Passenger train circuit board card detection system and method

Abstract

The invention provides a passenger train circuit board card detection system and method, and relates to the technical field of circuit fault detection, and the system comprises a data transmission module which supplies power to a to-be-detected board card in a non-contact manner through magnetic resonance coupling; the circuit fault detection module performs dynamic feature extraction and circuit fault detection based on the real-time feedback data; the self-heating repair analysis module generates a self-healing repair scheme when detecting an open circuit fault; the comparison and verification module obtains impedance characteristics after repairing the open circuit, and performs comparison and verification with a performance attenuation model; and if the verification is not passed, the parameter optimization module performs iterative repair based on the impedance characteristic difference until the verification is passed. The technical problem that in the prior art, passive overhaul depends on artificial experience and physical contact, so that it is difficult to accurately recognize and position hidden open circuit faults in complex scenes such as high-voltage isolation, and the overhaul efficiency of the circuit board card of the passenger train is affected can be solved, hidden damage is actively repaired, and the overhaul efficiency of the circuit board card of the passenger train is improved. And the maintenance efficiency of the passenger train circuit board card is improved.

Description

Technical Field

[0001] This application relates to the field of circuit fault detection technology, and in particular to a circuit board detection system and method for railway passenger cars. Background Technology

[0002] The inspection and repair of circuit boards has long relied on manual experience and physical contact equipment. Typically, probes or connectors are used to supply power and inject excitation signals to the board under test, and faults are diagnosed by measuring static electrical parameters such as voltage and current. The repair process itself is heavily reliant on manual operation. Minor damage or hidden open circuits often initially manifest as abnormal electrical characteristics or distorted signal waveforms, which traditional static measurement methods struggle to accurately detect and locate. Once these issues worsen, they can lead to circuit board failure during operation. This passive approach not only increases maintenance costs but also lacks quantitative evaluation of repair effectiveness and long-term reliability verification, leaving circuit board repair in a perpetually reactive, post-fault repair mode, further impacting the efficiency of circuit board repair in railway passenger cars.

[0003] In summary, existing technologies suffer from the technical problem of relying on manual experience and physical contact for passive maintenance, which makes it difficult to accurately identify and locate hidden circuit breakers in complex scenarios such as high-voltage isolation, further affecting the maintenance efficiency of railway passenger car circuit boards. Summary of the Invention

[0004] The purpose of this application is to provide a railway passenger car circuit board detection system and method to solve the technical problem in the prior art that the reliance on manual experience and physical contact for passive maintenance makes it difficult to accurately identify and locate hidden open circuit faults in complex scenarios such as high-voltage isolation, which further affects the maintenance efficiency of railway passenger car circuit boards.

[0005] In view of the above problems, this application provides a railway passenger car circuit board card testing system and method.

[0006] In a first aspect, this application provides a railway passenger car circuit board detection system, wherein the railway passenger car circuit board detection system includes: a data transmission module, used for non-contact transmission of electrical energy required for detection to the circuit board under test via magnetic resonance coupling, synchronous transmission of detection signals, and reception of real-time feedback data from the circuit board under test; a circuit fault detection module, used for dynamic feature extraction based on the real-time feedback data, and for circuit fault detection based on the extracted dynamic features, generating a fault detection result; and a self-heating repair analysis module, used for adjusting the transmission frequency to generate a high-frequency AC circuit when the fault detection result detects an open circuit fault. A variable magnetic field triggers the circuit board under test (PCB) to perform self-heating repair analysis, generating a self-healing repair scheme. The PCB is coated with a self-healing material capable of localized heating to repair open circuits. The heating parameters of the self-healing material are dynamically adjusted according to the location and severity of the open circuit. A comparison and verification module is used to obtain the impedance characteristics of the repaired area through multi-band impedance scanning after heating and repairing the open circuit based on the self-healing repair scheme, and compare and verify this with a pre-built performance degradation model. A parameter optimization module is used to automatically optimize the high-frequency alternating magnetic field parameters for iterative repair based on the impedance characteristic differences if the verification fails, until the verification passes.

[0007] Optionally, the data transmission module includes: a frequency band matching unit, used to match the resonant frequencies of the transmitting coil and the receiving coil in a preset low-frequency band to construct a non-contact energy transmission channel; a compensation calculation unit, used to sample the load current of the circuit board under test in real time during energy transmission, calculate the excitation voltage compensation amount of the transmitting coil based on the change of the load current, and generate a voltage adjustment signal; and a circuit control unit, used to control the driving circuit of the transmitting coil based on the voltage adjustment signal, output the excitation voltage after real-time compensation, and generate a dynamically stable power supply output.

[0008] Optionally, the data transmission module further includes: a signal encoding unit, used to encode the detection signal containing step, frequency sweep and pulse sequence into a digital command stream after the dynamically stable power supply output is established; a signal transmission unit, used to modulate the digital command stream onto the carrier of the non-contact energy transmission channel to generate a composite modulation signal and transmit it to the circuit board under test; and a demodulation and analog-to-digital conversion unit, used to receive the modulation response signal returned by the circuit board under test through the receiving coil coupling, and to demodulate and convert the modulation response signal into an analog-to-digital signal to generate real-time feedback data containing voltage waveform, current waveform, impedance spectrum and transient characteristics.

[0009] Optionally, the circuit fault detection module includes: a primary feature extraction unit, used to perform time-frequency domain joint transformation on the voltage waveform and current waveform in the real-time feedback data, extracting harmonic distortion rate, quality factor change curve, and transient response time features to form a primary feature set; a secondary feature extraction unit, used to perform feature decomposition on the impedance spectrum, extracting impedance amplitude, phase angle, and equivalent circuit parameters at characteristic frequency points to form a secondary feature set; a feature fusion unit, used to fuse the primary feature set and the secondary feature set, inputting them into a support vector machine classification model, and having the classification model output a fault probability distribution matrix; and a fault determination unit, used to generate a fault detection result based on the fault probability distribution matrix and a preset fault determination threshold, wherein the fault detection result includes fault coordinates, fault type code, and fault confidence weight.

[0010] Optionally, the self-heating repair analysis module includes: a mode switching unit, used to control the magnetic resonance coupling system to switch from a low-frequency energy transmission mode to a high-frequency magnetic field excitation mode when the fault mode in the fault detection result is an open circuit, raising the operating frequency to a preset high-frequency band, and generating a spatially focusable high-frequency alternating magnetic field above the circuit board under test; an energy control unit, used to control the energy of the high-frequency alternating magnetic field to concentrate on the open circuit area coated with self-healing material according to the fault location coordinates in the fault detection result; and a scheme generation unit, used to simulate and calculate the minimum temperature, heating rate, and holding time required for reliable repair based on the open circuit gap width, line layer depth, and confidence level in the fault detection result, combined with the phase transition dynamics model of the self-healing material, and generate the self-healing repair scheme including the target temperature curve, magnetic field strength distribution, and action sequence.

[0011] Optionally, the scheme generation unit includes: a model coupling subunit, used to construct a three-dimensional electromagnetic-thermal coupling model of the broken circuit region, including a substrate, wires, and a self-healing material coating, based on the fault coordinates and board layer information in the fault detection results; a target determination subunit, used to set a repair target in the three-dimensional electromagnetic-thermal coupling model of the broken circuit region based on the broken gap width and confidence weight, and generate the spatiotemporal distribution of thermal energy required to achieve the repair target through simulation calculation; and a magnetic field conversion subunit, used to convert the spatiotemporal distribution of thermal energy into magnetic field parameters, and generate the target temperature field, magnetic field intensity map, and heating sequence required to control the high-frequency alternating magnetic field to generate a predetermined temperature gradient in the broken circuit region.

[0012] Optionally, the self-heating repair analysis module further includes: a deviation calculation unit, used to monitor the surface temperature distribution of the broken area in real time through infrared thermometry during the repair process, compare the surface temperature distribution with the target temperature curve, and obtain the temperature distribution deviation; and a closed-loop control unit, used to dynamically correct the magnetic field strength distribution and action sequence based on the temperature distribution deviation to form closed-loop temperature control.

[0013] Optionally, the comparison and verification module includes: a frequency sweep measurement unit, used to control a vector network analyzer to perform frequency sweep measurements on the repair area within a preset frequency band after completing the current heating sequence; a result conversion unit, used to acquire the S-parameter measurement results of the repair area at multiple discrete frequency points, convert the S-parameter measurement results into complex impedance values, and generate the impedance characteristics of the repair area; an impedance characteristic error determination unit, used to extract the reference impedance characteristics and allowable error range of a similar intact board from the pre-built performance attenuation model; an impedance deviation calculation unit, used to compare the impedance characteristics of the repair area with the reference impedance characteristics point by point, and calculate the impedance deviation value at each frequency point in combination with the allowable error range; and a deviation fusion unit, used to weight and fuse the impedance deviation values ​​at all frequency points to generate a comprehensive repair quality index.

[0014] Optionally, the parameter optimization module includes: an impedance difference identification unit, used to determine that the verification has failed when the comprehensive repair quality score is lower than the verification threshold, analyze the difference between the impedance characteristics of the repair area and the reference impedance characteristics, and identify the frequency bands and components of impedance mismatch; an optimized parameter back-calculation unit, used to map the frequency bands and components of impedance mismatch to the corresponding circuit physical parameter defects, and back-calculate the direction and amplitude of the high-frequency alternating magnetic field parameters that need to be optimized according to the type and degree of the physical parameter defects, forming optimized high-frequency alternating magnetic field parameters; and a repair quality evaluation unit, used to re-execute the self-healing repair scheme using the optimized high-frequency alternating magnetic field parameters, and after each re-execution of the repair, re-acquire the impedance characteristics of the repair area and calculate a new comprehensive repair quality index until the new comprehensive repair quality index reaches or exceeds the quality threshold.

[0015] Secondly, this application also provides a method for detecting circuit boards in railway passenger cars. The method includes: non-contactly transmitting the electrical energy required for detection to the circuit board under test via magnetic resonance coupling, synchronously transmitting detection signals and receiving real-time feedback data from the circuit board under test; extracting dynamic features based on the real-time feedback data, and detecting circuit faults based on the extracted dynamic features to generate fault detection results; when the fault detection results detect an open circuit fault, adjusting the transmission frequency to generate a high-frequency alternating magnetic field, triggering the circuit board under test to perform self-heating repair analysis, and generating a self-healing repair scheme. The circuit board under test is coated with a self-healing material that can locally heat-repair the open circuit, and the heating parameters of the self-healing material are dynamically adjusted according to the open circuit location and fault severity; after heating and repairing the open circuit based on the self-healing repair scheme, obtaining the impedance characteristics of the repaired area through multi-band impedance scanning and comparing and verifying it with a pre-built performance attenuation model; if the verification fails, automatically optimizing the high-frequency alternating magnetic field parameters based on the impedance characteristic differences for iterative repair until verification is successful.

[0016] One or more technical solutions provided in this application have at least the following technical effects or advantages: non-contact transmission of the power required by the circuit board, detection signals and feedback data through magnetic resonance coupling technology; extraction of dynamic features based on real-time feedback data to determine the circuit status and generate fault detection results; when an open circuit fault is detected, the transmission frequency is adjusted to generate a high-frequency alternating magnetic field, triggering the self-healing material on the circuit board to heat up and repair the open circuit; after repair, the impedance of the repair area is scanned and compared with a pre-built performance attenuation model to determine whether the repair is effective; if the verification fails, the high-frequency magnetic field parameters are automatically adjusted, and the repair is carried out again until the repair is effective, thereby reducing interface wear, actively repairing hidden damage caused by open circuits, and reducing manual intervention, thus improving the maintenance efficiency of circuit boards for railway passenger cars.

[0017] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the railway passenger car circuit board detection system of this application.

[0020] Figure 2 This is a flowchart illustrating the railway passenger car circuit board testing method of this application.

[0021] Explanation of reference numerals in the attached diagram: Data transmission module 11, Circuit fault detection module 12, Self-heating repair analysis module 13, Comparison and verification module 14, Parameter optimization module 15. Detailed Implementation

[0022] This application provides a railway passenger car circuit board detection system and method, solving the technical problem in existing technologies where passive maintenance relying on manual experience and physical contact makes it difficult to accurately identify and locate hidden open circuit faults in complex scenarios such as high-voltage isolation, further affecting the maintenance efficiency of railway passenger car circuit boards. The system utilizes magnetic resonance coupling technology to transmit the required electrical energy, detection signals, and feedback data of the circuit board non-contactly; it extracts dynamic features based on real-time feedback data to determine the circuit status and generate fault detection results; when an open circuit fault is detected, the transmission frequency is adjusted to generate a high-frequency alternating magnetic field, triggering the self-healing material on the circuit board to heat up and repair the open circuit; after repair, the impedance of the repaired area is scanned and compared with a pre-built performance attenuation model to determine the effectiveness of the repair; if the verification fails, the high-frequency magnetic field parameters are automatically adjusted, and the heating and repair are repeated until the repair is effective. This reduces interface wear, actively repairs hidden open circuit damage, and reduces manual intervention, thereby improving the maintenance efficiency of railway passenger car circuit boards.

[0023] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0024] Example 1, please refer to the appendix. Figure 1 This application provides a railway passenger car circuit board detection system, wherein the railway passenger car circuit board detection system is used to implement the steps of the railway passenger car circuit board detection method, and the railway passenger car circuit board detection system includes:

[0025] The data transmission module 11 is used to transmit the electrical energy required for testing to the circuit board under test non-contactly via magnetic resonance coupling, synchronously transmit the testing signal, and receive the real-time feedback data from the circuit board under test.

[0026] Furthermore, the data transmission module 11 in the railway passenger car circuit board detection system further includes: a frequency band matching unit, used to match the resonant frequencies of the transmitting coil and the receiving coil in a preset low frequency band to construct a non-contact energy transmission channel; a compensation calculation unit, used to sample the load current of the circuit board under test in real time during energy transmission, calculate the excitation voltage compensation amount of the transmitting coil based on the change of the load current, and generate a voltage adjustment signal; and a circuit control unit, used to control the driving circuit of the transmitting coil based on the voltage adjustment signal, output the excitation voltage after real-time compensation, and generate a dynamically stable power supply output.

[0027] Specifically, a transmitting coil and its matching resonant and driving circuits are installed on the testing equipment. A receiving coil is pre-installed or temporarily fixed on the board under test (DUT), and this coil is connected to the board's power input interface. The receiving coil is usually designed to be small and thin, and can be temporarily attached to the board or integrated into the board testing fixture. Once the DUT is placed in the designated position on the testing equipment, the equipment begins preparations for power transfer. Whether the transmitting and receiving coils are within the effective coupling range is usually determined by short-pulse excitation and response detection. After confirming the correct position, the driving circuit is activated, causing the transmitting coil to generate a high-frequency alternating magnetic field. The frequency of this magnetic field is precisely controlled within a preset resonant frequency band, typically a fixed value between 85kHz and 205kHz. The receiving coil is located in this high-frequency alternating magnetic field, and due to the principle of electromagnetic induction, an induced electromotive force is generated inside the coil. More importantly, because the receiving and transmitting coils are tuned to the same resonant frequency, they form a magnetic resonance state, making the energy transfer efficiency far higher than ordinary electromagnetic induction. The AC power generated by the receiving coil is rectified and filtered to convert into the DC operating voltage required by the board under test (DUT), such as 5V, 3.3V, or 12V. As the receiving voltage stabilizes, the DUT powers on, and its internal main control chip, memory, communication interface, and other circuits sequentially enter their operational state. At this point, the board is ready to receive test commands, execute self-test programs, and send its operating status data back to the testing equipment via the feedback channel. Throughout the entire process, a gap of several millimeters to several centimeters is maintained between the transmitting and receiving coils, with no physical contact. This means that complete electrical isolation is achieved between the testing equipment and the DUT, allowing for safe operation even under high voltage conditions, while completely eliminating interface wear caused by repeated connector plugging and unplugging.

[0028] Before contactless power supply, a physical channel for energy transmission needs to be established. The resonant circuits containing the transmitting and receiving coils are tuned to the same resonant frequency, typically set to a fixed value between 85kHz and 205kHz. When the resonant frequencies of the two circuits are consistent, even with air gaps of several millimeters or centimeters between the transmitting and receiving coils, efficient magnetic field coupling can be formed, thus establishing a stable contactless energy transmission channel. Magnetic resonance coupling is a technology that utilizes high-frequency magnetic fields and resonance principles to achieve wireless power transmission. When the transmitting and receiving coils operate at the same resonant frequency, a strong magnetic field coupling is formed between them, achieving efficient energy transfer. Resonant frequency matching is used to adjust the natural frequencies of the transmitting and receiving LC resonant circuits to be consistent. When their frequencies are consistent, the circuit resonates, resulting in minimum impedance, maximum current, and peak energy transmission efficiency. The preset low-frequency band is usually selected between tens and hundreds of kilohertz. This band achieves high energy transmission efficiency while avoiding mutual interference with the high-frequency bands subsequently used to trigger self-healing material repair.

[0029] After the non-contact energy transmission channel is established, the circuit board under test (DUT) is powered on. Since the current consumed by the circuit board changes in real time during startup, execution of different functions, and standby, this change affects the electrical parameters of the transmitter via magnetic field coupling. The load current is monitored in real time using a current sampling circuit. When the sampling circuit detects a change in load current, it calculates the incremental excitation voltage that needs to be compensated at the transmitter to maintain a constant receiver voltage, and generates a voltage regulation signal accordingly. To achieve accurate compensation calculation, an equivalent circuit model based on mutual inductance coupling theory is first needed to describe the electrical relationship between the transmitter, coil coupling, receiver, and load. The equivalent circuit at the transmitter describes the detection device itself, including the inductive characteristics of the transmitting coil, the small resistance of the transmitting coil itself (i.e., copper loss), and the compensation capacitor added to generate resonance with the receiver. The combination of these three components determines the frequency at which the transmitter operates most efficiently. The equivalent circuit at the receiver describes the receiving side of the board under test (DUT), including the inductance characteristics of the receiving coil, its internal resistance, and the resonant compensation capacitor. The parameters of these three components are essentially symmetrical to those at the transmitter to ensure efficient energy transfer. The core of magnetic coupling is mutual inductance, the most critical parameter in the model, describing the tightness of the connection between the transmitting and receiving coils. The magnitude of mutual inductance depends on the relative positions of the two coils: the closer and more precisely aligned they are, the greater the mutual inductance; the farther apart and the greater the offset, the smaller the mutual inductance. In actual testing, the coil positions may vary slightly each time the DUT is placed, therefore, mutual inductance is a real-time variable and cannot be fixed in advance. The dynamically changing load is the DUT itself, which is the load of the model. The power consumed by the board during operation is constantly changing, represented in the model as a dynamically changing resistance value. This load variation is the root cause of supply voltage fluctuations. The equivalent circuits at the transmitter and receiver, mutual inductance, and dynamically changing loads are not isolated; they are interconnected through a set of equations based on circuit laws. The voltage applied at the transmitter induces current in the transmitting coil; the current in the transmitting coil induces voltage and current in the receiving coil through mutual inductance; the current flowing through the load at the receiver forms the final supply voltage; and simultaneously, the current at the receiver also acts back to the transmitter through mutual inductance, affecting the transmitter's operating state. This complete equivalent circuit model is the core of the calculations for compensation.

[0030] In a laboratory environment, using precision measuring instruments such as an LCR bridge, the static electrical parameters of the transmitting and receiving coils are directly measured, including their inductance and internal resistance values. The capacitance value of the resonant capacitor is then determined based on the design drawings. Once determined, these parameters remain largely unchanged under normal operating conditions and can be considered fixed constants for the model. The receiving coil is placed on a precisely movable fixture, with measurement points set at regular intervals within the expected operating range, such as coil spacing from 5mm to 30mm and horizontal offset from -10mm to +10mm. At each measurement point, a standard test load with a known resistance is connected. A fixed, low-power test voltage is applied to the transmitting end, and the voltage received at the receiving end is measured. Since the transmitting voltage, receiving voltage, load resistance, and all fixed parameters are known at this point, the mutual inductance value at that location can be deduced from the model equations. The mutual inductance value corresponding to each location is recorded, forming a lookup table of location and mutual inductance. This is equivalent to pre-installing vision on the model, allowing it to roughly know the approximate mutual inductance value when the coil is in a certain position. The equivalent circuit structure, fixed electrical parameters, and mathematical logic used for back-calculation described above are compiled into executable program code and burned into the control system chip of the testing equipment. Simultaneously, the mutual inductance lookup table generated in the second step is also stored in the system to determine the equivalent circuit model based on the mutual inductance coupling theory.

[0031] Load current and voltage are acquired at extremely high rates. Although an offline lookup table exists, slight deviations in board placement and parameter drift caused by temperature changes can lead to discrepancies between the actual mutual inductance and the lookup table. A very short time window is captured, such as when the board state is relatively stable, and the currently sampled current and voltage values ​​are substituted into the model. If the mutual inductance is a certain value, then under the current transmit voltage, the currently sampled current and voltage should be generated. By reverse engineering, the most reasonable mutual inductance value is found, minimizing the error between the model prediction and the measured result. This online calculated mutual inductance truly represents the actual mutual inductance at the current physical location and is used to replace the theoretical value in the lookup table for subsequent calculations, thereby achieving self-adaptation. The current load state, coupling state, fixed parameters, and control target are all input into the aforementioned equivalent circuit model for reverse engineering. The target transmit voltage is compared with the actual output voltage of the current transmitter to obtain the difference between the two, which is the amount of voltage compensation required. This compensation amount is converted into specific control commands, such as increasing the duty cycle of the pulse width modulation signal from the current 30% to 70%. The command is sent to the driver circuit at the transmitting end, which responds immediately by adjusting the output voltage to restore the receiving end voltage to a stable 5V. Because the model calculations may contain minor errors, or the load may change drastically, a single adjustment may not be perfect. Therefore, after adjusting the transmitting voltage, the receiving end voltage is continuously monitored. If the voltage is still slightly below 5V, this small deviation is calculated. This deviation is input into a proportional-integral-derivative (PID) controller. This controller generates a small correction based on the magnitude and trend of the deviation, which is added to the next calculation result. This dual protection of feedforward calculation and feedback correction ensures that the receiving end voltage remains stable near the target value under any operating conditions.

[0032] The calculated voltage regulation signal is sent to the drive circuit of the transmitting coil. The drive circuit typically consists of a full-bridge or half-bridge inverter and a pulse width modulation controller. Based on the received instructions, it adjusts the amplitude of its output excitation voltage or the pulse duty cycle in real time, outputting a precisely compensated high-frequency AC power to the transmitting coil. Through this dynamic, real-time closed-loop regulation, even if the load current of the circuit board under test fluctuates within a large range, the DC voltage output at the receiving end can always remain stable near the rated operating voltage required by the circuit board, thus generating a dynamically stable power supply output.

[0033] Furthermore, the data transmission module 11 in the railway passenger car circuit board detection system further includes: a signal encoding unit, used to encode the detection signal containing step, frequency sweep and pulse sequence into a digital command stream after the dynamic and stable power supply output is established; a signal transmission unit, used to modulate the digital command stream onto the carrier of the non-contact energy transmission channel to generate a composite modulation signal and transmit it to the circuit board under test; and a demodulation and analog-to-digital conversion unit, used to receive the modulation response signal returned by the circuit board under test through the receiving coil coupling, and to demodulate and convert the modulation response signal into an analog-to-digital signal to generate real-time feedback data containing voltage waveform, current waveform, impedance spectrum and transient characteristics.

[0034] Specifically, after the circuit board under test receives a stable power supply and completes startup, the testing equipment prepares to send testing commands. According to a preset testing strategy, a series of raw testing signals are generated to probe the board's status, including a step signal for testing transient response, a frequency sweep signal for testing frequency characteristics, and a pulse sequence for testing digital logic functions. A step signal is a signal that jumps instantaneously from one level to another, used to observe the circuit's response characteristics to sudden inputs, such as rise time, overshoot, and settling time; a frequency sweep signal is a signal whose frequency changes continuously over time, used to measure the circuit's impedance characteristics, amplitude-frequency characteristics, and phase-frequency characteristics at different frequencies; a pulse sequence is a series of pulses arranged according to a specific pattern, used to simulate the actual operating state of the digital circuit and detect logic functions and timing relationships.

[0035] These raw detection signals cannot be directly transmitted through the magnetic coupling channel because the channel is primarily used for transmitting electrical energy, and the signals require precise digital control. The detection signals, including step, frequency sweep, and pulse sequences, are encoded into a digital command stream. Parameters such as the amplitude and duration of the step, the start and end frequencies of the frequency sweep, the scan rate, and the width and interval of the pulse sequence are converted into binary data packets according to a predetermined communication protocol. After encoding, this digital command stream needs to be loaded onto an existing contactless energy transmission channel—this is modulation. Composite modulation technology is used, where tiny frequency, phase, or amplitude changes are used to represent 0s and 1s in the digital command stream on the carrier wave originally used for transmitting electrical energy. Because these changes are extremely small and short-lived, the power supply circuit at the receiving end treats them as noise and filters them out. However, the communication demodulation circuit on the board accurately identifies and decodes them, generating a composite modulated signal that simultaneously carries electrical energy and the detection signal. The composite modulated signal is a magnetically coupled carrier signal that simultaneously carries electrical energy and the detection signal.

[0036] The composite modulated signal is converted into a high-frequency alternating magnetic field by the transmitting coil, passes through the air gap, and reaches the receiving coil on the board under test. The receiving coil simultaneously receives two things: a continuous, stable electrical energy to power the board; and a weak detection command signal superimposed on the electrical energy. The board's power input is typically equipped with filter capacitors and large capacitors, which smooth out minute signal fluctuations, ensuring the detection command does not interfere with the board's normal power supply. Meanwhile, a specially designed demodulation circuit on the board extracts the detection command from the received composite signal. By detecting minute changes in the carrier wave, the demodulation circuit reconstructs the digital command stream encoded by the transmitting end and parses it into specific test commands. For example, when a command to perform a frequency sweep test is parsed, the test response circuit on the board will begin executing the corresponding operation according to the command requirements. For frequency sweep tests, the board may connect an internal test load to the circuit or adjust its operating state so that its overall impedance characteristics change with frequency; for step tests, the board may quickly switch an internal load, generating a transient response.

[0037] During the execution of detection commands, the electrical state of the circuit board changes, which is reflected in the overall impedance of the board as seen from the receiving end. This change in impedance, in turn, affects the load condition of the receiving coil. According to the principle of mutual inductance, the change in the load at the receiving end is reflected back to the transmitting end through the magnetic field, causing a small change in the voltage or current across the transmitting coil, which is the board's response signal to the detection command.

[0038] The voltage and current at the transmitting end are continuously monitored to extract minute changes caused by the board's response, i.e., the modulation response signal, which is transmitted back to the detection equipment via a magnetic coupling channel. Upon receiving the modulation response signal, demodulation is performed to reconstruct the analog information it contains. However, this information is still a continuous analog quantity and cannot be directly sent to a computer for analysis. Therefore, it needs to undergo analog-to-digital conversion (ADC) at a rate no less than twice the highest frequency of the signal to convert the continuous analog signal into discrete digital quantities. After ADC, real-time feedback data including voltage waveforms, current waveforms, impedance spectra, and transient characteristics are obtained. The voltage waveform is the curve showing the voltage change over time at the board's power supply point or key test point; the current waveform is the curve showing the current consumed by the board over time; the impedance spectrum is the impedance value exhibited by the board at different frequencies, including amplitude and phase; transient characteristics include rise time, settling time, and overshoot amplitude in the step response.

[0039] For example, the board under test is a door control unit of a certain type of vehicle, with a rated voltage of 5V. Non-contact power supply has been established, and the receiving voltage is stable at 5.02V. A set of detection signals includes three types: a step signal with an amplitude of 0.5V and a rise time of 10ns, superimposed on the supply voltage; a swept frequency signal with a frequency range of 100Hz to 10MHz, a scan time of 100ms, and an output amplitude of 100mV; and a pulse sequence with a frequency of 1kHz, a duty cycle of 50%, and a duration of 50ms. These signal parameters are encoded into a digital command stream with a data packet length of 128 bytes. Frequency shift keying modulation is used, with a carrier frequency of 100kHz as the center, and an offset of ±5kHz representing digits 0 and 1. The modulation depth is set to 10% to ensure no impact on the main power transmission. The composite modulated signal is emitted through the transmitting coil; the measured transmitting voltage is 24V, and the amplitude of the superimposed modulated signal is 2.4V. The measured received voltage at the receiving coil was 5.02V, the amplitude of the superimposed modulation signal was 0.5V, and the signal-to-noise ratio was 32dB, meeting the demodulation requirements. The demodulation circuit on the board successfully parsed all 128 bytes of instructions, with a 100% instruction parsing accuracy. The board started the corresponding test program according to the instructions. During the frequency sweep test, the internal test circuit of the board switched the load point by point according to the instructions, and the detection equipment monitored the changes in voltage and current at the transmitting end in real time. Data from 200 frequency points were collected, each point including impedance amplitude and phase angle. The analog-to-digital converter sampling rate was set to 20MHz, and the resolution was 12 bits. The collected raw data volume was 4MB. After processing, the following real-time feedback data was generated: voltage waveform data with a sampling time of 200ms, including step response and impulse response waveforms; current waveform data sampled synchronously, corresponding to the voltage waveform; impedance spectrum data of impedance amplitude and phase at 200 points in the 100Hz to 10MHz frequency band, with typical data points showing an impedance amplitude of 120Ω and a phase angle of -5° at 1kHz, and an impedance amplitude of 45Ω and a phase angle of -32° at 1MHz; transient characteristic data of step response rise time of 12μs, settling time of 45μs, and overshoot amplitude of 3.2%. The above data was packaged into JSON format, with a total size of 4.2MB, and transmitted to the diagnostic analysis module for further processing. The entire signal transmission and feedback reception process took 320ms, of which the frequency sweep test accounted for 200ms and data acquisition and processing accounted for 120ms.

[0040] The receiving coil on the board under test simultaneously receives two things: a continuous and stable power supply to maintain normal operation, and a detection command signal superimposed on the power supply. The board executes corresponding operations according to the command. For example, when a frequency sweep command is received, the test response circuit inside the board will change its load state or operating mode point by point according to the preset scan range and step; when a step command is received, the board will quickly switch a certain internal load, generating a transient response. According to the principle of mutual inductance, the change in the impedance at the receiving end will be reflected back to the transmitting end through the magnetic field, causing a small change in the voltage or current across the transmitting coil. The detection equipment continuously monitors the voltage and current at the transmitting end, extracts the modulation response signal caused by the board's response, demodulates it, restores the analog information transmitted back by the board, and then performs high-speed analog-to-digital conversion to finally generate multi-dimensional real-time feedback data including voltage waveform, current waveform, impedance spectrum, and transient characteristics. Power transmission and bidirectional data communication are completed simultaneously on the same magnetic coupling channel. The entire process requires no physical contact, completely eliminating the interface wear problem caused by repeated plugging and unplugging, and achieving natural electrical isolation in high-voltage scenarios, ensuring operational safety.

[0041] The circuit fault detection module 12 is used to extract dynamic features based on the real-time feedback data, and to detect circuit faults based on the extracted dynamic features, thereby generating fault detection results.

[0042] Furthermore, the circuit fault detection module 12 in the railway passenger car circuit board detection system further includes: a primary feature extraction unit, used to perform time-frequency domain joint transformation on the voltage waveform and current waveform in the real-time feedback data, extracting harmonic distortion rate, quality factor change curve, and transient response time features to form a primary feature set; a secondary feature extraction unit, used to perform feature decomposition on the impedance spectrum, extracting impedance amplitude, phase angle, and equivalent circuit parameters at characteristic frequency points to form a secondary feature set; a feature fusion unit, used to fuse the primary feature set and the secondary feature set, inputting them into a support vector machine classification model, and outputting a fault probability distribution matrix from the classification model; and a fault determination unit, used to generate a fault detection result based on the fault probability distribution matrix and a preset fault determination threshold, wherein the fault detection result includes fault coordinates, fault type code, and fault confidence weight.

[0043] Specifically, a joint time-frequency domain transformation is performed on the voltage and current waveforms in the real-time feedback data. Taking the step response waveform as an example, transient response characteristics are directly measured in the time domain, including the delay time from the application of excitation to the start of response change, the overshoot amplitude after the response exceeds the steady-state value, and the settling time required for the fluctuation to stabilize within the allowable error range, reflecting the dynamic performance of the circuit. A frequency domain transformation is performed on the waveform, converting the time signal into a frequency signal. Fourier analysis is used to obtain the waveform's spectral distribution, and then the harmonic distortion rate is calculated. If component damage in the circuit leads to nonlinear distortion, the harmonic content will increase significantly, and the distortion rate will rise accordingly. For the waveform data obtained from the frequency sweep test, the quality factor at each frequency is calculated point-by-point, forming a curve showing the quality factor changing with frequency, reflecting the loss of the resonant circuit. The primary feature set is a collection of time-frequency domain features extracted from the voltage and current waveforms, including the harmonic distortion rate, the quality factor variation curve, and transient response time characteristics.

[0044] Directly using all frequency points of the impedance spectrum leads to excessively high feature dimensionality, easily causing model overfitting and incurring enormous computational costs. Feature decomposition is performed on the impedance spectrum in real-time feedback data, extracting the most representative and discriminative key features from massive impedance data. Characteristic frequency points in the impedance spectrum are identified, such as the resonant frequency where the impedance amplitude reaches its maximum, the frequency corresponding to the zero-crossing point of the phase angle, and the frequency points where the impedance amplitude changes most drastically. The actual impedance spectrum is fitted to the aforementioned equivalent circuit model to find the equivalent circuit parameters that best approximate the actual impedance spectrum, including resistance, capacitance, inductance, and quality factor. The secondary feature set is a collection of features extracted from the impedance spectrum, including impedance amplitude, phase angle, and equivalent circuit parameters at characteristic frequency points.

[0045] Primary and secondary features are concatenated in a fixed order to form a one-dimensional composite feature vector. The support vector machine (SVM) model needs to be trained beforehand. During offline training, a large number of circuit board samples in known states are collected, including normal boards and various faulty boards. The aforementioned features are extracted from each sample and labeled with the true fault type. These labeled feature vectors are then input into the SVM for training. The architecture of this SVM is a conventional SVM, which will not be detailed here. The training objective is to find an optimal classification hyperplane that separates samples of different categories as much as possible in the feature space. When the real-time feature vector of the board under test is input into the trained SVM model, the probability of it belonging to each fault type is calculated based on the position of the feature vector in the feature space. The output is a fault probability distribution matrix, where each row represents a possible fault type and each column represents the corresponding probability value.

[0046] The fault probability distribution matrix is ​​compared with a pre-set fault determination threshold. The fault determination threshold is a pre-defined probability threshold value, typically set based on the requirements for false alarm and false negative rates in the actual application scenario. When the probability of a certain fault type output by the model exceeds this threshold, the fault is determined to have occurred, thereby improving the reliability of the judgment and reducing false alarms. The probability distribution matrix is ​​traversed to determine if the preset threshold is exceeded. If it exceeds the threshold, it is determined to be a fault of that type; if the highest probability value is still below the threshold, it is determined to be either indeterminate or normal. Once a fault is determined, detailed fault detection results are generated, including fault coordinates, fault type code, and fault confidence weight. The fault type code uses a predefined encoding to represent the fault type, such as F001 representing a power short circuit, F102 representing a signal line open circuit, and F205 representing capacitor aging. The fault confidence weight is the fault probability value output by the model, used to measure the reliability of the judgment; for example, a confidence level of 98% indicates very reliable, and 82% indicates relatively reliable. For example, if the preset fault detection threshold is set to 80%, and the highest probability in the fault probability distribution matrix is ​​87.5%, the corresponding fault type is output capacitor aging exceeding the threshold. A fault detection result is generated, and based on the abnormal characteristic frequency of 127kHz in the impedance spectrum, the location of capacitor C23 at the board's output end is determined (coordinates: X=45.2mm, Y=32.8mm); fault type code: F205; fault confidence weight: 87.5%. This result is then transmitted to the next stage to guide the self-healing repair step, specifically the heating repair of the area near capacitor C23 using self-healing materials.

[0047] By using methods such as harmonic analysis, quality factor variation, and equivalent circuit parameter extraction, latent faults such as capacitor value drift, inductor core microcracks, and increased solder joint contact resistance can be detected, allowing for early warning and repair before the faults worsen.

[0048] The self-heating repair analysis module 13 is used to adjust the transmission frequency to generate a high-frequency alternating magnetic field when the fault detection result detects an open circuit fault, trigger the circuit board under test to perform self-heating repair analysis, and generate a self-healing repair scheme. The circuit board under test is coated with a self-healing material that can be locally heated to repair the open circuit. The heating parameters of the self-healing material are dynamically adjusted according to the open circuit location and the degree of fault.

[0049] Furthermore, the self-heating repair analysis module 13 in the railway passenger car circuit board detection system also includes: a mode switching unit, used to control the magnetic resonance coupling system to switch from a low-frequency energy transmission mode to a high-frequency magnetic field excitation mode when the fault mode in the fault detection result is an open circuit, raising the operating frequency to a preset high-frequency band, and generating a spatially focusable high-frequency alternating magnetic field above the circuit board under test; an energy control unit, used to control the energy of the high-frequency alternating magnetic field to concentrate on the open circuit area coated with self-healing material according to the fault location coordinates in the fault detection result; and a scheme generation unit, used to simulate and calculate the minimum temperature, heating rate, and holding time required for reliable repair based on the open circuit gap width, line layer depth, and confidence level in the fault detection result, combined with the phase transition dynamics model of the self-healing material, and generate the self-healing repair scheme including the target temperature curve, magnetic field strength distribution, and action sequence.

[0050] Furthermore, the self-heating repair analysis module 13 in the railway passenger car circuit board detection system further includes: a model coupling subunit, used to construct a three-dimensional electromagnetic-thermal coupling model of the broken circuit area, including the substrate, conductors, and self-healing material coating, based on the fault coordinates and board layer information in the fault detection results; a target determination subunit, used to set a repair target in the three-dimensional electromagnetic-thermal coupling model of the broken circuit area based on the broken gap width and confidence weight, and generate the spatiotemporal distribution of thermal energy required to achieve the repair target through simulation calculation; and a magnetic field conversion subunit, used to convert the spatiotemporal distribution of thermal energy into magnetic field parameters, and generate the target temperature field, magnetic field intensity map, and heating sequence required to control the high-frequency alternating magnetic field to generate a predetermined temperature gradient in the broken circuit area.

[0051] Specifically, if the fault type code in the fault detection results is identified as an open circuit fault, the repair process is initiated. The operating mode of the magnetic resonance coupling system is changed, gradually reducing the low-frequency output of the control drive circuit while simultaneously activating the high-frequency excitation module. The operating frequency is increased from the low-frequency band to a preset high-frequency band, typically set between 1MHz and 3MHz. This frequency band is chosen because the eddy current loss and hysteresis loss generated by the high-frequency magnetic field in the ferrite nanoparticles are the most significant, resulting in the highest heating efficiency. Furthermore, the magnetic field penetration depth is sufficient to cover the common thickness of the circuit board. After the mode switch is completed, the transmitting coil no longer transmits electrical energy but generates a high-frequency, high-intensity alternating magnetic field, covering the area where the board under test is located. However, at this point, the magnetic field is uniformly distributed and has not yet been focused on a specific location.

[0052] A uniformly distributed magnetic field would heat the entire board, damaging areas where the self-healing material applied could not be repaired. Therefore, the magnetic field energy needed to be concentrated at the fault location. By reading the fault location coordinates from the fault detection results, the amplitude and phase of the excitation signal required for each transmitting coil were calculated. This ensured that the magnetic fields generated by all coils were in phase and maximized at the fault point, while canceling each other out in other areas and minimizing their intensity. After focusing control, the energy of the high-frequency alternating magnetic field was precisely concentrated in the open circuit area corresponding to the fault location coordinates. The magnetic field strength in this area reached its maximum, while the magnetic field strength in the surrounding areas significantly decreased, achieving precise heating of the open circuit area.

[0053] After the magnetic field is focused, specific heating parameters need to be determined, including the temperature to be reached, the heating rate, and the holding time. These parameters are based on the fault detection results, including the width of the circuit breaker gap, the depth of the circuit layer, and the confidence level. The width of the circuit breaker gap determines the amount of material to be filled; a wider gap requires more heating time and a larger amount of material to melt. The depth of the circuit layer determines the thickness the magnetic field needs to penetrate; a greater depth requires a stronger magnetic field. The confidence level is used to adjust the safety margin of the scheme; a high confidence level allows for more precise parameters, while a low confidence level requires a certain margin. A pre-stored self-healing material phase transition kinetic model is used, containing a series of physical parameters of the self-healing material: melting point temperature, glass transition temperature, thermal conductivity, specific heat capacity, density, viscosity as a function of temperature, surface tension, etc. By substituting parameters such as the circuit breaker gap width and the depth of the circuit layer into the model, combined with material properties, thermodynamic and fluid dynamic simulations are performed. The goal of the simulation is to find a set of heating parameters, including minimum temperature, heating rate, and holding time, so that the temperature in the broken area reaches above the melting point of the self-healing material, allowing the material to fully melt and fill the broken gap through surface tension and capillary flow, without damaging the circuit board substrate or adjacent components due to excessive temperature. The phase transition kinetics model of the self-healing material takes the fault detection results as input, including the width of the broken gap, the depth of the circuit layer, and the confidence weight. It also calls upon the complete set of phase transition kinetic parameters of the self-healing material pre-stored in the database, including the complete curves of melting point temperature, glass transition temperature, thermal conductivity, specific heat capacity, density, viscosity as a function of temperature, and surface tension coefficient. These input parameters are sequentially fed into three interconnected simulation modules: First, the electromagnetic field simulation module calculates the penetration depth and energy deposition distribution of the high-frequency magnetic field in the multilayer structure based on the circuit layer depth and the fault area; then, this energy distribution is input as a heat source into the heat conduction simulation module, which, combined with the material's thermal conductivity and specific heat capacity, simulates the transient temperature field evolution of the fault area and its surroundings; finally, the temperature field results are substituted into the fluid dynamics simulation module, which, combined with the material's viscosity-temperature curve and surface tension coefficient, simulates the flow, spreading, and filling behavior of the molten self-healing material in the circuit breaker gap. These three modules are not simply executed sequentially, but rather iteratively solved under the dynamic coupling relationship where the temperature field affects viscosity, and viscosity changes, in turn, affect flow and heat transfer, until convergence. The training process of this simulation system requires the pre-construction of hundreds of sets of sample data, including different circuit widths, different line layer depths, and different material ratios. The actual repair process is observed through high-speed photography and infrared thermal imaging. The observation results are compared with the simulation results. Bayesian optimization methods are used to calibrate key empirical parameters in the model, such as the interface heat transfer coefficient and magnetocaloric conversion efficiency, so that the simulation accuracy reaches more than 95%.Comparative experiments verified that the repair success rate based solely on empirical formula heating was 65%, the success rate using only electromagnetic-thermal coupling simulation was 72%, and the success rate using only thermal-fluid coupling simulation was 88%. However, the three-field fully coupled simulation method used in this solution, combined with the dynamic adjustment of the safety margin based on the confidence weight in the fault detection results, increased the repair success rate to 98%. At the same time, it reduced the radius of the heat-affected zone from 8 mm when using empirical formulas to less than 3 mm, fully demonstrating that the deep coupling between fault characteristics and the material physics model brings about an overall technical effect far exceeding that of simply superimposing individual modules.

[0054] Based on the minimum temperature, heating rate, and holding time, a self-healing repair scheme is generated, including a target temperature curve, magnetic field strength distribution, and application timing. The target temperature curve, plotted with time on the horizontal axis and temperature on the vertical axis, defines the ideal trajectory that the temperature of the broken area should follow throughout the entire process from the start of heating to the end of cooling. The magnetic field strength distribution specifies the spatial magnetic field strength distribution that the transmitting coil array needs to generate during the repair process to ensure that the broken area receives the required heating power. The application timing specifies the time sequence of the entire repair process, including when the magnetic field is first applied, how the power changes, when to maintain a constant temperature, when to stop heating, and whether multiple pulses are required.

[0055] Based on the fault coordinates and board layer information from the fault detection results, the fault coordinates provide the precise location of the circuit break on the board plane, while the board layer information indicates which layer of the multilayer board the circuit break is located on. Detailed design data for the corresponding area of ​​this type of board is retrieved from the computer-aided design database, including the substrate material's stack-up structure, the thickness and width of the copper foil conductors, and the coverage of the solder mask layer. If precise design data cannot be retrieved, an approximate model is created using pre-stored typical parameters. In the simulation software, a three-dimensional electromagnetic-thermal coupling model of the circuit break region, including the substrate, conductors, and self-healing material coating, is constructed. For the substrate, the material properties are set based on the board layer information, including density, thermal conductivity, specific heat capacity, and relative permittivity. For multilayer boards, the substrate materials for different layers may be the same or different, requiring separate settings. For the conductors, the specific conductor is located based on the fault coordinates. A geometric entity for this conductor segment is created in the model, and copper material properties are assigned, including electrical conductivity, thermal conductivity, density, and specific heat capacity. The circuit break gap is simulated by leaving a gap in the middle of the conductor, with the gap width set to the value provided by the fault detection results. The self-healing material coating is a geometric entity created within the broken area and a certain surrounding region, with the coating thickness set according to the process design. After the model is built, it needs to be meshed, discretizing the continuous geometric entity into hundreds of thousands or even millions of tiny units, each with material properties and coordinate information.

[0056] The frequency of the high-frequency alternating magnetic field is set, such as 2.45 MHz, and an initial excitation condition is applied, such as applying a uniform magnetic field on the model boundary. By solving Maxwell's equations, the simulation software calculates the magnetic field intensity distribution at various points within the model. Since the self-healing material contains ferrite nanoparticles, which have high permeability and certain electrical conductivity, eddy current losses and hysteresis losses are generated in the alternating magnetic field, and these two losses are converted into heat. Based on the magnetic field intensity and material properties, the simulation software calculates the heat generation rate on each grid cell, i.e., the heat generated per unit volume per unit time. The heat generation rate distribution is input as a heat source into the heat conduction simulation module. The heat conduction simulation is based on Fourier's law of heat conduction, considering the thermal conductivity, specific heat capacity, and density of the substrate, wires, and self-healing material coating, as well as the interfacial thermal resistance between them, to calculate the temperature distribution evolution throughout the model over time. Through the above iterative simulation, the spatiotemporal distribution of thermal energy required to meet the repair target is obtained. The repair target is set based on the width of the circuit breaker gap and the confidence weight. The wider the circuit breaker gap, the greater the heat required, and the higher the target temperature. The confidence level weight is used to adjust the safety margin: when the confidence level is high, the target can be set more precisely, such as just reaching 10°C above the melting point of the self-healing material; when the confidence level is low, a margin needs to be left, such as setting it to 20°C above the melting point, to ensure that the repair can be completed even if there is a detection error.

[0057] Using the spatiotemporal distribution of thermal energy as the optimization objective, the spatial distribution of the magnetic field required to generate this thermal energy distribution is derived. Based on electromagnetic field theory, the spatial distribution of the external high-frequency magnetic field intensity required to generate this eddy current is further derived from the eddy current density distribution, thus generating a magnetic field intensity map. This map indicates the required magnetic field intensity values ​​at different locations in space above the fault area. Combining the electromagnetic characteristic parameters of the coil, the spatial magnetic field intensity map is converted into requirements for the current magnitude, phase, and frequency of the specific drive coil, and a sequence of these current parameters changing over time is planned, forming a precise heating timing control command set. Based on the magnetic field parameters, the required target temperature field, magnetic field intensity map, and heating sequence are generated. The target temperature field is the ideal spatial temperature distribution that the fault area needs to achieve during the repair process; the magnetic field intensity map describes the spatial distribution of the intensity of the high-frequency alternating magnetic field within the repair area; the heating sequence is the temporal order and duration of each step in the repair process, including the start and end times of magnetic field application, power adjustment time points, and constant temperature maintenance time.

[0058] The circuit board under test is coated with a self-healing material that can locally heat up and repair broken circuits. Under the influence of a high-frequency alternating magnetic field, it locally heats up to a molten state, flows to fill the broken gap, and restores conductivity upon cooling. The heating parameters of the self-healing material are dynamically adjusted according to the location and severity of the broken circuit. In other words, the heating parameters are specific values ​​that control the repair process, including target temperature, heating rate, holding time, magnetic field strength, and application sequence. These parameters are not fixed values ​​but are dynamically calculated based on the location and severity of the broken circuit.

[0059] Furthermore, the self-heating repair analysis module 13 in the railway passenger car circuit board detection system also includes: a deviation calculation unit, used to monitor the surface temperature distribution of the broken area in real time through infrared thermometry during the repair process, compare the surface temperature distribution with the target temperature curve, and obtain the temperature distribution deviation; and a closed-loop control unit, used to dynamically correct the magnetic field strength distribution and action sequence based on the temperature distribution deviation, forming a closed-loop temperature control.

[0060] Specifically, an infrared thermal imager or infrared temperature sensor array is mounted above the circuit board under test, covering the entire open circuit area and its surroundings. During the repair process, extremely high frame rates are used to continuously acquire surface infrared radiation images of the open circuit area. Each frame contains hundreds or even thousands of pixels, with each pixel corresponding to the temperature value of a tiny area on the board surface. Using calibrated emissivity parameters, these infrared radiation intensity data are converted into temperature values ​​in real time, thus obtaining the two-dimensional surface temperature distribution of the open circuit area at each moment.

[0061] The real-time measured surface temperature distribution is compared point-by-point with the target temperature curve. This involves two aspects: first, whether the temperature value at the center point or key points matches the target curve; and second, whether the temperature gradient distribution across the entire area is consistent with the target temperature field. The comparison results are output as temperature distribution deviations, including the center temperature deviation value, the temperature deviation value at each point, and the degree of deviation in the temperature gradient.

[0062] Based on the temperature distribution deviations, analyze the causes of these deviations and determine how to adjust subsequent magnetic field parameters to correct them. Temperature deviations may be caused by various factors, such as a slight offset in the coil position leading to low magnetic field focusing accuracy, uneven thickness of the self-healing material coating causing local differences in heat generation rate, or changes in ambient temperature affecting heat dissipation conditions.

[0063] Based on the type and degree of deviation, the magnetic field parameters requiring correction are determined. The main objects of correction include the magnetic field strength distribution and the timing of application. Correction of the magnetic field strength distribution is achieved by adjusting the amplitude and phase of the excitation current in each transmitting coil. Correction of the application timing involves adjustments in the time dimension. For example, if the heating rate is too slow, the magnetic field strength is increased earlier to accelerate the subsequent heating process; if temperature fluctuations occur during the holding phase, the magnetic field strength is adjusted in real time to maintain a constant temperature; if excessive temperature is detected as potentially damaging the substrate, heating is stopped early or the power is reduced. Based on these decisions, specific correction instructions are generated, including adjustments to the current amplitude and phase of each coil, as well as power settings for subsequent time periods.

[0064] Once the correction command is generated, it is immediately sent to the drive circuit of the transmitting coil, which responds within microseconds, adjusting the excitation current of each coil to change the magnetic field strength distribution and timing. After the new magnetic field acts on the disconnected area, the heat generation rate distribution of the self-healing material changes, thus affecting the temperature field evolution. The infrared thermography system continues to acquire temperature data at a high frame rate, comparing the new temperature distribution with the target curve again to obtain new deviations, which are then analyzed and corrected repeatedly, at a frequency of tens or even hundreds of times per second. Through closed-loop regulation, even with initial magnetic field focusing errors, material inhomogeneities, and environmental disturbances, the actual temperature can always be controlled near the target temperature curve. At the end of the heating phase, the center temperature is ensured to accurately reach the preset target value; during the holding phase, real-time fine-tuning maintains a constant temperature, with fluctuations typically controlled within ±2 degrees Celsius; when entering the cooling phase, heating stops, but the temperature drop process continues to be monitored to ensure the cooling rate meets requirements and avoids excessively rapid cooling that could lead to thermal stress. The entire closed-loop control process continues until the repair operation is completed, ensuring that the self-healing material heats up uniformly within the circuit breaker area and precisely follows the target temperature curve. Temperature is monitored in real time and compared with the target curve, dynamically correcting the magnetic field parameters to ensure that the actual temperature always follows the preset trajectory. This avoids temperature deviations caused by various disturbances in open-loop control. By monitoring the temperature distribution throughout the area and making timely adjustments, a uniform temperature rise within the circuit breaker area is ensured.

[0065] The comparison and verification module 14 is used to obtain the impedance characteristics of the repaired area by multi-band impedance scanning after the circuit is repaired by heating based on the self-healing repair scheme, and compare and verify it with the pre-built performance attenuation model.

[0066] Furthermore, the comparison and verification module 14 in the railway passenger car circuit board detection system further includes: a frequency sweep measurement unit, used to control a vector network analyzer to perform frequency sweep measurement on the repair area within a preset frequency band after completing the current heating sequence; a result conversion unit, used to obtain the S-parameter measurement results of the repair area at multiple discrete frequency points, convert the S-parameter measurement results into complex impedance values, and generate the impedance characteristics of the repair area; an impedance characteristic error determination unit, used to extract the reference impedance characteristics and allowable error range of the same type of intact circuit board from the pre-built performance attenuation model; an impedance deviation calculation unit, used to compare the impedance characteristics of the repair area with the reference impedance characteristics point by point, and calculate the impedance deviation value at each frequency point in combination with the allowable error range; and a deviation fusion unit, used to weight and fuse the impedance deviation values ​​at all frequency points to generate a comprehensive repair quality index.

[0067] Specifically, after closed-loop repair according to the self-healing repair scheme, the complete electrical characteristics of the repaired area are obtained through multi-band impedance scanning and compared with benchmark data to generate a comprehensive evaluation index to determine whether the repair meets the requirements. Once the heating sequence is complete and the temperature of the repaired area drops to ambient temperature, the repair effect verification process is initiated. At this time, the magnetic resonance coupling system switches back to measurement mode, but no longer supplies power; instead, it is dedicated to signal transmission. The control system sends instructions to the vector network analyzer to set the measurement parameters. The preset frequency band is determined based on the circuit characteristics and fault features of the specific board model. For example, for a power management board, it is set to 100Hz to 10MHz. Within this frequency band, the system selects a set of discrete frequency points, typically evenly distributed on logarithmic coordinates, such as ten points per decade, to ensure sufficient measurement resolution in both low and high frequency bands. The vector network analyzer applies a low-power sweep excitation signal to the repaired area of ​​the board under test through a coupler connected to the transmitter. The excitation signal is transmitted to the board through the magnetic coupling channel and acts on the repaired area. Since the circuit board is powered off at this point, only the repair area and its connecting lines are in passive measurement mode. The impedance characteristics of the repair area affect signal reflection. The reflected signal returns to the transmitter through the magnetic coupling channel and is captured by the receiving port of the vector network analyzer. For each selected discrete frequency point, the vector network analyzer measures the amplitude ratio and phase difference between the incident and reflected signals to obtain the S-parameters at that frequency point, mainly the S11 parameter, i.e., the reflection coefficient. The entire frequency sweep measurement process usually takes only a few seconds to complete.

[0068] The S-parameters directly output by a vector network analyzer are the reflection coefficients, expressed as amplitude and phase, or real and imaginary parts. Based on transmission line theory and circuit principles, using a preset reference impedance value, the S11 parameters at each frequency point are converted into corresponding complex impedance values. The converted complex impedance contains two parts: the real part represents the resistive component, and the imaginary part represents the reactive component. A positive reactance indicates inductive behavior, and a negative reactance indicates capacitive behavior. After conversion, a set of data is obtained, with corresponding real and imaginary impedance values, or equivalent impedance amplitude and phase angle, for each measurement frequency point. S-parameters are scattering parameters, a set of parameters describing the signal reflection and transmission characteristics in an RF network. For single-port measurements, the focus is mainly on the S11 parameters, i.e., the reflection coefficient, which has a definite conversion relationship with the input impedance of the measured object. The complex impedance value is the impedance expressed in complex form, containing a real part (resistance) and an imaginary part (reactance, including inductive and capacitive reactance). The complex impedance describes the electrical behavior of the repaired area under AC signals.

[0069] The reference impedance characteristics and allowable error range of a similar, intact circuit board are extracted from a pre-built performance degradation model. The reference impedance characteristics are the curve showing the standard impedance value that the board should possess during its corresponding usage phase, changing with frequency. The allowable error range is the maximum permissible deviation between the repaired impedance characteristics and the reference impedance characteristics, typically including amplitude and phase deviations. For each measurement frequency point, the measured impedance amplitude and phase are compared with the values ​​of the reference curve at that frequency point, and the deviation value is calculated. After point-by-point comparison of all frequency points, a set of deviation value sequences is obtained, with each frequency point corresponding to one amplitude deviation and one phase deviation. Deviation calculation usually considers amplitude and phase, or directly calculates the complex difference. Then, it is checked whether the deviation falls within the allowable error range defined by the model.

[0070] The deviation values ​​of all frequency points are integrated into a comprehensive evaluation index. According to a pre-defined weighting scheme, the amplitude and phase deviations of each frequency point are multiplied by their corresponding weighting coefficients, and then summed or averaged to obtain the comprehensive deviation value. The weighting scheme is predetermined based on circuit design characteristics and failure mode analysis and stored in the system database. The comprehensive deviation value is normalized and converted into a comprehensive repair quality index between zero and one hundred. Generally, a higher index indicates better repair quality. Finally, a comprehensive repair quality index between 0 and 1 is calculated, reflecting the degree of conformity between the impedance characteristics of the repaired area and the baseline characteristics, used to determine the success of the repair. By comparing with a pre-built performance degradation model, it not only determines whether the circuit is currently conductive but also assesses whether the electrical characteristics after repair conform to the normal state of a board with this service life, indirectly predicting the long-term reliability of the repaired area. Simultaneously considering amplitude and phase deviations, covering multiple frequency bands, and assigning different weights avoids the one-sidedness of single-frequency point testing and can detect anomalies in local frequency bands.

[0071] The parameter optimization module 15 is used to automatically optimize the high-frequency alternating magnetic field parameters based on the impedance characteristic difference for iterative repair if the verification fails, until the verification passes.

[0072] Furthermore, the parameter optimization module 15 in the railway passenger car circuit board detection system further includes: an impedance difference identification unit, used to determine that the verification has failed when the comprehensive repair quality score is lower than the verification threshold, analyze the difference between the impedance characteristics of the repair area and the reference impedance characteristics, and identify the frequency bands and components of impedance mismatch; an optimized parameter back-reasoning unit, used to map the frequency bands and components of impedance mismatch to the corresponding circuit physical parameter defects, and back-reason the direction and amplitude of the high-frequency alternating magnetic field parameters that need to be optimized according to the type and degree of the physical parameter defects, forming optimized high-frequency alternating magnetic field parameters; and a repair quality evaluation unit, used to re-execute the self-healing repair scheme using the optimized high-frequency alternating magnetic field parameters, and after each re-execution of the repair, re-acquire the impedance characteristics of the repair area and calculate a new comprehensive repair quality index until the new comprehensive repair quality index reaches or exceeds the quality threshold.

[0073] Specifically, when the overall repair quality index falls below a preset verification threshold, the repair is deemed to have failed verification, and an iterative repair process is immediately initiated. The verification threshold is a pre-set critical value for the overall quality index used to determine whether a repair is satisfactory. For example, 85 points is set as the passing score; below this value, the repair is considered unsuccessful. The impedance characteristic data after the repair, along with the reference impedance characteristic data for comparison, are retrieved. Through point-by-point comparison, it is identified which frequency bands and components exhibit significant deviations. The frequency bands and components with impedance mismatches are specific frequency ranges where the measured impedance characteristics differ significantly from the reference impedance characteristics, discovered during the point-by-point comparison process. It is also determined whether amplitude deviation or phase deviation is predominant within these frequency bands. For example, in the low-frequency range, such as below 1Hz, the measured impedance amplitude is generally more than 15% higher than the reference value, while the phase deviation is smaller. According to electromagnetic field theory, the impedance in the low-frequency range mainly reflects DC resistance and low-frequency loss; a higher amplitude indicates that the resistance in the repaired area is too high. In addition, considering the characteristics of self-healing materials, a high resistance may indicate poor interfacial contact between the self-healing material and the wire, or that the self-healing material itself is not fully densified.

[0074] An electrical characteristic-physical defect mapping table was pre-established, associating different types of electrical deviations with possible physical defect types. For example, high low-frequency resistance indicates poor interface contact or incomplete material fusion; low low-frequency resistance indicates excessive self-healing material spreading, leading to short-circuit risk; resonant frequency shift indicates changes in equivalent capacitance or inductance, potentially involving the material's dielectric constant or geometry; increased high-frequency loss indicates microscopic defects or voids in the material itself; and phase anomalies indicate changes in parasitic parameters, such as additional inductance or capacitance. Through table lookup and comprehensive analysis, the main physical defect types and degrees of this repair failure were determined. For example, if poor interface contact was identified, the contact resistance was approximately 0.15Ω higher than normal.

[0075] After determining the type and extent of physical defects, the process works backward to deduce how to adjust the magnetic field parameters to compensate for these defects. Based on the type of physical defect, the dimensions of the magnetic field parameters that need optimization are determined. For example, poor interface contact indicates that the heating temperature may be insufficient or the holding time inadequate, resulting in insufficient melting and flow of the self-healing material, failing to form a good bond with the broken wire. The self-healing repair scheme used in the first repair is retrieved, including the target temperature curve, magnetic field strength distribution, and application sequence. The parameter adjustment amount is calculated based on the defect extent. For example, if the contact resistance is 0.15Ω higher than normal, simulation calculations based on the resistivity-temperature relationship and flow characteristics of the self-healing material indicate that the maximum temperature needs to be increased by 8°C, or the holding time extended by 1.5s, to ensure sufficient interface fusion. The magnetic field strength distribution is adjusted based on the spatial characteristics of the defect. If the defect manifests as poor contact due to a lower temperature on one side, it may be necessary to adjust the amplitude and phase of each coil to slightly shift the magnetic field focus point, or to make the energy distribution more biased towards the defect side. The application sequence is adjusted based on the heating rate and holding time. If the defect indicates that the temperature rise was too rapid, causing the material to solidify before it had sufficient time to flow, then the heating rate may need to be reduced; if the defect indicates that the holding time was insufficient, then the holding period should be extended. A set of optimized high-frequency alternating magnetic field parameters was generated through a series of calculations. This was an adjustment based on the original scheme, rather than a complete regeneration. It retained the effective parts of the original scheme while specifically addressing its shortcomings.

[0076] After the optimized magnetic field parameters are generated, the second repair process is immediately initiated using the optimized parameters, and the heating repair is re-executed according to the closed-loop control method. During the second repair, the infrared temperature measurement system continues to monitor the temperature in real time to ensure that the actual temperature evolves according to the new target temperature curve. Since the parameters have been optimized to address the reasons for the failure of the first repair, the second repair is expected to achieve better results. After the repair is completed, the verification process is executed again, and a multi-band impedance scan is performed again to calculate a new comprehensive repair quality index. The new index is compared with the quality threshold. If the new index reaches or exceeds the threshold, the repair is considered successful, the process ends, and the final conclusion of successful repair is output. If the new index is still below the threshold, the first stage is resumed to analyze the difference in impedance characteristics after the second repair, identify new or residual defects, optimize the parameters again, and perform a third repair. This cycle continues until the comprehensive repair quality index meets the standard, or the preset maximum number of iterations is reached, at which point an alarm is triggered for manual intervention. This iterative mechanism ensures that even if the first repair is imperfect, satisfactory repair quality is eventually achieved through self-learning and adjustment; that is, iteration stops only when the new comprehensive repair quality index reaches or exceeds the quality threshold, thus improving the overall repair success rate.

[0077] By mapping electrical test defects back to their physical causes and deriving process parameters in reverse, each parameter adjustment has a clear objective and theoretical basis, avoiding blind trial and error and improving optimization efficiency. Combined with a comprehensive repair quality index, iterative repair can be performed to ensure that open circuits are fully conductive and meet electrical performance standards. From fault detection and repair plan generation to closed-loop heating control, effect verification, and iterative optimization, a closed loop is formed, enabling high-quality circuit board repair without any manual intervention.

[0078] In summary, the railway passenger car circuit board inspection system provided in this application has the following technical effects: It transmits the required electrical energy, detection signals, and feedback data of the circuit board non-contactly through magnetic resonance coupling technology; it extracts dynamic features based on real-time feedback data to determine the circuit status and generate fault detection results; when an open circuit fault is detected, it adjusts the transmission frequency to generate a high-frequency alternating magnetic field, triggering the self-healing material on the circuit board to heat up and repair the open circuit; after repair, it scans the impedance of the repaired area and compares it with a pre-built performance attenuation model to determine whether the repair is effective; if the verification fails, it automatically adjusts the high-frequency magnetic field parameters and reheats the repair until it is effective, reducing interface wear, actively repairing hidden damage caused by open circuits, and reducing manual intervention, thereby improving the maintenance efficiency of railway passenger car circuit boards.

[0079] Example 2: Based on the same inventive concept as the railway passenger car circuit board detection system in Example 1, this application also provides a railway passenger car circuit board detection method. Please refer to the appendix. Figure 2The railway passenger car circuit board detection method includes: non-contactly transmitting the electrical energy required for detection to the circuit board under test via magnetic resonance coupling, synchronously transmitting detection signals and receiving real-time feedback data from the circuit board under test; performing dynamic feature extraction based on the real-time feedback data, and performing circuit fault detection based on the extracted dynamic features to generate fault detection results; when the fault detection results detect an open circuit fault, adjusting the transmission frequency to generate a high-frequency alternating magnetic field, triggering the circuit board under test to perform self-heating repair analysis, and generating a self-healing repair scheme, wherein the circuit board under test is coated with a self-healing material that can locally heat and repair the open circuit, and the heating parameters of the self-healing material are dynamically adjusted according to the open circuit location and fault degree; after heating and repairing the open circuit based on the self-healing repair scheme, obtaining the impedance characteristics of the repair area through multi-band impedance scanning and comparing and verifying it with a pre-built performance attenuation model; if the verification fails, automatically optimizing the high-frequency alternating magnetic field parameters based on the impedance characteristic differences for iterative repair until the verification passes.

[0080] Furthermore, the non-contact transmission of the required electrical energy to the circuit board under test via magnetic resonance coupling includes: matching the resonant frequencies of the transmitting coil and the receiving coil in a preset low-frequency band to construct a non-contact energy transmission channel; during the energy transmission process, sampling the load current of the circuit board under test in real time, calculating the excitation voltage compensation amount of the transmitting coil based on the change of the load current, and generating a voltage adjustment signal; controlling the driving circuit of the transmitting coil based on the voltage adjustment signal to output the excitation voltage after real-time compensation, thereby generating a dynamically stable power supply output.

[0081] Furthermore, the synchronous transmission of detection signals and reception of real-time feedback data from the circuit board under test includes: after the establishment of the dynamically stable power supply output, encoding the detection signal containing step, frequency sweep, and pulse sequences into a digital command stream; modulating the digital command stream onto the carrier of the non-contact energy transmission channel to generate a composite modulation signal and transmitting it to the circuit board under test; receiving the modulation response signal returned by the circuit board under test through the receiving coil coupling, and demodulating and converting the modulation response signal into analog and digital signals to generate real-time feedback data containing voltage waveforms, current waveforms, impedance spectra, and transient characteristics.

[0082] Furthermore, the step of extracting dynamic features based on the real-time feedback data and detecting circuit faults based on the extracted dynamic features to generate fault detection results includes: performing time-frequency domain joint transformation on the voltage and current waveforms in the real-time feedback data to extract harmonic distortion rate, quality factor variation curves, and transient response time features to form a primary feature set; performing feature decomposition on the impedance spectrum to extract impedance amplitude, phase angle, and equivalent circuit parameters at characteristic frequency points to form a secondary feature set; fusing the primary and secondary feature sets and inputting them into a support vector machine classification model, from which the classification model outputs a fault probability distribution matrix; and generating fault detection results based on the fault probability distribution matrix and a preset fault judgment threshold, wherein the fault detection results include fault coordinates, fault type codes, and fault confidence weights.

[0083] Furthermore, when the fault detection result detects an open circuit fault, the transmission frequency is adjusted to generate a high-frequency alternating magnetic field, triggering the circuit board under test to perform self-heating repair analysis and generate a self-healing repair scheme. This includes: when the fault mode in the fault detection result is an open circuit, controlling the magnetic resonance coupling system to switch from a low-frequency energy transmission mode to a high-frequency magnetic field excitation mode, raising the operating frequency to a preset high-frequency band, and generating a spatially focusable high-frequency alternating magnetic field above the circuit board under test; according to the fault location coordinates in the fault detection result, controlling the energy of the high-frequency alternating magnetic field to concentrate on the open circuit area coated with self-healing material; based on the open circuit gap width, line layer depth, and confidence level in the fault detection result, combined with the phase transition dynamics model of the self-healing material, simulating and calculating the minimum temperature, heating rate, and holding time required for reliable repair, and generating a self-healing repair scheme that includes a target temperature curve, magnetic field strength distribution, and action sequence.

[0084] Furthermore, the generation of the self-healing repair scheme, which includes the target temperature curve, magnetic field strength distribution, and action sequence, includes: constructing a three-dimensional electromagnetic-thermal coupling model of the broken circuit region, including the substrate, conductors, and self-healing material coating, based on the fault coordinates and plate information in the fault detection results; setting a repair target in the three-dimensional electromagnetic-thermal coupling model of the broken circuit region based on the broken gap width and confidence weight, and generating the spatiotemporal distribution of thermal energy required to achieve the repair target through simulation calculation; converting the spatiotemporal distribution of thermal energy into magnetic field parameters, and generating the target temperature field, magnetic field strength map, and heating sequence required to control the high-frequency alternating magnetic field to generate a predetermined temperature gradient in the broken circuit region.

[0085] Furthermore, the railway passenger car circuit board detection method includes: during the repair process, real-time monitoring of the surface temperature distribution of the broken area by infrared thermometry, comparing the surface temperature distribution with the target temperature curve to obtain the temperature distribution deviation; and dynamically correcting the magnetic field strength distribution and action sequence based on the temperature distribution deviation to form a closed-loop temperature control.

[0086] Furthermore, based on the self-healing repair scheme, after heating to repair the open circuit, the impedance characteristics of the repaired area are obtained through multi-band impedance scanning and compared with a pre-built performance degradation model for verification. This includes: after completing the current heating sequence, controlling a vector network analyzer to perform frequency sweep measurements on the repaired area within a preset frequency band; obtaining the S-parameter measurement results of the repaired area at multiple discrete frequency points, converting the S-parameter measurement results into complex impedance values ​​to generate the impedance characteristics of the repaired area; extracting the reference impedance characteristics and allowable error range of a similar intact board from the pre-built performance degradation model; comparing the impedance characteristics of the repaired area with the reference impedance characteristics point by point, and calculating the impedance deviation value at each frequency point in conjunction with the allowable error range; and weighting and fusing the impedance deviation values ​​at all frequency points to generate a comprehensive repair quality index.

[0087] Furthermore, if the verification fails, the high-frequency alternating magnetic field parameters are automatically optimized based on impedance characteristic differences for iterative repair until the verification passes. This includes: when the comprehensive repair quality score is lower than the verification threshold, the verification is deemed to have failed; the difference between the impedance characteristics of the repaired area and the reference impedance characteristics is analyzed to identify the frequency bands and components of impedance mismatch; the frequency bands and components of impedance mismatch are mapped to the corresponding circuit physical parameter defects; based on the type and degree of the physical parameter defects, the direction and amplitude of the high-frequency alternating magnetic field parameters that need to be optimized are deduced in reverse to form optimized high-frequency alternating magnetic field parameters; the self-healing repair scheme is re-executed using the optimized high-frequency alternating magnetic field parameters; after each re-execution of the repair, the impedance characteristics of the repaired area are re-acquired and a new comprehensive repair quality index is calculated until the new comprehensive repair quality index reaches or exceeds the quality threshold.

[0088] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Figure 1 The railway passenger car circuit board detection system and specific examples in Embodiment 1 are also applicable to the railway passenger car circuit board detection method in this embodiment. Through the foregoing detailed description of the railway passenger car circuit board detection system, those skilled in the art can clearly understand the railway passenger car circuit board detection method in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

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

[0090] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A railway passenger car circuit board detection system, characterized in that, include: The data transmission module is used to transmit the electrical energy required for testing to the circuit board under test non-contactly via magnetic resonance coupling, synchronously transmit the testing signal, and receive the real-time feedback data from the circuit board under test. The circuit fault detection module is used to extract dynamic features based on the real-time feedback data, and to detect circuit faults based on the extracted dynamic features, thereby generating fault detection results. The self-heating repair analysis module is used to adjust the transmission frequency to generate a high-frequency alternating magnetic field when the fault detection result detects an open circuit fault, triggering the circuit board under test to perform self-heating repair analysis and generate a self-healing repair scheme. The circuit board under test is coated with a self-healing material that can be locally heated to repair the open circuit. The heating parameters of the self-healing material are dynamically adjusted according to the location of the open circuit and the degree of the fault. The comparison and verification module is used to obtain the impedance characteristics of the repaired area by multi-band impedance scanning after the circuit is repaired by heating based on the self-healing repair scheme, and compare and verify it with the pre-built performance attenuation model. The parameter optimization module is used to automatically optimize the high-frequency alternating magnetic field parameters based on the impedance characteristic differences and perform iterative repairs until the verification passes.

2. The railway passenger car circuit board detection system as described in claim 1, characterized in that, The data transmission module includes: The frequency band matching unit is used to match the resonant frequencies of the transmitting coil and the receiving coil in a preset low-frequency band to build a non-contact energy transmission channel. The compensation calculation unit is used to sample the load current of the circuit board under test in real time during the energy transmission process, calculate the excitation voltage compensation amount of the transmitting coil based on the change of the load current, and generate a voltage adjustment signal. The circuit control unit is used to control the drive circuit of the transmitting coil based on the voltage regulation signal, and output the excitation voltage after real-time compensation to generate a dynamically stable power supply output.

3. The railway passenger car circuit board detection system as described in claim 2, characterized in that, The data transmission module further includes: The signal encoding unit is used to encode the detection signal containing step, frequency sweep and pulse sequence into a digital command stream after the dynamic and stable power supply output is established; The signal transmission unit is used to modulate the digital command stream onto the carrier wave of the non-contact energy transmission channel, generate a composite modulated signal, and transmit it to the circuit board under test. The demodulation and analog-to-digital conversion unit is used to receive the modulation response signal returned by the circuit board under test through the receiving coil, and to demodulate and convert the modulation response signal into analog-to-digital data to generate real-time feedback data including voltage waveform, current waveform, impedance spectrum and transient characteristics.

4. The railway passenger car circuit board detection system as described in claim 3, characterized in that, The circuit fault detection module includes: The primary feature extraction unit is used to perform time-frequency domain joint transformation on the voltage waveform and current waveform in the real-time feedback data, and extract the harmonic distortion rate, quality factor change curve, and transient response time features to form a primary feature set. The secondary feature extraction unit is used to perform feature decomposition on the impedance spectrum, extract the impedance amplitude, phase angle and equivalent circuit parameters at the feature frequency points, and form a secondary feature set. The feature fusion unit is used to fuse the primary feature set and the secondary feature set, input the fusion into the support vector machine classification model, and output the fault probability distribution matrix from the classification model. The fault determination unit is used to generate a fault detection result based on the fault probability distribution matrix and a preset fault determination threshold. The fault detection result includes fault coordinates, fault type code and fault confidence weight.

5. The railway passenger car circuit board detection system as described in claim 1, characterized in that, The self-heating repair analysis module includes: The mode switching unit is used to control the magnetic resonance coupling system to switch from low-frequency energy transmission mode to high-frequency magnetic field excitation mode when the fault mode in the fault detection result is open circuit, thereby raising the operating frequency to a preset high-frequency band and generating a spatially focusable high-frequency alternating magnetic field above the circuit board under test. An energy control unit is used to control the energy of the high-frequency alternating magnetic field to be concentrated on the circuit break area coated with self-healing material, based on the fault location coordinates in the fault detection results. The scheme generation unit is used to simulate and calculate the minimum temperature, heating rate and holding time required for reliable repair based on the open gap width, line layer depth and confidence level in the fault detection results, combined with the phase change dynamics model of the self-healing material, and generate the self-healing repair scheme including the target temperature curve, magnetic field strength distribution and action sequence.

6. The railway passenger car circuit board detection system as described in claim 5, characterized in that, The scheme generation unit includes: The model coupling subunit is used to construct a three-dimensional electromagnetic-thermal coupling model of the circuit breaker region, including the substrate, wires and self-healing material coating, based on the fault coordinates and board layer information in the fault detection results. The target determination subunit is used to set the repair target in the electromagnetic-thermal coupling model of the three-dimensional circuit breaker region according to the width of the circuit breaker gap and the confidence weight, and to generate the spatiotemporal distribution of thermal energy required to achieve the repair target through simulation calculation. The magnetic field conversion subunit is used to convert the spatiotemporal distribution of thermal energy into magnetic field parameters, and generate the target temperature field, magnetic field intensity map and heating sequence required to control the high-frequency alternating magnetic field to produce a predetermined temperature gradient in the circuit-breaking area.

7. The railway passenger car circuit board detection system as described in claim 6, characterized in that, The self-heating repair analysis module also includes: The deviation calculation unit is used to monitor the surface temperature distribution of the broken area in real time through infrared thermometry during the repair process, compare the surface temperature distribution with the target temperature curve, and obtain the temperature distribution deviation. A closed-loop control unit is used to dynamically correct the magnetic field strength distribution and application timing based on the temperature distribution deviation, thereby forming a closed-loop temperature control.

8. The railway passenger car circuit board detection system as described in claim 1, characterized in that, The comparison and verification module includes: The frequency sweep measurement unit is used to control the vector network analyzer to perform frequency sweep measurement on the repair area within a preset frequency band after the current heating sequence is completed; The result conversion unit is used to obtain the S-parameter measurement results of the repair area at multiple discrete frequency points, convert the S-parameter measurement results into complex impedance values, and generate the impedance characteristics of the repair area. Impedance characteristic error determination unit, used to extract the reference impedance characteristics and allowable error range of the same type of intact board from the pre-built performance degradation model; The impedance deviation calculation unit is used to compare the impedance characteristics of the repair area with the reference impedance characteristics point by point, and calculate the impedance deviation value at each frequency point in combination with the allowable error range. The deviation fusion unit is used to weight and fuse the impedance deviation values ​​at all frequency points to generate a comprehensive repair quality index.

9. The railway passenger car circuit board detection system as described in claim 8, characterized in that, The parameter optimization module includes: The impedance difference identification unit is used to determine that the verification has failed when the comprehensive repair quality score is lower than the verification threshold, analyze the difference between the impedance characteristics of the repaired area and the reference impedance characteristics, and identify the frequency bands and components of impedance mismatch. The optimization parameter back-reasoning unit is used to map the frequency bands and components of the impedance characteristic mismatch to the corresponding circuit physical parameter defects. Based on the type and degree of the physical parameter defects, it back-reasons the direction and amplitude of the high-frequency alternating magnetic field parameters that need to be optimized, thus forming optimized high-frequency alternating magnetic field parameters. The repair quality assessment unit is used to re-execute the self-healing repair scheme using the optimized high-frequency alternating magnetic field parameters. After each re-execution of the repair, the impedance characteristics of the repaired area are re-acquired and a new comprehensive repair quality index is calculated until the new comprehensive repair quality index reaches or exceeds the quality threshold.

10. A method for testing circuit boards in railway passenger cars, characterized in that, The railway passenger car circuit board detection method, performed by any one of claims 1 to 9, comprises: Through magnetic resonance coupling, the electrical energy required for testing is transmitted to the circuit board under test in a non-contact manner, and the testing signal is transmitted synchronously and the real-time feedback data of the circuit board under test is received. Dynamic features are extracted based on the real-time feedback data, and circuit fault detection is performed based on the extracted dynamic features to generate fault detection results. When the fault detection result detects an open circuit fault, the transmission frequency is adjusted to generate a high-frequency alternating magnetic field, triggering the circuit board under test to perform self-heating repair analysis and generate a self-healing repair scheme. The circuit board under test is coated with a self-healing material that can be locally heated to repair the open circuit. The heating parameters of the self-healing material are dynamically adjusted according to the location of the open circuit and the degree of the fault. After the circuit is repaired by heating based on the self-healing repair scheme, the impedance characteristics of the repaired area are obtained by multi-band impedance scanning and compared with the pre-built performance attenuation model for verification. If the verification fails, the high-frequency alternating magnetic field parameters will be automatically optimized based on the difference in impedance characteristics for iterative repair until the verification passes.

Images