V from the differential amplifier. diff The signal is converted into a digital signal by an analog-to-digital converter and then transmitted to the edge computing system via an SPI / I2C interface; the edge computing system calculates the voltage signal V. diff The ratio between the contact impedance and the current I is used to obtain the contact resistance.

[0044] It should also be noted that the multidirectional adjustable probe array in this embodiment has only 8 probes, but the physical connection between it and the multiple terminals (far more than 8) of the terminal block is not one-to-one. Instead, efficient detection is achieved through the following mechanism:

[0045] (1) Probe dynamic adjustment and adaptive positioning

[0046] Each probe, controlled by an electromagnetic drive unit, can achieve axial extension and radial deflection. Combined with a star-shaped distribution, the probe array can cover multi-angle contact requirements. The edge computing system dynamically plans the movement path of the multi-directional adjustable probe array based on the layout of the terminal blocks under test and the detection requirements. For example, the multi-directional adjustable probe array can be adjusted to contact the terminals in a certain area first, and then move to the next area after the detection is completed, achieving time-division and area-division coverage detection.

[0047] (2) Multi-terminal group detection

[0048] Probe multiplexing technology: The same probe can be adjusted in position to sequentially contact multiple adjacent terminals. For example, a multi-directional adjustable probe array can perform segmented detection on a terminal block in a "scanning" manner, with each group of detections covering a portion of the terminals, ultimately completing the parameter acquisition for all terminals.

[0049] Adaptability: The probe material and star-shaped distribution design allow it to adapt to different sizes (3-40mm). 2 Terminals of various shapes and sizes ensure reliable contact with different terminals.

[0050] (3) Real-time feedback and calibration of contact impedance

[0051] Contact status self-test circuit: Real-time monitoring of the contact impedance between the probe and the terminal using a four-wire measurement method. If a probe has poor contact (such as terminal oxidation or misalignment), the system immediately triggers the electromagnetic drive unit to adjust the probe pressure or angle until impedance matching is achieved (automatic adjustment from 0-100Ω).

[0052] Magnetic-assisted positioning: The permanent magnet at the end of the probe forms a magnetic attraction with the magnetic material of the terminal block (such as iron terminals), which enhances contact stability and reduces miscontact caused by vibration or displacement.

[0053] As an example, assuming a terminal block has 24 terminals, the detection system can divide the 24 terminals into 3 detection areas. An electromagnetic drive unit sequentially moves probes to the corresponding positions in each area to complete the detection. During this process:

[0054] Step 1: Adjust the multi-directional adjustable probe array to the terminal position of the first area, and all 8 probes simultaneously contact the terminal of that area and collect data.

[0055] Step 2: After the test is completed, the probe retracts and moves to the terminal position of the second area, and the test process is repeated.

[0056] Step 3: The edge computing system integrates the detection data from all regions to generate global analysis results.

[0057] It should also be noted that the adaptive probe system ensures stable signals through impedance matching, referring to the electrical parameters (voltage, current, insulation resistance, phase) and high-frequency transient signals (CT secondary open-circuit characteristic signals) of the measured terminal block during acquisition and transmission. Impedance matching in this scheme is primarily used to eliminate signal attenuation, reflection, or noise interference caused by poor contact or impedance differences between the sensor and the terminal block. For example, with high-frequency transient signals (MHz level), if the probe contact impedance is mismatched, the high-frequency signal will be distorted due to reflection, affecting the accuracy of CT secondary open-circuit detection. Furthermore, when multiple electrical signals such as voltage and current are transmitted synchronously through the same probe array, impedance matching can reduce signal crosstalk, ensuring the independence and accuracy of each electrical parameter.

[0058] Traditional testing methods do not require impedance matching because conventional tools (such as multimeters and insulation meters) primarily detect power frequency (50Hz) or DC signals. Low-frequency signals are less sensitive to changes in contact impedance, and slight impedance fluctuations have negligible impact on measurement results. Traditional tools measure only a single parameter (such as voltage or resistance) at a time, eliminating the need for simultaneous transmission of multiple signals and thus avoiding signal crosstalk issues. Traditional methods cannot capture high-frequency transient signals (such as MHz-level signals generated by open circuits in the secondary windings of a current transformer), therefore, impedance matching for high-frequency transmission is unnecessary. Traditional testing methods rely on operators manually adjusting probe contact pressure and time, judging contact quality based on experience, rather than relying on automatic system adjustments.

[0059] Therefore, the detection system requirements of this solution include high-frequency signal transmission, multi-parameter synchronous transmission, and automated long-term monitoring, necessitating impedance matching. Traditional methods, limited by low frequency, single-parameter operation, and manual control, do not require impedance matching. For example, when measuring terminal block voltage with a multimeter, even a contact resistance of 1Ω has only a negligible impact on the measurement result (ΔV = I × 1Ω, where I is the measurement circuit current, typically extremely small). However, when this system detects MHz-level transient signals, a 1Ω contact impedance can lead to a signal reflectivity as high as 20% (impedance mismatch). Dynamic impedance matching must be used to reduce the contact impedance to below 0.1Ω to ensure signal integrity.

[0060] The principle of the probe extension or deflection driven by electromagnetic force in this embodiment is as follows: after the coil of each electromagnet is energized, it generates a magnetic field, which pushes the permanent magnet at the end of the probe to realize the axial extension or radial deflection of the probe. For example, in the vertical direction, the upper and lower probes are controlled to extend and retract synchronously to adapt to terminal blocks of different heights; in the horizontal direction, the deflection angle of the left and right probes is controlled to adapt to inclined or irregularly shaped terminal blocks.

[0061] The displacement of the probe can be precisely adjusted (accuracy ±0.1mm) by controlling the direction and intensity of the current in the electromagnet.

[0062] As an example, in the control process of an electromagnet, a hysteresis compensation coefficient (e.g., hysteresis = 0.2) is set to avoid frequent state switching caused by signal fluctuations; the signal range is determined and the corresponding operation is executed, as follows:

[0063] Condition 1: If the signal value is too low (contact pressure needs to be increased)

[0064] Judgment condition: If the signal value is <0.3 (i.e., 0.5-hysteresis).

[0065] Perform the following operation: Activate the electromagnet to level 4 power (level=4) to increase the contact pressure of the probe on the terminal block and ensure a stable connection.

[0066] Condition 2: Signal value is within the normal range (maintain current state)

[0067] Judgment condition: If the signal value is between 0.3 and 2.2 (i.e., 0.5 - hysteresis ≤ signal ≤ 2 + hysteresis).

[0068] Operation: Maintain the current of the electromagnet (I = 10mA) and keep the contact pressure between the probe and the terminal block constant.

[0069] Condition 3: Signal value too high (probe needs to be retracted immediately)

[0070] Judgment condition: If the signal value > 2.2 (i.e., 2 + hysteresis).

[0071] Execution: Immediately trigger probe retraction (trigger_retract(timeout=100ms)) and complete retraction within 100 milliseconds to prevent equipment damage due to excessive pressure.

[0072] Condition 4: The probe deflection angle error is ≤0.5°. If it exceeds this, position calibration will be initiated.

[0073] This embodiment improves control stability by introducing a hysteresis range (±hysteresis) to avoid frequent state switching when the signal jitters near the critical value. Through a multi-level response strategy of low-pressure enhancement, normal maintenance, and overpressure protection, it actively increases contact pressure when the signal value is weak to ensure a reliable connection between the probe and the terminal block; reduces unnecessary energy consumption and extends equipment life when the signal value is within the normal range; and quickly retracts the probe in case of overpressure to avoid mechanical damage or safety accidents. A timeout limit is set for the probe retraction action to ensure rapid response in emergency situations.

[0074] In this embodiment, the electromagnetic drive unit and the multi-directional adjustable probe array are connected via a flexible circuit board (FPC) to ensure signal transmission stability.

[0075] Therefore, the adaptive probe system of this embodiment realizes the magnetic dynamic impedance matching function (such as 0-100Ω automatic adjustment), and achieves a tight connection between the probe and the terminal through the magnetic principle. At the same time, it can automatically adjust the impedance according to different detection environments to ensure the stability of signal transmission.

[0076] In some embodiments, the detection system is provided with a dual isolation barrier, which includes optical coupling and magnetic isolation. The optical coupling is used to isolate the sensing signals, contact impedance signals and electromagnetic control signals between the edge computing system and the composite sensing array and the adaptive probe system; the magnetic isolation is used to isolate the high-frequency signals or power supply signals of the detection system.

[0077] In one implementation, the output signals of multiple sensors in the composite sensing array are input to the edge computing system through an optocoupler isolation circuit, the contact impedance feedback signal of the adaptive probe system is input to the edge computing system through an optocoupler isolation circuit, and the output signal of the high-frequency sampling module is input to the edge computing system through a magnetic coupling element.

[0078] The high-voltage detection end (hazard side) of this solution includes components that directly contact the high-voltage power circuit, such as the composite sensor array, adaptive probe system, and terminal blocks. The low-voltage control end (safety side) includes low-voltage control and processing units such as the edge computing system, power module, and communication interface. Optocouplers are used for isolation of digital signals or low-frequency analog signals, including optocoupler isolation between the output signal of the composite sensor array (such as temperature sensors or Hall sensors) and the edge computing system, and optocoupler isolation between the contact impedance feedback signal of the adaptive probe system and the edge computing system. Magnetic isolation is used for isolation of high-frequency signals or power supplies, including isolation between the transient signals of the high-frequency sampling module and the edge computing system, isolation between the PWM control signal of the electromagnetic drive unit and the adaptive probe system, isolation between the input power supply of the power module (such as 220V AC) and the isolation transformer and low-voltage DC-DC converter, and isolation between the sensor power supply at the high-voltage detection end and the power supply at the low-voltage control end. Magnetic isolation refers to the technology of achieving electrical isolation through magnetic coupling elements (such as isolation transformers, magnetic isolation chips, non-contact coupling, etc.). For example, the magnetic isolation chip can be selected from ADI. Alternatively, TI's ISO series digital isolators can transmit signals via magnetic coupling, with isolation voltages reaching over 5kV. They can be used for isolating high-frequency sampling signals and communication interfaces (such as SPI and I2C). Isolation transformers can be used for isolated power supply of power modules, converting high-voltage input to safe low voltage while blocking common-mode interference. For example, flyback transformers or push-pull isolated DC-DC converters use Rogowski coils to capture high-frequency transient signals from the secondary circuit of current transformers (CTs) in a non-contact manner, which is essentially a magnetic isolation technology.

[0079] In some embodiments, the electromagnetic drive unit is further provided with a safety retraction structure for the electromagnet, and the control of the electromagnet implements the following three-level protection mechanism:

[0080] Level 1: Software interruption (the electromagnet power supply stops when the signal is abnormal, <1ms).

[0081] Level 2: Hardware latch-up (FPGA forcibly cuts off the electromagnet's drive circuit, <10μs).

[0082] Level 3: Mechanical spring reset (after the electromagnet is de-energized, the spring pushes the probe back to the initial position, <500μs).

[0083] The three-tiered protection mechanism mentioned above will be explained in detail below.

[0084] The triggering conditions for a Level 1 software interrupt response include any of the following exceptions:

[0085] (1) Abnormal electrical parameters: Voltage surge / drop (e.g., exceeding ±20% of the rated value) or instantaneous current overload (e.g., exceeding 150% of the rated current) is detected.

[0086] (2) Abnormal contact condition: The contact impedance between the probe and the terminal block continuously deviates from the safe range (e.g., >50Ω or <0.1Ω).

[0087] (3) High-frequency transient signals: The intensity of the CT secondary open-circuit characteristic signal exceeds the threshold (e.g., >5V / μs).

[0088] The handling procedures for a Level 1 software interrupt include the following:

[0089] (1) Immediately interrupt the current operation: Pause all detection processes to prevent further data acquisition or probe action.

[0090] (2) Record abnormal data: Store abnormal parameters (such as voltage, current and impedance values) in the log of the edge computing core.

[0091] (3) Send alarm signals: Send real-time alarm information to operators through human-machine interface (HMI) or wireless module.

[0092] (4) Attempt self-recovery: If the abnormality is a transient interference (such as electromagnetic noise), it will automatically reset and re-detect within 1ms.

[0093] The triggering conditions for a Level 2 hardware latch-up response include any of the following:

[0094] (1) Continuous electrical abnormality: The current continuously exceeds 200% of the rated value and lasts for more than 1ms (software interruption not recovered), and the voltage rise / fall exceeds ±30% of the rated value and persists.

[0095] (2) Direct hardware signal abnormality: The FPGA detects that the level of a critical signal line (such as the overcurrent protection signal) jumps to a high level, and the temperature sensor detects that the terminal block temperature is >100℃ (preset threshold).

[0096] (3) Probe mechanical failure: The probe is detected to be stuck or exceeds the maximum allowable displacement range (e.g., ±2mm) by the probe displacement sensor.

[0097] Hardware latch implementation (hardware latch is implemented through pure hardware circuitry, ensuring a response time of <10μs):

[0098] (1) The sensor signal (such as current and voltage) is compared with the preset threshold (set by a precision resistor network) in real time by the comparator circuit, and the high and low levels are output; then the FPGA logic judges: the level change of the key signal line (such as the overcurrent signal line) is directly monitored by the hardware description language (HDL) programming, and the latching signal is triggered.

[0099] (2) Power cut-off and signal isolation are achieved through MOSFET or solid-state relay (SSR): After the hardware latch signal is triggered, the MOSFET gate voltage is pulled low or the SSR control terminal is de-energized, immediately cutting off the power supply (5VDC) of the probe's electromagnetic drive unit.

[0100] (3) Hardware logic gate blocking: Using a combination of AND gate and NOT gate, the sensor signal input path is blocked to prevent abnormal data from entering the edge computing system.

[0101] The triggering conditions for a Level 3 physical isolation response include any of the following:

[0102] (1) Extreme danger signal: Short circuit current > 10kA and insulation resistance < 1kΩ are detected (direct grounding risk).

[0103] (2) Secondary hardware interlock failure: If the fault persists after hardware interlock (e.g., the arc is not extinguished).

[0104] (3) Mechanical jamming: The probe is stuck due to foreign objects and cannot retract, triggering the mechanical stress sensor alarm.

[0105] The following actions are included after a Level 3 physical isolation response is triggered:

[0106] (1) Activate the physical isolation device:

[0107] Electromagnetic fuse: By blowing the fuse with a large current, the electrical connection between the detection device and the terminal block is completely broken.

[0108] Mechanical spring reset: Releases the preloaded spring, forcibly pulling all probes back to their initial position (disengaging from terminal block contact).

[0109] (2) Enable the backup isolation path: switch to the optocoupler isolation channel to ensure that the control signal is completely isolated from the high voltage circuit.

[0110] (4) System self-test and locking:

[0111] Perform the self-test procedure of the full detection system to confirm the fault type (such as short circuit or over-temperature).

[0112] The device operation permissions are locked, and manual intervention (such as replacing the fuse or clearing the obstruction) is required before it can be restarted.

[0113] In some embodiments, the edge computing system includes an embedded central processing unit and a neural network processor, and a lightweight deep learning model is configured on the edge computing system.

[0114] As an example, the edge computing system employs a dual-core heterogeneous architecture (Cortex-M7 + NPU). The Cortex-M7 handles routine data processing and control tasks, while the NPU is dedicated to accelerating deep learning and artificial intelligence algorithms. Working together, they improve computational efficiency. The lightweight deep learning model uses an embedded lightweight TensorFlow Lite model (<500KB). The original TensorFlow model is optimized and trimmed, removing unnecessary parameters and layers. Simultaneously, quantization techniques are used to compress the model size to <500KB while maintaining performance, facilitating operation on embedded devices. The edge computing system has an edge computing latency of ≤50ms, meeting the real-time requirements of the IEC 61850-5 standard, ensuring timely processing and feedback of detection data, and improving the real-time performance and effectiveness of detection.

[0115] In some embodiments, the edge computing system is also configured with a device health index model to analyze and process multiple electrical parameter data, predict the device health index (HI), and provide a scientific basis for the health assessment of the device.

[0116] As an example, the formula for calculating the health index is as follows:

[0117] HI=α·Vn o rm+β·In o rm+γ·Rn o rm

[0118] Where Vnorm, Inorm, and Rnorm are the normalized voltage, current, and insulation resistance parameters, respectively, and α, β, and γ are the dynamic weighting coefficients.

[0119]

[0120] Wherein, τ is the equipment aging time constant (derived from historical data fitting); R0 is the initial insulation resistance value (the reference value when the equipment leaves the factory or is calibrated); ΔI is the instantaneous change in current (captured by a high-frequency sampling module); the time weight α decreases as the equipment runs longer, reflecting the gradual weakening of the health impact of aging (needs to be corrected in conjunction with maintenance records); the current change weight β increases as the current fluctuations are greater, highlighting the potential risks of sudden load changes to the equipment; and the insulation resistance weight γ increases significantly when the insulation performance deteriorates, serving as a warning of the risk of insulation failure.

[0121] Acquire data (voltage, current, insulation, temperature, phase) synchronously collected through the composite sensor array and transient signals from the high-frequency sampling module; obtain dynamic weight coefficients based on the dynamic weight adjustment logic above; input the above data into the equipment health index model to obtain the health index value; when HI≥0.85, issue an early warning signal, which can be directly used for equipment maintenance priority ranking.

[0122] This embodiment solves the problem that traditional detection methods cannot interact with the equipment management system in real time, resulting in data silos. It can identify the operating status of equipment by analyzing the collected electrical data in a timely manner, which is conducive to the intelligent management of equipment.

[0123] The data for each indicator item are shown in Table 1, compared with traditional detection methods.

[0124] Table 1: Comparison results between traditional methods and this scheme

[0125] Indicator Item Traditional methods This invention Increase Single-point detection time 120s 15s 87.5％ False positive rate 3.2％ 0.12％ 96.25％ weight 2.3kg 0.38kg 83.5％

[0126] As can be seen, this solution effectively improves detection efficiency, with a single-point detection time of ≤15 seconds, an improvement of 87.5% compared to traditional detection methods, and the accuracy is also significantly improved. Furthermore, the electromagnetic compatibility of the detection system in this embodiment passes the Level 4 test of the GB / T 17626 series standards, and the probe array withstands ≥500,000 insertion / removal cycles (IEC 60512-7 test). Due to the improved detection efficiency, this solution reduces the number of inspection personnel and the inspection time, while also reducing maintenance costs caused by equipment failures, effectively reducing detection costs. For example, compared with the measured data of a pumped storage power station in 2024, it saves approximately 42,000 yuan in labor costs annually. The AR display module can realize the three-dimensional visualization of data analysis results, such as... Figure 3 As shown.

[0127] Based on any of the above embodiments, this application also provides a multi-parameter intelligent detection method for power secondary circuits. The execution subject of this method is an edge computing system, such as... Figure 4 As shown, the method includes the following steps:

[0128] Step S101: Obtain the contact impedance between each probe on the adaptive probe system and the terminal of the terminal block to be tested, and adjust the impedance matching between each probe of the adaptive probe system and the terminal of the terminal block to be tested if the contact impedance does not meet the preset impedance threshold.

[0129] The adaptive probe system utilizes magnetic dynamic impedance matching to automatically adjust the impedance between each probe and the terminals of the terminal block, ensuring optimal signal transmission across the terminal block.

[0130] Step S102: Simultaneously acquire multiple electrical parameters of the terminal block under test through the composite sensor array.

[0131] By utilizing a composite sensor array to simultaneously acquire five electrical parameters—voltage, current, insulation resistance, temperature, and phase—detection efficiency is improved.

[0132] Step S103: Feature extraction is performed on multiple electrical parameters to obtain feature values ​​that reflect the operating status of the power equipment related to the terminal block.

[0133] As one implementation method, a multi-parameter feature fusion method based on Fourier transform (FFT) and wavelet transform is adopted to extract feature values ​​that can reflect the operating status of relevant equipment from the collected raw electrical data.

[0134] Step S104: Input the feature values ​​into the deep learning model to identify whether there are abnormal patterns in the power equipment and present the identification results in a visual manner.

[0135] As one implementation method, the extracted feature values ​​are analyzed and identified using an embedded lightweight TensorFlow Lite model to determine whether the device has abnormal patterns; and the detection results are presented to the operator in an AR visualization manner to facilitate an intuitive understanding of the device's operating status.

[0136] Therefore, through in-depth analysis and model training of multiple electrical parameter data, various abnormal modes can be accurately identified. This embodiment can realize the identification of abnormal modes of 12 types of secondary circuits (e.g., hidden defects such as CT secondary open circuit and PT multi-point grounding), providing a guarantee for the safe operation of equipment.

[0137] As an example, such as Figure 5 As shown, multiple electrical parameters are first preprocessed, then spectrum analysis is performed using FFT, and then feature values ​​are extracted from the spectrum analysis results using wavelet packet transform. The feature values ​​are then input into the TensorFlowLite model to obtain the model prediction results, which are then output through an AR visualization rendering system.

[0138] In some embodiments, a method for adjusting the impedance matching between each probe of an adaptive probe system and the terminals of a terminal block to be tested includes: determining a target probe whose contact impedance does not meet a preset impedance threshold; and controlling the electromagnet of the target probe based on the contact impedance of the target probe to achieve impedance matching between the target probe and the terminals of the terminal block to be tested.

[0139] The specific preset impedance threshold and control implementation method are described in the corresponding content of the above embodiments.

[0140] In some embodiments, the detection method of this scheme further includes a three-level protection mechanism, as follows:

[0141] Determine whether any of the soft interrupt triggering conditions are met. Soft interrupt triggering conditions include abnormal electrical parameters, abnormal contact impedance, and abnormal open circuit characteristic signal.

[0142] If any of the software interrupt triggering conditions are met, a Level 1 software interrupt response is executed; the Level 1 software interrupt response includes detecting pause, alarm, and detecting self-recovery operations.

[0143] Determine whether any one of the hardware latching trigger conditions is met. The hardware latching trigger conditions include any electrical parameter among multiple electrical parameters satisfying the first abnormal condition of the corresponding electrical parameter, hardware signal abnormality, and probe mechanical failure.

[0144] If any one of the hardware latching trigger conditions is met, a secondary hardware latching response is executed; the secondary hardware latching response includes cutting off the power supply to the electromagnets of each probe and shutting down the data transmission path of the composite sensor array.

[0145] When any of the following is detected: extreme danger signal, second-level hardware lockout response delay, or probe jammed by foreign object, a third-level physical isolation response is executed. The third-level physical isolation response includes disconnecting the electrical connection between the detection system and the terminal block under test through an electromagnetic fuse, forcibly pulling all probes back from the terminal block to their initial position through a mechanical spring, and switching the direct electrical connection between the edge computing system and the composite sensor array and adaptive probe system to an optocoupler isolation channel to ensure complete isolation of the control signal from the high-voltage circuit.

[0146] It should be noted that before enabling the optocoupler isolation channel, the connection between the detection system and the high-voltage circuit was a direct electrical connection: control signals from the edge computing system (such as PWM signals and probe adjustment commands) were transmitted directly from the edge computing system to the electromagnetic drive unit on the high-voltage side via unisolated wires; sensor signals (such as voltage, current, and temperature) were transmitted from the composite sensor array on the high-voltage side to the edge computing system via unisolated circuitry. The potential risks of this direct electrical connection include: transient interference from the high-voltage circuit (such as arcing and surges) may couple to the low-voltage control terminal via wires, damaging equipment or causing misjudgments. For example, the PWM control line of the electromagnetic drive unit was directly connected to the GPIO pin of the edge computing system without isolation, and the current signal from the Hall sensor was input to the ADC via a standard operational amplifier circuit without optocoupler isolation.

[0147] The optical coupling isolation channel here refers to the isolation of control signals between the edge computing system and the electromagnetic drive unit through optical coupling, the isolation of sensor signals between the composite sensor array and the edge computing system through optical coupling, and the isolation of the communication interface between the feedback signal of the adaptive probe system and the edge computing system through optical coupling.

[0148] As an example, the PWM control signal of the edge computing system is transmitted via optocoupler isolation, with the input connected to the low-voltage control terminal and the output connected to the electromagnetic drive circuit on the high-voltage side. For instance, the PWM output of the edge computing system is connected to the input of the optocoupler isolation circuit, and the output of the optocoupler isolation circuit is connected to the electromagnetic drive unit.

[0149] As an example, sensor signals between the composite sensor array and the edge computing system are isolated via an optocoupler isolation channel. Sensor output signals (such as temperature and voltage) are transmitted through optocoupler isolation to prevent high-voltage interference from affecting the ADC module. For instance, the analog signal output of the temperature sensor in the composite sensor array is connected to the input of the optocoupler isolation circuit via a voltage follower, and the output of the optocoupler isolation circuit is connected to the ADC input of the edge computing system.

[0150] As an example, the contact impedance feedback signal of the adaptive probe system is transmitted through optocoupler isolation to ensure signal integrity.

[0151] One implementation method for switching to the optocoupler-isolated channel is either hardware switching (physical switch) or software-controlled switching. Hardware switching can be relay switching, where a relay is inserted into the signal path, initially connecting to the non-isolated channel. When a high-voltage risk is detected, the relay closes, switching to the optocoupler-isolated channel. For example, when the insulation resistance is detected to be <1kΩ, the relay is triggered, using a MOSFET or solid-state relay (SSR) to achieve rapid switching (response time <100μs). Software-controlled switching involves the edge computing system automatically switching the signal path based on real-time monitored electrical parameters (such as voltage spikes, abnormal contact impedance, etc.). For example, when the voltage exceeds 120% of the rated voltage or the contact impedance exceeds 50 ohms, the optocoupler-isolated channel is activated; otherwise, the non-isolated channel, i.e., the direct electrical connection channel, is used.

[0152] Specific examples of the three-level protection mechanism are described in the corresponding content of the above system implementation embodiment, and will not be repeated here.

[0153] In the foregoing descriptions of the embodiments, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0154] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0155] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A multi-parameter intelligent detection system for power secondary circuits based on edge computing, characterized in that, The system includes a composite sensor array, an adaptive probe system, and an edge computing system. The composite sensor array includes multiple sensors, and the adaptive probe system includes a multi-directional adjustable probe array. The ends of each probe in the multi-directional adjustable probe array are embedded with permanent magnets. The multi-directional adjustable probe array transmits the electrical signals of the terminal block under test collected by the probe array to one or more of the multiple sensors through a multiplexer. Both the composite sensor array and the adaptive probe system are connected to the edge computing system.

2. The detection system according to claim 1, characterized in that, The multiple sensors include a temperature sensor, a Hall sensor, a voltage divider insulation detection module, and a high-frequency sampling module. The temperature sensor is used to collect temperature parameters, the Hall sensor is used to collect current parameters, the voltage divider insulation detection module is used to collect voltage parameters and insulation resistance to ground, and the high-frequency sampling module is used to collect open-circuit characteristic signals of the secondary circuit of the current transformer (CT).

3. The detection system according to claim 1 or 2, characterized in that, The adaptive probe system further includes a contact state self-test circuit and an electromagnetic drive unit. The contact state self-test circuit is used in conjunction with the edge computing system to obtain the contact impedance between each probe and the terminal of the terminal block to be tested. The electromagnetic drive unit includes a PWM voltage regulation module and an electromagnet independently configured for each probe. The electromagnet includes an iron core and an electromagnet coil. The edge computing system is also used to control the energizing direction and current intensity of the electromagnet through the PWM voltage regulation module based on the obtained contact impedance between each probe and the terminal of the terminal block to be tested, so as to drive the permanent magnet at the end of the probe to realize the axial extension and radial deflection of the probe.

4. The detection system according to claim 3, characterized in that, The multidirectional adjustable probe array includes eight probes arranged in a star shape, each probe is made of tungsten steel and has a gold-plated surface.

5. The detection system according to claim 3, characterized in that, The contact state self-test circuit includes multiple probe voltage acquisition circuits and a multiplexer. The multiple probe voltage acquisition circuits transmit the voltage difference signal to the edge computing system through the multiplexer. The probe voltage acquisition circuit includes a constant current source, a differential amplifier, and an analog-to-digital converter. The constant current source is connected to the distal end A and distal end B of the probe through two current excitation lines, respectively. The differential amplifier is connected to the contact point C and contact point D of the probe through two voltage detection lines, respectively, to acquire and amplify the analog voltage difference between the two contact points of the probe. The analog-to-digital converter converts the analog voltage difference into a digital voltage difference. The edge computing system is used to obtain the contact impedance between each probe and the terminal of the terminal block to be tested based on the voltage difference signal and the current parameters of the constant current source.

6. The detection system according to claim 3, characterized in that, The edge computing system includes an embedded central processing unit and a neural network processor. The edge computing system is equipped with a lightweight deep learning model and a device health index model. The device health index model is used to predict the device health index by calculating the weighted sum of multiple electrical parameters. The weight coefficients in the device health index model are all dynamic weight coefficients.

7. The detection system according to claim 3, characterized in that, The detection system is equipped with a dual isolation barrier, which includes optical coupling and magnetic isolation. The optical coupling is used to isolate the sensing signals, contact impedance signals and electromagnetic control signals between the edge computing system and the composite sensing array and adaptive probe system. The magnetic isolation is used to isolate the high-frequency signals or power supply signals of the detection system.

8. A multi-parameter intelligent detection method for power secondary circuits, characterized in that, The method is implemented using the detection system as described in any one of claims 1 to 7, and the method includes the following steps: The adaptive probe system obtains the contact impedance between each probe and the terminal of the terminal block to be tested, and adjusts the impedance matching between each probe of the adaptive probe system and the terminal of the terminal block to be tested when the contact impedance does not meet the preset impedance threshold. The composite sensor array simultaneously acquires multiple electrical parameters of the terminal block under test; Feature extraction is performed on the multiple electrical parameters to obtain feature values ​​that reflect the operating status of the power equipment related to the terminal block; The feature values ​​are input into a deep learning model to identify whether there are abnormal patterns in power equipment and the identification results are presented in a visual manner.

9. The method according to claim 8, characterized in that, The method is implemented using the detection system as described in any one of claims 3 to 7, wherein the method for adjusting the impedance matching between each probe of the adaptive probe system and the terminals of the terminal block to be detected includes: Identify target probes whose contact impedance does not meet the preset impedance threshold. Based on the contact impedance of the target probe, the electromagnet of the target probe is controlled to achieve impedance matching between the target probe and the terminals of the terminal block to be tested.

10. The method according to claim 9, characterized in that, The method further includes: Determine whether any one of the soft interrupt triggering conditions is met, including abnormal electrical parameters, abnormal contact impedance, and abnormal open circuit characteristic signal. If any of the software interrupt triggering conditions are met, a Level 1 software interrupt response is executed; the Level 1 software interrupt response includes detecting pause, alarm, and detecting self-recovery operation. Determine whether any one of the hardware interlocking trigger conditions is met. The hardware interlocking trigger conditions include any electrical parameter among a plurality of electrical parameters satisfying the first abnormal condition of the corresponding electrical parameter, hardware signal abnormality, and probe mechanical failure. If any one of the hardware latching trigger conditions is met, a secondary hardware latching response is executed; the secondary hardware latching response includes cutting off the power supply to the electromagnets of each probe and shutting down the data transmission path of the composite sensor array. When any of the following is detected: an extreme danger signal, a second-level hardware interlock response delay, or a probe jammed by a foreign object, a third-level physical isolation response is executed. The third-level physical isolation response includes disconnecting the electrical connection between the detection system and the terminal block under test, forcibly pulling all probes back from the terminal block to their initial position using mechanical springs, and switching the direct electrical connection between the edge computing system and the composite sensor array and adaptive probe system to an optocoupler isolation channel to ensure complete isolation between the control signal and the high-voltage circuit.

Images