Detection method of high-speed connector and related device

Abstract

The invention discloses a detection method of a high-speed connector and a related device, and the method comprises the steps: carrying out the plugging motion control of the high-speed connector, injecting a pulse sequence carrying a displacement mark into a differential signal pair in the process, and obtaining the initial reflection data containing a plugging displacement-time index; space-time boundary information of a gap coupling section, an initial touch deformation section and a full-insertion steady-state section is analyzed and determined in a segmented mode through reflection characteristics. And generating a mutual excitation pulse pair in the initial touch deformation section, and executing differential sampling processing by using a thermal-electric coupling effect to obtain a thermally induced group delay sequence and a geometrically induced impedance differential sequence. A dynamic zero-reflection anchor point is generated based on double-sequence fusion, and real-time reference plane migration processing is realized to obtain a calibration impedance curve. And repeating the processes of insertion and extraction, generating an elastic difference curve through displacement-time coaxial difference processing of a bidirectional calibration impedance curve, and positioning an irreversible abnormal section of the impedance. According to the method, the plugging transient impedance change can be continuously monitored, and the early failure risk of the high-speed connector can be evaluated in time.

Description

Technical Field

[0001] The present invention relates to the technical field of online electrical parameter measurement and signal integrity evaluation of high-speed differential connectors, and in particular to a detection method and related devices for high-speed connectors. Background Art

[0002] High-speed connectors are core components for high-frequency, broadband signal transmission in modern electronic systems and are widely used in data centers, 5G communication base stations, high-performance computing, and other fields. To meet the modularization and maintainability requirements of modern electronic systems, most high-speed connectors are designed to be pluggable, using a mechanical locking mechanism to establish a stable electrical connection during insertion and removal.

[0003] Currently, the quality inspection of high-speed connectors mainly adopts static testing methods, that is, using a vector network analyzer to measure the S-parameter characteristics of the connector when it is fully plugged in. However, this static testing method ignores one problem: in hot-swappable applications, the connector must maintain the continuity and integrity of signal transmission during the plugging and unplugging process. Specifically, as the pin is gradually inserted into the socket, the geometry of the contact interface continuously changes within milliseconds, causing the dynamic evolution of the transmission line impedance. At the same time, the shielded cavity and stepped terminal structure inside the connector will produce complex electromagnetic mode conversion, forming time-varying multipath reflection and resonance effects during the plugging and unplugging process. These transient electromagnetic phenomena can cause the impedance to fluctuate violently on the nanosecond to microsecond time scale. The existing test equipment is based on the measurement principle of point-by-point frequency sweep or time domain averaging. Its time resolution is simply unable to capture such rapid dynamic changes, making it impossible to evaluate the true performance of the connector in actual plugging and unplugging applications. Summary of the Invention

[0004] The main purpose of the present invention is to solve the technical problem that the existing high-speed connector detection method cannot capture nanosecond-microsecond impedance fluctuations in real time and accurately locate irreversible mismatch defects during the hot plugging process.

[0005] A first aspect of the present invention provides a method for detecting a high-speed connector, the method comprising: S1. Perform non-constant speed plugging and unplugging motion control on the high-speed connector. In the initial stage of plugging and unplugging, a pulse sequence carrying a displacement marker is continuously injected into the differential signal pair. The pulses and their echoes are collected to obtain initial reflection data containing the plugging and unplugging displacement-time index. S2. Performing segmented reflection feature analysis on the initial reflection data, determining the boundaries of the gap coupling segment, the initial contact deformation segment, and the full insertion steady-state segment based on changes in scattered energy distribution and echo drift rate, and outputting spatiotemporal boundary information; S3. In the initial deformation section, generating mutually excited pulse pairs according to the spatiotemporal boundary information, performing differential sampling processing on the group delay and reflection coefficient of the mutually excited pulse pairs, and obtaining a thermally induced group delay sequence and a geometrically induced impedance differential sequence, respectively; S4. Generate a dynamic zero reflection anchor point based on the fusion of the thermally induced group delay sequence and the geometrically induced impedance difference sequence, and use the dynamic zero reflection anchor point to perform real-time reference plane migration processing on the corresponding reflection coefficient to obtain a current direction calibration impedance curve; S5. Execute steps S1 to S4 for the insertion process and the removal process respectively to obtain an insertion direction calibration impedance curve and a removal direction calibration impedance curve, perform displacement-time coaxial differential processing on the insertion direction calibration impedance curve and the removal direction calibration impedance curve to generate an elastic differential curve, and determine the position and amplitude of the irreversible impedance abnormal section based on the elastic differential curve.

[0006] Preferably, the non-constant speed plugging and unplugging motion control of the high-speed connector is performed, and a pulse sequence carrying a displacement marker is continuously injected into the differential signal pair at the initial stage of plugging and unplugging, and the pulse and its echo are collected to obtain initial reflection data containing the plugging and unplugging displacement-time index, including: Perform three-phase speed curve synthesis processing on the plug-in drive, and generate a plug-in path scheduling table with two speed inflection points based on the connector terminal strain model and the allowable acceleration threshold; Performing interpolation and synchronization processing on the real-time displacement count values according to the plug-in and pull-out path scheduling table to obtain an equally spaced, high-resolution virtual displacement grid; injecting a displacement mark pulse sequence carrying orthogonal displacement coding into the differential signal pair according to the virtual displacement grid to obtain a displacement mark pulse sequence; performing displacement-synchronized recursive gated sampling processing on the displacement-marked pulse sequence, intercepting echoes according to the round-trip delay predicted by the virtual displacement grid, and obtaining a displacement segmented echo matrix; An index fusion process is performed based on the displacement segmented echo matrix and the virtual displacement grid to obtain initial reflection data containing plug-in displacement-time indexes.

[0007] Preferably, performing displacement-synchronized recursive gated sampling processing on the displacement-marked pulse sequence, intercepting echoes according to the round-trip delay predicted by the virtual displacement grid, and obtaining a displacement segmented echo matrix includes: Calculating an equivalent round-trip delay estimate for each grid point in the virtual displacement grid, and generating a gating delay list based on the equivalent round-trip delay estimate; Setting the sampling window start and end times of the displacement marker pulse sequence point by point according to the gated delay list, and performing recursive self-adjustment processing on the sampling window width so that the sampling window width gradually converges according to the dynamic characteristics of the insertion and removal depth; Double-threshold gating is performed on the echo data intercepted within the sampling window, the main echo and the first-level multipath echo are retained, and the displacement segmented echo matrix is output in grid order.

[0008] Preferably, the performing of segmented reflection feature analysis on the initial reflection data, determining the boundaries of the gap coupling segment, the initial contact deformation segment, and the full insertion steady-state segment based on the changes in scattered energy distribution and echo drift rate, and outputting spatiotemporal boundary information includes: Performing energy integration processing on the initial reflection data to obtain a scattered energy envelope expanded along the plug-in displacement; Performing a threshold comparison process on the energy gradient according to the scattered energy envelope to obtain an energy transition boundary candidate set; Performing echo time delay extraction processing on the initial reflection data to obtain an echo drift trajectory expanded along the plug-in displacement, performing inflection point detection processing on the first-order derivative of the time delay according to the echo drift trajectory to obtain a drift inflection point boundary candidate set; Intersection fusion processing is performed on the energy transition boundary candidate set and the drift inflection point boundary candidate set to obtain the gap coupling segment boundary, the initial contact deformation segment boundary, and the full insertion steady-state segment boundary arranged in chronological order, and the boundaries marked by the displacement-time coordinates are output as spatiotemporal boundary information.

[0009] Preferably, in the initial deformation section, mutually excited pulse pairs are generated according to the spatiotemporal boundary information, and differential sampling processing is performed on the group delay and reflection coefficient of the mutually excited pulse pairs to obtain a thermally induced group delay sequence and a geometrically induced impedance differential sequence, respectively, including: In the initial contact deformation stage, frequency domain slope layering processing is performed on the spatiotemporal boundary information, a displacement window sequence arranged along the plugging and unplugging displacement direction is obtained according to the change in the slope of the reflection energy spectrum, and the thermal diffusion association time is calculated for each displacement window; Injecting a mutual excitation pulse pair into the differential signal pair according to the shift window sequence, setting a variable delay matching the duration associated with its thermal diffusion for each shift window, and applying an amplitude inversion mask to the subsequent pulses to obtain a set of mutual excitation pulse pairs with adaptive duration and interleaved polarity; Performing group delay differential sampling processing on the mutually excited pulse pair set, calculating the group delay difference between the front pulse and the rear pulse after variable delay correction, and obtaining a thermally induced group delay sequence; Polarity return-to-zero differential sampling processing is performed on the mutual excitation pulse pair set, and the reflection coefficient difference between the front pulse and the rear pulse under the amplitude inversion mask is calculated to obtain a geometric impedance differential sequence.

[0010] Preferably, the step of injecting a mutual excitation pulse pair into the differential signal pair according to the shift window sequence, setting a variable delay matching the duration associated with the thermal diffusion of each shift window, and applying an amplitude inversion mask to the subsequent pulses to obtain a set of mutual excitation pulse pairs with adaptive duration and interleaved polarity includes: Calculating the thermal diffusion time constant for each displacement window in the displacement window sequence to generate a thermal diffusion time vector; According to the heat diffusion time vector, a variable delay is configured for the mutual excitation pulse pair so that the time interval between the front pulse and the rear pulse is equal to a preset ratio of the corresponding heat diffusion time constant, and an amplitude inversion mask is applied to the rear pulse to obtain a set of polarity-interleaved mutual excitation pulse pairs; A delay parameter recursive updating process is performed on the polarity-interleaved mutual excitation pulse pair set, the updated delay parameter is written into a delay lookup table vector, and a time-length-adapted and polarity-interleaved mutual excitation pulse pair set is output.

[0011] Preferably, the step of generating a dynamic zero reflection anchor point based on the fusion of the thermally induced group delay sequence and the geometrically induced impedance difference sequence, and performing real-time reference plane migration processing on the corresponding reflection coefficient using the dynamic zero reflection anchor point to obtain a current direction calibration impedance curve includes: Performing displacement synchronous interpolation processing on the thermally induced group delay sequence and the geometrically induced impedance difference sequence, correcting the sampling time deviation of the two sequences based on cross-correlation analysis, and obtaining synchronously aligned delay vectors and impedance vectors; Calculating first-order gradients for the time delay vector and the impedance vector respectively, generating a temperature gradient weight vector and a geometric gradient weight vector according to the absolute values of the gradients, performing normalization processing on the two weight vectors and multiplying them point by point according to the displacement to obtain a gradient weighting factor sequence; Performing sliding window minimization processing on the impedance vector according to the gradient weighting factor sequence, searching for the minimum weighted absolute reflection value at each bit of the sliding window, and obtaining a dynamic zero reflection anchor point sequence; The reference plane migration process is recursively performed on the original reflection coefficient sequence according to the dynamic zero reflection anchor point sequence, and phase unwrapping correction is synchronously performed at each migration step to obtain a current direction calibration impedance curve.

[0012] Preferably, the steps S1 to S4 are performed on the insertion process and the removal process respectively to obtain an insertion direction calibration impedance curve and a removal direction calibration impedance curve, and displacement-time coaxial difference processing is performed on the insertion direction calibration impedance curve and the removal direction calibration impedance curve to generate an elastic difference curve, and the position and amplitude of the irreversible impedance abnormal section are determined based on the elastic difference curve, including: During the insertion process, steps S1 to S4 are executed to obtain an insertion direction calibration impedance curve; during the subsequent removal process, steps S1 to S4 are executed to obtain a removal direction calibration impedance curve, and the two curves are anchor-point aligned according to the displacement index to obtain an aligned calibration curve pair; performing a displacement-by-displacement difference process on the aligned calibration curve pair to obtain an elastic difference curve; Performing sliding window variance calculation processing on the elastic difference curve, and generating an adaptive elastic threshold sequence according to the sliding window variance; performing an above-threshold truncation process on the elastic difference curve according to the adaptive elastic threshold sequence to obtain an irreversible difference subsequence; A segment integration process is performed on the irreversible difference subsequence to obtain an impedance damage area factor vector, and the position and amplitude of the impedance irreversible abnormal segment are determined according to the impedance damage area factor vector.

[0013] A second aspect of the present invention provides a high-speed connector detection device, comprising: a memory and at least one processor, wherein the memory stores instructions, and the memory and the at least one processor are interconnected via a line; the at least one processor calls the instructions in the memory to cause the high-speed connector detection device to execute the steps of the above-mentioned high-speed connector detection method.

[0014] A third aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the steps of the above-mentioned high-speed connector detection method.

[0015] When high-speed connectors are plugged in and out, the impedance fluctuates dramatically on the nanosecond to microsecond scale. Most traditional detection methods have difficulty capturing this process completely due to the stationary reference plane and sparse sampling. This solution transforms transient behavior into a resolvable data trajectory through continuous mechanical and electromagnetic dual excitation and dynamic calibration. The plug-in drive uses a variable speed curve to create a predictable distribution of the terminal strain rate and Joule heating rate throughout the stroke; the differential channel simultaneously injects a pulse train with a displacement marker, and each pulse is uniquely located to a specific displacement in the time domain. This coupling design allows the reflection data to naturally carry motion coordinates, solving the problem of synchronization between sampling and geometric state under high-speed motion.

[0016] The collected reflection sequence is first divided into a gap coupling section, an initial contact deformation section, and a fully plugged steady-state section based on scattered energy and time delay drift. This segmentation not only eliminates the interference of strong cavity echoes on subsequent processing, but also focuses the observation on the initial contact deformation zone, which is most likely to expose defects. A pair of mutually excited pulses is further injected into this section: the first pulse instantaneously heats the contact interface, and the second pulse follows before heat diffusion is complete, quantifying the change in dielectric constant through group delay perturbation; the difference in reflection coefficient records the impedance jump caused by geometric deformation. The two sequences respectively map thermal effects and mechanical deformations, and then a dynamic zero-reflection anchor point is generated through gradient weighted fusion. The reference plane continuously migrates with the anchor point, allowing the de-embedding matrix to remain effective in real time throughout the plug-in window, fundamentally eliminating the cumulative deviation of phase and amplitude caused by thermal drift.

[0017] After the full stroke, the device is immediately unplugged in the opposite direction and the same sampling and calibration process is repeated. The two calibration impedance curves for insertion and removal have identical displacement coordinates, but represent opposite loading paths. When the two are coaxially differentiated, pure elastic rebound cancels each other out, while any plastic deflection or coating delamination leaves an irreversible residual. By performing sliding window variance and segment integration on this residual curve, both defect depth and damage magnitude can be determined, ultimately enabling complete monitoring of the transient impedance continuity of high-speed connectors during insertion and removal, and precise location of irreversible faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying any creative work.

[0019] Figure 1 Schematic diagram of an embodiment of a high-speed connector detection method according to an embodiment of the present invention; Figure 2 FIG. 1 is a schematic diagram of an embodiment of a detection device for a high-speed connector according to an embodiment of the present invention.

[0020] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0022] It should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the various components under a certain specific posture. If the specific posture changes, the directional indications will also change accordingly.

[0023] In addition, the descriptions of "first", "second", etc. in the present invention are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, "and / or" in the full text includes three solutions. Taking A and / or B as an example, it includes technical solution A, technical solution B, and technical solution that satisfies both A and B. In addition, the technical solutions between the various embodiments can be combined with each other, and must be based on the ability of ordinary technicians in this field to implement. When the combination of technical solutions is mutually contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0024] An embodiment of the present application provides a method for detecting a high-speed connector. Figure 1 A flow chart of a high-speed connector detection method provided in one embodiment of the present application. In this embodiment, the method includes: See also Figure 1 S1: Perform non-constant speed plugging and unplugging motion control on the high-speed connector. In the initial stage of plugging and unplugging, a pulse sequence carrying displacement markers is continuously injected into the differential signal pair, and the pulses and their echoes are collected to obtain initial reflection data containing plugging and unplugging displacement-time index; In one embodiment of the present invention, the non-constant speed plugging and unplugging motion control of the high-speed connector is performed, and a pulse sequence carrying a displacement marker is continuously injected into the differential signal pair at the initial stage of plugging and unplugging, and the pulses and their echoes are collected to obtain initial reflection data containing the plugging and unplugging displacement-time index, including: Perform three-phase speed curve synthesis processing on the plug-in drive, and generate a plug-in path scheduling table with two speed inflection points based on the connector terminal strain model and the allowable acceleration threshold; Performing interpolation and synchronization processing on the real-time displacement count values according to the plug-in and pull-out path scheduling table to obtain an equally spaced, high-resolution virtual displacement grid; injecting a displacement mark pulse sequence carrying orthogonal displacement coding into the differential signal pair according to the virtual displacement grid to obtain a displacement mark pulse sequence; performing displacement-synchronized recursive gated sampling processing on the displacement-marked pulse sequence, intercepting echoes according to the round-trip delay predicted by the virtual displacement grid, and obtaining a displacement segmented echo matrix; An index fusion process is performed based on the displacement segmented echo matrix and the virtual displacement grid to obtain initial reflection data containing plug-in displacement-time indexes.

[0025] The following is a detailed description of the steps involved in the above embodiment: Before testing begins, the motor controller issues a three-phase velocity curve control command to the linear actuator. This velocity curve divides the insertion and removal process into three phases: the first phase uses high acceleration and constant speed, primarily to overcome initial insertion resistance; the second phase switches to lower acceleration and speed at approximately 20% of the full insertion depth to reduce mechanical shock; and the third phase increases to medium speed and maintains zero acceleration in the final 5% of travel to ensure smooth mating between the terminal and the spring. The locations of the two velocity inflection points were determined through finite element simulation, corresponding to the maximum gap between adjacent terminals and the physical location where the spring enters the elastic deformation zone. The permissible acceleration threshold is calculated based on the yield strength and safety factor of the connector terminal material to ensure that plastic deformation is not generated during the insertion and removal process. The control software records the actuator's displacement, velocity, and acceleration data with sub-millisecond time resolution. The relationships between the three velocity curves and the corresponding displacements are compiled into a data table for insertion and removal path scheduling, which is stored in the system buffer for subsequent access. This schedule establishes a precise time-displacement-velocity mapping, ensuring that subsequent electromagnetic excitation is strictly synchronized with the actual mechanical state, providing a stable physical basis for accurate impedance measurement.

[0026] The optical encoder outputs real-time displacement counts with micron-level resolution. The control software performs cubic spline interpolation on these discrete displacement data. The interpolation algorithm reads adjacent displacement nodes from the schedule table, calculates the cubic polynomial coefficients between the nodes, and generates equally spaced sampling points. The interpolation spacing is chosen based on the typical dimensions of connector terminals, ensuring that each terminal unit is covered by multiple sampling segments. This avoids redundant data generated by oversampling and prevents the omission of narrow impedance mismatch regions. The resulting virtual displacement grid contains a large number of sampling points, each associated with an absolute timestamp, instantaneous velocity, and acceleration parameters. Compared to the sampling interval of the original schedule table, the temporal and spatial resolution of the virtual displacement grid is significantly improved. For example, for a typical connector insertion and removal stroke, the virtual displacement grid generates hundreds of precise sampling points, each of which records the complete mechanical state information at that moment. This high-density displacement grid provides a unified spatial reference for the precise positioning of subsequent pulse trains.

[0027] An arbitrary waveform generator (AWG) emits a pair of orthogonally encoded displacement marker pulses at each node of the virtual displacement grid. Pulse parameters are selected based on the typical operating frequency band of high-speed connectors to ensure excellent signal transmission characteristics. Orthogonal displacement encoding utilizes the phase mapping technique of the Walsh sequence. The two pulses use orthogonal carrier phases to ensure zero cross-correlation, thus preventing interference between echoes at different locations. Specifically, the encoding implementation uses phase modulation to carry the high-order displacement information, while the second pulse carries the low-order information. The pulses are differentially injected into the connector under test through appropriate attenuators to maintain differential signal balance. The pulse emission interval is strictly synchronized with the spatial spacing of the virtual displacement grid, meaning a pair of marker pulses is emitted each time the plugging and unplugging displacement advances a preset distance. This encoding method ensures that each reflected echo carries the precise displacement information at the time of emission, providing a reliable data foundation for the subsequent accurate correlation between displacement and time.

[0028] The oscilloscope front-end opens the sampling gate window appropriately in advance based on the theoretical round-trip delay of each node in the virtual displacement grid. The theoretical round-trip delay is calculated as twice the signal propagation distance divided by the effective propagation velocity in the connector medium. The sampling gate utilizes a recursive window adjustment strategy. The initial window width is set to a large value to ensure complete signal capture. The window width is then gradually reduced according to a preset narrowing factor, reaching a minimum value during the deep insertion phase to improve resolution. This recursive narrowing strategy effectively suppresses high-density multipath noise during the late insertion and removal phases while ensuring the integrity of the main echo envelope. Dual-threshold gating is performed within each sampling window. First, a lower threshold is set to eliminate ambient noise, determined based on the system's inherent noise level. Then, an upper threshold is set to filter out direct coupling interference from the transmitted signal, retaining only the main echo and major multipath reflections. The system stacks the sampling results row by row according to the sequence of the virtual displacement grid, forming a displacement-segmented echo matrix. The number of rows corresponds to the number of grid points, and the number of columns corresponds to the order of the retained echo. Each matrix element contains complete time-domain amplitude and phase information for the corresponding echo.

[0029] Data fusion processing precisely associates the displacement segmented echo matrix with the virtual displacement grid. For each grid index, the system extracts the corresponding displacement coordinates, timestamp, and primary echo and multipath data from the echo matrix. This information is combined into a data structure containing complete parameters: displacement value, timestamp, primary echo amplitude, primary echo phase, multipath echo amplitude, and multipath echo phase. The data is stored in a structured format, fully recording the information at each sampling point. For example, a sampling point contains the specific displacement value, corresponding timestamp, primary echo amplitude and phase information, and multipath echo amplitude and phase parameters. This structured data organization eliminates potential clock skew between the test system and the actuator, as all time stamps are uniformly aligned using the virtual displacement grid. The resulting initial reflection dataset contains a large number of complete data records, each of which establishes a precise correspondence between plug-in displacement, propagation time, and electrical parameters. This dataset provides an accurate displacement benchmark for subsequent segmented reflection feature analysis, completely resolving the technical challenge of accurately mapping signal parameters to physical locations in hot-swap applications.

[0030] In one embodiment of the present invention, performing displacement-synchronized recursive gated sampling processing on the displacement-marked pulse sequence, intercepting echoes according to the round-trip delay predicted by the virtual displacement grid, and obtaining a displacement segmented echo matrix includes: Calculating an equivalent round-trip delay estimate for each grid point in the virtual displacement grid, and generating a gating delay list based on the equivalent round-trip delay estimate; Setting the sampling window start and end times of the displacement marker pulse sequence point by point according to the gated delay list, and performing recursive self-adjustment processing on the sampling window width so that the sampling window width gradually converges according to the dynamic characteristics of the insertion and removal depth; Double-threshold gating is performed on the echo data intercepted within the sampling window, the main echo and the first-level multipath echo are retained, and the displacement segmented echo matrix is output in grid order.

[0031] The following is a detailed description of the steps involved in the above embodiment: The system calculates an estimated equivalent round-trip delay for each grid point in the virtual displacement grid. The equivalent round-trip delay is the total time it takes for an electromagnetic signal to propagate from the transmitter to the plug-in / plug-out contact and then reflect back to the receiver. The calculation first extracts the plug-in / plug-out depth value corresponding to each grid point in the virtual displacement grid. This depth value is then multiplied by two to determine the signal's round-trip propagation distance. The propagation time calculation takes into account the signal's propagation speed in the various media of the connector, including air, insulation, and metal conductor segments. In air segments, the signal propagates at a speed close to the speed of light; in insulation segments, the propagation speed is converted based on the relative permittivity of the medium; and in metal conductor segments, the signal propagates as a surface wave. The system accumulates the propagation times for each segment to obtain a total estimated equivalent round-trip delay. The gated delay list is an array structure that stores delay data for each grid point. It contains four fields: grid index, plug-in / plug-out depth, equivalent round-trip delay, and expected echo arrival time. For example, when the insertion depth is 2.5 mm, assuming a total round-trip propagation distance of 5 mm and an average propagation speed of 70% of the speed of light in the mixed medium, the calculated equivalent round-trip delay is approximately 79 picoseconds. This pre-calculated delay estimation mechanism eliminates the latency of real-time calculations, ensuring that sampling gating is precisely synchronized with the actual signal propagation process, and avoiding echo signal loss caused by inaccurate delay prediction.

[0032] The oscilloscope sampling system sets the sampling window start and end times for each pulse in the displacement marker pulse train based on the gated delay table. The sampling window start time is set to the pulse emission time plus the corresponding equivalent round-trip delay estimate, minus a preset lead time that accounts for system response delay and signal propagation uncertainty. The sampling window end time is determined by adding the window width to the start time. Recursive self-tuning refers to an algorithm that dynamically adjusts the sampling window width based on changes in signal characteristics during the plugging and unplugging process. During the initial plugging and unplugging process, the connector's internal structure is relatively simple, and multipath effects are minimal. Therefore, the sampling window can be set wider to ensure complete capture of the echo signal. However, as the plugging and unplugging depth increases, the metal structure and dielectric interfaces within the connector increase, and multipath reflections become more complex. At this point, the sampling window width needs to be narrowed to improve temporal resolution and reduce noise interference. The recursive self-tuning algorithm uses a piecewise linear adjustment strategy based on the plugging and unplugging depth, dividing the plugging and unplugging travel into several segments, each with a different window width adjustment factor. For example, during the first 30% of the insertion / unplugging stroke, the window width remains at its initial value; during the middle 40%, the window width gradually decreases according to a preset ratio; and during the final 30%, the window width converges to its minimum value. This gradually convergent window adjustment strategy adapts to the complex changes in the signal propagation environment during the insertion / unplugging process, maximizing measurement accuracy while ensuring signal integrity.

[0033] The system performs dual-threshold gating on echo data captured within the sampling window to filter out noise and retain valid signal components. This dual-threshold gating process consists of two criteria: an upper threshold and a lower threshold. The lower threshold is used to filter out low-amplitude ambient noise and system noise floor. This threshold is determined based on the system's noise level in a no-signal state and is set to the noise power mean plus a multiple of the standard deviation. The upper threshold is used to eliminate excessively strong direct-coupled signals and transmission leakage to prevent receiver saturation. This threshold is calculated based on the transmit power and the expected maximum reflection coefficient. After dual-threshold gating, the system retains signal components with amplitudes between the two thresholds. The primary echo is the direct echo returned after a single reflection and has the shortest propagation path and the highest signal strength. The primary multipath echo is a secondary echo formed after two or three reflections. It has a longer propagation path but still has sufficient signal strength for analysis. The system identifies echoes of different orders based on their time delay differences: the arrival time of the primary echo is closest to the theoretically calculated value, while the arrival time of the primary multipath echo is slightly delayed. The displacement segmented echo matrix is a two-dimensional data structure arranged in the order of a virtual displacement grid. The rows of the matrix correspond to different plug-in and unplug positions, and the columns correspond to echo data of different orders. Each matrix element contains the time domain waveform data of the corresponding echo, including the amplitude sequence and the phase sequence. For example, for a plug-in and unplug process containing 500 grid points, if the main echo and the first-level multipath echo are retained at each position, the output displacement segmented echo matrix dimension is 500 rows and 2 columns. This structured data organization method establishes a direct mapping relationship between plug-in and unplug positions and echo characteristics, providing a complete data foundation for subsequent spatiotemporal boundary analysis. At the same time, by retaining multipath information, it enhances sensitivity to changes in the internal structure of the connector.

[0034] Please continue reading Figure 1 S2, performing segmented reflection feature analysis on the initial reflection data, determining the boundaries of the gap coupling segment, the initial contact deformation segment, and the full insertion steady-state segment based on the changes in the scattered energy distribution and the echo drift rate, and outputting the spatiotemporal boundary information; In one embodiment of the present invention, the segmented analysis of reflection characteristics of the initial reflection data is performed, and the boundaries of the gap coupling segment, the initial contact deformation segment, and the full insertion steady-state segment are determined based on the changes in the scattered energy distribution and the echo drift rate, and the spatiotemporal boundary information is output, including: Performing energy integration processing on the initial reflection data to obtain a scattered energy envelope expanded along the plug-in displacement; Performing a threshold comparison process on the energy gradient according to the scattered energy envelope to obtain an energy transition boundary candidate set; Performing echo time delay extraction processing on the initial reflection data to obtain an echo drift trajectory expanded along the plug-in displacement, performing inflection point detection processing on the first-order derivative of the time delay according to the echo drift trajectory to obtain a drift inflection point boundary candidate set; Intersection fusion processing is performed on the energy transition boundary candidate set and the drift inflection point boundary candidate set to obtain the gap coupling segment boundary, the initial contact deformation segment boundary, and the full insertion steady-state segment boundary arranged in chronological order, and the boundaries marked by the displacement-time coordinates are output as spatiotemporal boundary information.

[0035] The following is a detailed description of the steps involved in the above embodiment: The system performs energy integration on the reflected signal at each insertion / unplug displacement point in the initial reflection data. Energy integration involves calculating the total energy of the time-domain reflection waveform corresponding to each displacement point. Specifically, the amplitude of the reflected signal is squared and numerically integrated over the entire time window. The initial reflection data contains all reflection signals from the insertion / unplug start position to the fully plugged position, with each displacement point corresponding to a complete time-domain waveform. During processing, the digital signal processor reads the reflected waveform data for each displacement point one by one, calculates the squared amplitude values for all sample points in the waveform, and then sums these values to obtain the scattered energy value for that displacement point. The scattered energy envelope is a curve formed by connecting the scattered energy values corresponding to all insertion / unplug displacement points in the order of displacement. The horizontal axis of this envelope curve represents the insertion / unplug displacement, and the vertical axis represents the scattered energy at the corresponding displacement. For example, if the insertion / unplug process progresses from 0 mm to 5 mm and contains 500 displacement sampling points, the scattered energy envelope will contain 500 energy value points. During the initial mating phase, the connector's interior is primarily a hollow structure, resulting in low scattering energy. As the pins begin to contact the terminals, reflection from the metal interface increases significantly, causing a sharp rise in scattering energy. During the fully mated phase, the structure stabilizes, and the scattering energy gradually levels off. This energy integration process compresses complex time-domain waveform information into a single energy metric, highlighting the differences in electromagnetic scattering intensity during the different mating phases and providing clear characteristic parameters for subsequent segment recognition.

[0036] Energy gradient execution processing identifies energy transition locations by calculating the rate of energy change between adjacent displacement points in the scattered energy envelope. The energy gradient refers to the rate of change of scattered energy along the plug-in displacement direction. It is calculated by subtracting the energy value of the previous displacement point from the energy value of the current displacement point, and then dividing it by the displacement interval. The system traverses the entire scattered energy envelope, calculates the energy gradient value corresponding to each displacement point, and forms an energy gradient sequence. Threshold comparison processing refers to comparing the calculated energy gradient value with the preset threshold to identify displacement points where the absolute value of the gradient exceeds the threshold. The threshold setting is based on the statistical characteristics of the scattered energy envelope, and the standard deviation of the energy gradient sequence multiplied by a preset multiple is used as the judgment criterion. When the absolute value of the energy gradient exceeds the threshold, it indicates that there is a significant change in the electromagnetic environment at the displacement point, and it is marked as an energy transition candidate point. The energy transition boundary candidate set is an array of all marked transition candidate points arranged in displacement order. Each element contains the displacement coordinates, the corresponding timestamp, and the energy gradient value. For example, during a complete plugging and unplugging process, the energy transition boundary candidate set contains approximately 10-20 candidate points, corresponding to the locations where key physical events occur, such as when the pin contacts the terminal, enters the elastic zone, and fully locks. This energy gradient-based transition detection method accurately captures the sudden changes in the electromagnetic environment during the plugging and unplugging process, avoiding errors caused by subjective judgment.

[0037] The echo delay extraction process extracts the arrival time of the main echo corresponding to each insertion / unplug displacement point from the initial reflection data. Echo delay is the time interval from pulse emission to the arrival of the main echo peak, reflecting the round-trip propagation distance of the signal. During the extraction process, the system performs peak detection on the time-domain reflection waveform at each displacement point, identifies the echo signal with the largest amplitude, and records its arrival time. The difference between this time and the pulse emission time is the echo delay. The echo drift trajectory is a curve formed by connecting the echo delay values corresponding to all insertion / unplug displacement points in displacement order, with the horizontal axis representing insertion / unplug displacement and the vertical axis representing echo delay. Ideally, echo delay should increase linearly with insertion / unplug depth, but in practice, delay drift occurs due to changes in contact state and geometric structure adjustments during insertion / unplugging. The first-order derivative of the delay is the slope of the echo drift trajectory, indicating the rate of change of echo delay with insertion / unplug displacement. The inflection point detection process identifies turning points in the trajectory by calculating the sign change of the first-order derivative of the delay. In specific implementation, the system calculates the first-order derivative value of each displacement point in the echo drift trajectory and then detects whether the signs of adjacent derivative values change. When the derivative sign changes from positive to negative or from negative to positive, the corresponding displacement point is marked as an inflection point. The drift inflection point boundary candidate set contains all detected inflection point locations and their corresponding delay derivative values. For example, when the pin first contacts the terminal, the growth rate of the echo delay will suddenly change, forming a clear inflection point in the echo drift trajectory. This inflection point detection method based on delay derivatives can accurately identify the transition moment of the geometric fit state during the insertion and removal process, complementing the shortcomings of the energy analysis method.

[0038] The intersection fusion process compares and analyzes the energy transition boundary candidate set with the drift inflection point boundary candidate set, selecting displacement points that meet both criteria as the final segment boundary. The intersection fusion criterion is that the distance between the displacement points in the two candidate sets is less than a preset tolerance range, which is determined by the spacing of the virtual displacement grid and the system's measurement error. The system iterates through each candidate point in the energy transition boundary candidate set and searches for the closest candidate point in the drift inflection point boundary candidate set. If the distance between the two points is less than the tolerance range, the displacement point is confirmed as a valid segment boundary. The gap coupling segment refers to the insertion and removal phase before the pin contacts the terminal. During this phase, the signal propagates primarily through the cavity, characterized by low scattering energy and a gradual delay change. The initial contact deformation segment refers to the phase when the pin contacts the terminal and undergoes elastic deformation. It is characterized by a sharp increase in scattering energy and a sudden change in the delay rate. The fully plugged steady-state segment refers to the stable phase after the connector is fully plugged in. It is characterized by a gradual scattering energy and a near-constant delay rate. Segment boundaries are determined by analyzing the physical significance of candidate points and their temporal sequence during the insertion and removal process: the first intersection point corresponds to the boundary between the gap coupling segment and the initial contact deformation segment, and the second intersection point corresponds to the boundary between the initial contact deformation segment and the fully inserted steady-state segment. Spatiotemporal boundary information is structured data containing the displacement coordinates, timestamp, and segment identifier of each segment boundary, allowing subsequent processing steps to accurately locate each insertion and removal phase. This dual-verification boundary determination method combines the advantages of energy and time delay characteristics, significantly improving the accuracy and reliability of segment segmentation and ensuring that subsequent mutual excitation pulse processing can accurately focus on the most critical initial contact deformation segment.

[0039] Please continue reading Figure 1 S3, in the initial deformation section, generating a mutual excitation pulse pair according to the space-time boundary information, performing differential sampling processing on the group delay and reflection coefficient of the mutual excitation pulse pair, and obtaining a thermally induced group delay sequence and a geometrically induced impedance differential sequence respectively; In one embodiment of the present invention, in the initial contact deformation section, generating a mutual excitation pulse pair according to the spatiotemporal boundary information, performing differential sampling processing on the group delay and reflection coefficient of the mutual excitation pulse pair to obtain a thermally induced group delay sequence and a geometrically induced impedance differential sequence, respectively, includes: In the initial contact deformation stage, frequency domain slope layering processing is performed on the spatiotemporal boundary information, a displacement window sequence arranged along the plugging and unplugging displacement direction is obtained according to the change in the slope of the reflection energy spectrum, and the thermal diffusion association time is calculated for each displacement window; Injecting a mutual excitation pulse pair into the differential signal pair according to the shift window sequence, setting a variable delay matching the duration associated with its thermal diffusion for each shift window, and applying an amplitude inversion mask to the subsequent pulses to obtain a set of mutual excitation pulse pairs with adaptive duration and interleaved polarity; Performing group delay differential sampling processing on the mutually excited pulse pair set, calculating the group delay difference between the front pulse and the rear pulse after variable delay correction, and obtaining a thermally induced group delay sequence; Polarity return-to-zero differential sampling processing is performed on the mutual excitation pulse pair set, and the reflection coefficient difference between the front pulse and the rear pulse under the amplitude inversion mask is calculated to obtain a geometric impedance differential sequence.

[0040] The following is a detailed description of the steps involved in the above embodiment: The system performs frequency-domain slope stratification on the spatiotemporal boundary information within the initial contact deformation segment. Frequency-domain slope stratification involves converting the reflected signal within the initial contact deformation segment to the frequency domain, analyzing the changing characteristics of its spectral slope, and then stratifying it based on the slope differences. Specifically, a fast Fourier transform (FFT) is performed on the time-domain reflection signal at each displacement point within the initial contact deformation segment to obtain the corresponding frequency-domain reflection spectrum. The reflection spectrum is the amplitude spectrum of the frequency-domain reflection signal, with the horizontal axis representing frequency and the vertical axis representing the amplitude of the reflection signal at that frequency. The system calculates the slope of each frequency-domain reflection spectrum within a preset frequency range, which reflects the frequency response characteristics of the reflection signal. The slope variation of the reflection spectrum refers to the difference in the slope of the reflection spectrum between adjacent displacement points. When the slope variation exceeds a preset threshold, it indicates a significant change in the electromagnetic environment at that location. Based on the distribution characteristics of the slope variation, the system divides the initial contact deformation segment into several subregions, each corresponding to a displacement window. The displacement window sequence is an array of these subregions arranged in the order of insertion and removal displacement. Each displacement window contains three parameters: a starting displacement, an ending displacement, and a window length. The thermal diffusion correlation time refers to the characteristic time required for the Joule heat generated by the contact resistance to diffuse to the surrounding medium within the displacement window. During the calculation process, the system estimates the thermal diffusion time constant based on the contact area, material thermal conductivity, and heat capacity parameters within the displacement window. For example, in the 2 mm plugging and unplugging stroke of the initial contact deformation section, if the change in the slope of the reflected energy spectrum detects 8 significant change points, 7 displacement windows are formed, each with a length of approximately 0.3 mm, and the corresponding thermal diffusion correlation time is in the range of tens to hundreds of nanoseconds. This hierarchical processing method based on the frequency domain slope can identify subtle changes in the electromagnetic propagation characteristics during the plugging and unplugging process, providing suitable spatial resolution for subsequent refined measurements.

[0041] The arbitrary waveform generator injects a mutual excitation pulse pair into the differential signal pair according to a sequence of displacement windows. A mutual excitation pulse pair is two pulses emitted consecutively within a short time interval. The front pulse is used to stimulate the connector to generate transient thermal effects, while the back pulse is used to detect the impact of this thermal effect on electromagnetic propagation characteristics. The variable delay is the time interval between the front and back pulses, which is set based on the thermal diffusion correlation duration of the corresponding displacement window. In implementation, the system reads the thermal diffusion correlation duration value for each displacement window and uses it as the delay time for the mutual excitation pulse pair within that window. When the thermal diffusion correlation duration is short, the pulse interval is set to a smaller value to capture rapid thermal responses; when the thermal diffusion correlation duration is longer, the pulse interval is increased to allow for sufficient thermal diffusion. Amplitude inversion masking is a signal processing operation performed on the back pulse. This operation inverts the amplitude of the back pulse, multiplying it by negative one, giving it the opposite polarity to the front pulse. The purpose of this inversion operation is to eliminate common background signals in the subsequent differential processing and highlight subtle changes caused by thermal effects. The set of time-adaptive and polarity-interleaved mutual excitation pulse pairs is a sequence of mutual excitation pulse pairs corresponding to all displacement windows, arranged in displacement order. Each pulse pair has a delay time and phase inversion characteristics that match its displacement window. For example, in an initial deformation segment containing seven displacement windows, the system will generate seven pairs of mutual excitation pulses. The first pair has a delay time of 50 nanoseconds, the second pair has a delay time of 80 nanoseconds, and so on. All subsequent pulses are amplitude-inverted. This adaptive pulse parameter setting mechanism ensures that the thermal-electric coupling effect within each displacement window is optimally stimulated and detected.

[0042] The system performs group delay differential sampling on a collection of mutually excited pulse pairs. Group delay refers to the time required for a signal envelope to propagate, reflecting the actual propagation velocity of the signal in the transmission medium. Differential sampling involves measuring the group delay of the preceding and following pulses separately and then calculating the difference between them. Specifically, the system measures the group delay of the preceding and following pulses of each mutually excited pulse pair. Group delay is determined by analyzing the slope of the reflected signal's phase variation with frequency. Variable delay correction deducts the manually set interpulse delay when calculating the group delay difference, retaining only delay variations caused by physical effects. The calculation is performed by subtracting the preceding pulse group delay from the following pulse group delay, then subtracting the set variable delay. The thermally induced group delay sequence is an array of group delay differences corresponding to all displacement windows, arranged in displacement order. This sequence reflects the variation in signal propagation delay caused by contact thermal effects during insertion and removal. For example, when Joule heating generated by contact resistance raises the temperature of the surrounding medium, the dielectric constant of the medium changes, causing a change in signal propagation velocity, resulting in a measurable, small difference in group delay. During a typical initial contact deformation period, the thermally induced group delay sequence may contain seven values, each representing the thermally induced delay variation within the corresponding displacement window, ranging from a few picoseconds to tens of picoseconds. This group delay differential measurement method can detect extremely small thermally induced dielectric constant changes, providing a high-precision electrical means for indirect measurement of temperature fields.

[0043] The system performs polarity-zero differential sampling on the set of mutually excited pulse pairs. Polarity-zero differential sampling involves performing a differential operation on the front pulse and the back pulse, which has been processed with an amplitude inversion mask. This operation eliminates common mode signals and highlights differential signals. In practice, the system measures the reflection coefficients of the front pulse and back pulse separately. Because the back pulse has been amplitude-inverted, summing the reflection coefficients of the two pulses automatically eliminates the common background reflection component, retaining only the difference caused by physical state changes. The reflection coefficient is the complex ratio of the incident signal to the reflected signal and consists of two components: amplitude and phase. The amplitude inversion mask ensures that the reflection coefficient of the back pulse has the opposite sign to that of the front pulse. When the two are added together, the fixed geometric reflection components cancel each other out, while the dynamic variation components caused by thermal effects are retained and amplified. The geometrically induced impedance differential sequence is an array of reflection coefficient differences corresponding to all displacement windows, arranged in displacement order. This sequence reflects the impedance characteristic differences caused by changes in the geometric fit during insertion and removal. For example, when the contact area between the pin and the terminal changes slightly, the local impedance changes accordingly. This change is amplified in the geometric impedance differential sequence after polarity-zeroed differential processing. In an initial contact deformation phase containing seven displacement windows, the geometric impedance differential sequence will contain seven complex values, each representing the impedance change characteristic within the corresponding displacement window. The real and imaginary parts of the value reflect the resistive and reactive components of the impedance change, respectively. This polarity-zeroed differential measurement technique significantly improves the detection sensitivity of small impedance changes and can identify subtle geometric adjustments that are undetectable by conventional measurement methods.

[0044] In one embodiment of the present invention, injecting a mutual excitation pulse pair into the differential signal pair according to the shift window sequence, setting a variable delay matching the duration associated with the heat diffusion of each shift window, and applying an amplitude inversion mask to the subsequent pulse to obtain a set of mutual excitation pulse pairs with adaptive duration and interleaved polarity, includes: Calculating the thermal diffusion time constant for each displacement window in the displacement window sequence to generate a thermal diffusion time vector; According to the heat diffusion time vector, a variable delay is configured for the mutual excitation pulse pair so that the time interval between the front pulse and the rear pulse is equal to a preset ratio of the corresponding heat diffusion time constant, and an amplitude inversion mask is applied to the rear pulse to obtain a set of polarity-interleaved mutual excitation pulse pairs; A delay parameter recursive updating process is performed on the polarity-interleaved mutual excitation pulse pair set, the updated delay parameter is written into a delay lookup table vector, and a time-length-adapted and polarity-interleaved mutual excitation pulse pair set is output.

[0045] The following is a detailed description of the steps involved in the above embodiment: The system calculates the corresponding thermal diffusion time constant for each displacement window in the displacement window sequence. The thermal diffusion time constant is the characteristic time required for the Joule heat generated at the contact interface within that displacement window to diffuse into the surrounding medium and reach a steady-state distribution. During the calculation process, the digital signal processor first reads the geometric parameters of each displacement window, including the window length, expected contact area, and contact pressure distribution. The contact area is estimated based on the geometric overlap between the pin and the terminal within the displacement window, while the contact pressure is determined based on the elastic properties of the spring and the insertion and extraction depth. The system then calculates the equivalent heat capacity and thermal resistance within that displacement window. The heat capacity depends on the specific heat capacity and mass of the metal and dielectric materials in the contact area, while the thermal resistance is determined by the thermal conductivity and geometric dimensions of the materials. The thermal diffusion time constant is calculated by multiplying the heat capacity and thermal resistance and reflects the time scale for establishing thermal equilibrium within that displacement window. The thermal diffusion time vector is a one-dimensional array of the thermal diffusion time constants corresponding to all displacement windows arranged in displacement order. For example, in the initial contact deformation phase, which consists of seven displacement windows, the heat diffusion time vector will contain seven time constant values. The time constant in the first window is approximately 20 nanoseconds, and as the insertion depth increases and the contact area expands, the time constant in the last window can reach 150 nanoseconds. This physical parameter-based time constant calculation method ensures the accuracy of subsequent pulse delay settings, allowing the thermal-electric coupling effects within each displacement window to be detected at the optimal time.

[0046] The arbitrary waveform generator configures variable delay parameters for each mutual pulse pair within each displacement window based on the heat diffusion time vector. Variable delay configuration sets the time interval between the front and back pulses based on the value in the heat diffusion time vector. The system reads the time constant corresponding to each displacement window in the heat diffusion time vector and multiplies it by a preset scaling factor to obtain the actual pulse delay. The preset scaling factor is typically set between 0.5 and 0.8, based on the dynamics of the heat diffusion process: a too small scaling factor results in detection before the thermal effect has fully developed, while a too large scaling factor misses the optimal thermal response detection window. For example, if the heat diffusion time constant for a displacement window is 100 nanoseconds, a preset scaling factor of 0.6 would set the delay time for the mutual pulse pair within that window to 60 nanoseconds. Amplitude inversion masking performs an amplitude inversion operation on the back pulse of each mutual pulse pair, multiplying its amplitude by -1 to give it the opposite polarity to the front pulse. This inversion process is implemented by sign-shifting the pulse waveform data using a digital signal processor. A polarity-interleaved interexcitation pulse pair set consists of all interexcitation pulse pairs that have undergone variable delay configuration and amplitude inversion, arranged in a displacement window sequence. Within this set, each pulse pair has a delay time that matches the thermal characteristics of its displacement window, and the trailing pulse undergoes polarity inversion. This adaptive delay configuration and polarity manipulation mechanism optimizes the stimulation and discrimination of thermally induced changes within each displacement window, significantly improving the detection accuracy of weak thermo-electric coupling signals.

[0047] The system recursively updates the delay parameters for a set of polarity-interleaved interexcitation pulse pairs. This recursive delay parameter update is an iterative optimization process that dynamically adjusts the current pulse delay parameters based on the previous measurement results. In practice, the system first performs a complete round of interexcitation pulse measurements using the initial delay parameters. It then analyzes the signal-to-noise ratio and feature clarity of the resulting thermally induced group delay sequence and geometrically induced impedance differential sequence. If the signal quality in a displacement window is unsatisfactory, the system adjusts the corresponding delay parameters. If the thermally induced signal is too weak, indicating that the delay is too short, the system increases the delay; if the signal is saturated or distorted, indicating that the delay is too long, the system decreases the delay. The update algorithm uses a proportional adjustment strategy, with adjustments ranging from 5% to 15% of the original delay value to ensure smooth parameter changes. The delay lookup table vector is a data structure that stores the updated delay parameters. It contains three fields: the original delay value for each displacement window, the adjustment amount, and the final delay value. The system writes the optimized delay parameters into the delay lookup table vector for use in the next measurement round. The set of mutually excited pulse pairs, which undergo time adaptation and polarity interleaving, is the final pulse sequence after recursive updating. Each pulse pair in this set has optimized delay parameters and polarity configuration. For example, in an initial deformation segment containing seven displacement windows, after three recursive updates, the delay lookup vector may show that the delay of the third window has been adjusted from the initial 80 nanoseconds to 92 nanoseconds, and the delay of the fifth window has been adjusted from 120 nanoseconds to 108 nanoseconds. This adaptive parameter optimization mechanism compensates for the deviation between theoretical calculations and actual physical processes, ensuring optimal measurement accuracy and reliability for thermal-electric coupling detection within each displacement window.

[0048] Please continue reading Figure 1 S4, generating a dynamic zero reflection anchor point based on the fusion of the thermally induced group delay sequence and the geometrically induced impedance difference sequence, and performing real-time reference plane migration processing on the corresponding reflection coefficient using the dynamic zero reflection anchor point to obtain a current direction calibration impedance curve; In one embodiment of the present invention, generating a dynamic zero reflection anchor point based on the fusion of the thermally induced group delay sequence and the geometrically induced impedance difference sequence, and performing real-time reference plane migration processing on the corresponding reflection coefficient using the dynamic zero reflection anchor point to obtain a current direction calibration impedance curve includes: Performing displacement synchronous interpolation processing on the thermally induced group delay sequence and the geometrically induced impedance difference sequence, correcting the sampling time deviation of the two sequences based on cross-correlation analysis, and obtaining synchronously aligned delay vectors and impedance vectors; Calculating first-order gradients for the time delay vector and the impedance vector respectively, generating a temperature gradient weight vector and a geometric gradient weight vector according to the absolute values of the gradients, performing normalization processing on the two weight vectors and multiplying them point by point according to the displacement to obtain a gradient weighting factor sequence; Performing sliding window minimization processing on the impedance vector according to the gradient weighting factor sequence, searching for the minimum weighted absolute reflection value at each bit of the sliding window, and obtaining a dynamic zero reflection anchor point sequence; The reference plane migration process is recursively performed on the original reflection coefficient sequence according to the dynamic zero reflection anchor point sequence, and phase unwrapping correction is synchronously performed at each migration step to obtain a current direction calibration impedance curve.

[0049] The following is a detailed description of the steps involved in the above embodiment: The system performs shift-synchronized interpolation on the thermally induced group delay sequence and the geometrically induced impedance difference sequence to eliminate sampling inconsistencies between the two sequences. Shift-synchronized interpolation involves resampling the two sequences in the same shift coordinate system, ensuring that each shift point contains both thermally induced delay data and geometrically induced impedance data. In specific implementation, the digital signal processor first compares the shift sampling point distributions of the two sequences, identifying the sequence with the higher sampling density as the benchmark. It then performs cubic spline interpolation on the other sequence, calculating the corresponding interpolated data at each shift point in the benchmark sequence. Cross-correlation analysis detects time delay or shift offset between the two sequences by calculating the cross-correlation function. The system performs a cross-correlation operation on the interpolated sequences, searching for the offset that maximizes the cross-correlation function. This offset is the sampling time offset between the two sequences. Sampling time offset correction is achieved by applying the detected offset to one of the sequences. Specifically, the sequence with the larger offset is shifted along the shift axis accordingly. The synchronized time delay vector represents the thermally induced group delay data after shift synchronization and offset correction, while the impedance vector represents the geometrically induced impedance difference data, which has undergone the same processing. For example, when a thermally induced group delay sequence contains seven data points and a geometrically induced impedance differential sequence contains nine data points, the system interpolates the former to nine data points. Cross-correlation analysis then reveals a 0.05mm displacement error, which is corrected to yield two completely synchronized vectors. This precise spatiotemporal synchronization eliminates systematic deviations between measurement channels and ensures the accuracy of subsequent fusion analysis.

[0050] The system calculates the first-order gradient of the time delay vector and the impedance vector respectively to quantify the rate of change of the two physical quantities along the insertion and removal displacement direction. The first-order gradient calculation is achieved by dividing the difference between adjacent data points by the displacement interval, reflecting the sensitivity of the corresponding physical quantity to change at that position. For the time delay vector, the system calculates the rate of change of the time delay value at each displacement point relative to the previous displacement point to obtain a temperature gradient vector, each element of which represents the intensity of the thermal effect at the corresponding displacement. For the impedance vector, the system performs the same gradient calculation to obtain a geometric gradient vector, which reflects the degree of change of the geometric fit state at each displacement point. The temperature gradient weight vector is obtained by taking the absolute value of the temperature gradient vector and normalizing it, and the geometric gradient weight vector is obtained from the geometric gradient vector in the same way. Normalization refers to dividing all elements of each vector by the maximum value of the vector so that all weight values are between 0 and 1. Point-by-point multiplication by displacement refers to multiplying the elements of the temperature gradient weight vector and the geometric gradient weight vector at the same displacement index one by one. The gradient weighting factor sequence is the result of the point-by-point multiplication of two weight vectors. Each element comprehensively reflects the combined strength of the thermal and geometric effects at the corresponding displacement point. For example, at a displacement point, if the temperature gradient weight is 0.8 and the geometric gradient weight is 0.6, the gradient weighting factor at that point is 0.48. This dual gradient weighting mechanism can highlight displacement regions with the most significant thermal-electric coupling effects, providing reliable weight guidance for subsequent anchor point search.

[0051] The system performs a sliding window minimization on the impedance vector based on a sequence of gradient weighting factors to identify the optimal reference plane location. Sliding window minimization involves searching for the minimum weighted reflectance value within a sliding window of a preset length. In practice, the system sets the sliding window length to three to five displacement points. This length is chosen based on the spatial correlation of the connector geometry to ensure that the window encompasses local features while not overly smoothing critical details. Within each sliding window, the system calculates the absolute value of each element of the impedance vector multiplied by the corresponding gradient weighting factor to produce a sequence of weighted absolute reflectance values. The displacement point corresponding to the minimum weighted absolute reflectance value is designated as a zero-reflection candidate within that window. The system slides the window along the entire mating and unmating stroke, performing the same minimization search at each window position, ultimately generating a series of zero-reflection candidate points. These candidate points are further screened and verified to form the final set of anchor points, a dynamic zero-reflection anchor point sequence. Each anchor point represents an ideal reference plane location where the combined effects of thermal and geometric effects drive the reflection coefficient close to zero. For example, in a plugging and unplugging process involving 200 displacement points, sliding window minimization may identify 15-20 zero-reflection candidate points, retaining 8-12 valid dynamic zero-reflection anchor points after verification. This adaptive anchor point identification method fully utilizes the changes in physical properties during the plugging and unplugging process and avoids the misalignment problem of the fixed reference plane method under conditions of thermally induced geometric deformation.

[0052] The system recursively performs reference plane migration on the original reflection coefficient sequence based on a sequence of dynamic zero-reflection anchor points. Reference plane migration dynamically adjusts the measurement reference plane based on the anchor point positions to eliminate the effects of thermally induced geometric deformation on impedance measurements. The original reflection coefficient sequence represents uncorrected reflection coefficient measurements during the plugging and unplugging process, containing systematic deviations due to factors such as thermal expansion and contact deformation. Recursive processing involves applying dynamic zero-reflection anchor points one by one to perform reference plane correction in the chronological order of plugging and unplugging displacement. In practice, the system begins with the first anchor point, sets it as the current reference plane zero point, and performs phase and amplitude correction on all reflection coefficient data between this and the next anchor point. This correction is achieved through complex arithmetic, subtracting the reflection coefficient reference value at the anchor point from the original reflection coefficient. Phase unwrapping correction is an algorithm that handles phase jumps in the complex reflection coefficients. When the phase difference exceeds 180 degrees, it automatically performs a ±360-degree phase adjustment to ensure phase continuity. The system performs the same migration and correction operations for each anchor point interval, gradually eliminating accumulated errors throughout the plugging and unplugging process. The current-direction calibrated impedance curve is the final impedance measurement result after a complete reference plane migration and phase unwrapping correction. This curve truly reflects the connector's intrinsic impedance characteristics in the current plugging and unplugging direction, eliminating the effects of thermally induced geometric deformation. For example, during a plugging and unplugging process involving eight dynamic zero-reflection anchors, the system will perform eight recursive migration operations, each correcting 10-30 raw reflection coefficient data points within the corresponding displacement range, ultimately outputting a complete impedance curve containing 200 calibration data points. This dynamic reference plane migration technology significantly improves the accuracy and stability of impedance measurements under high-temperature hot-plugging conditions.

[0053] Please continue reading Figure 1 , S5. Execute steps S1 to S4 for the insertion process and the removal process respectively to obtain the insertion direction calibration impedance curve and the removal direction calibration impedance curve, perform displacement-time coaxial differential processing on the insertion direction calibration impedance curve and the removal direction calibration impedance curve to generate an elastic differential curve, and determine the position and amplitude of the impedance irreversible abnormal section based on the elastic differential curve.

[0054] In one embodiment of the present invention, performing steps S1 to S4 on the insertion process and the removal process respectively to obtain an insertion direction calibration impedance curve and a removal direction calibration impedance curve, performing displacement-time coaxial difference processing on the insertion direction calibration impedance curve and the removal direction calibration impedance curve to generate an elastic difference curve, and determining the position and amplitude of the irreversible impedance abnormal section based on the elastic difference curve includes: During the insertion process, steps S1 to S4 are executed to obtain an insertion direction calibration impedance curve; during the subsequent removal process, steps S1 to S4 are executed to obtain a removal direction calibration impedance curve, and the two curves are anchor-point aligned according to the displacement index to obtain an aligned calibration curve pair; performing a displacement-by-displacement difference process on the aligned calibration curve pair to obtain an elastic difference curve; Performing sliding window variance calculation processing on the elastic difference curve, and generating an adaptive elastic threshold sequence according to the sliding window variance; performing an above-threshold truncation process on the elastic difference curve according to the adaptive elastic threshold sequence to obtain an irreversible difference subsequence; A segment integration process is performed on the irreversible difference subsequence to obtain an impedance damage area factor vector, and the position and amplitude of the impedance irreversible abnormal segment are determined according to the impedance damage area factor vector.

[0055] The following is a detailed description of the steps involved in the above embodiment: After completing steps S1 to S4 of the insertion process, the system immediately performs the reverse removal operation and repeats the same inspection process. The insertion process includes non-constant speed control of the high-speed connector, injection of a displacement marker pulse sequence, segmented reflection signature analysis, and generation of dynamic zero-reflection anchor points. This ultimately produces a calibration impedance curve in the insertion direction. This curve reflects the actual impedance changes of the connector during the insertion process. The removal process uses the opposite motion direction of the insertion process, but the inspection method and processing steps are identical, including three-phase velocity curve control, generation of mutually excited pulse pairs, and differential sampling of thermally induced group delay and geometrically induced impedance. This ultimately produces a calibration impedance curve in the removal direction. The displacement index refers to the insertion and removal displacement coordinates corresponding to each data point in the two curves. Because the displacement ranges of the insertion and removal processes are the same but in opposite directions, a coordinate transformation is required to establish a corresponding relationship. Anchor point alignment identifies key displacement points with similar impedance characteristics in the two curves to achieve precise alignment. The system first searches for local extreme impedance values in the two curves as characteristic anchor points. It then calculates the displacement deviations between these anchor points and eliminates these deviations through interpolation and coordinate adjustment. An aligned calibration curve pair is a pair of curves that have been anchored to ensure that data points at the same displacement index correspond to the same physical location. For example, if the insertion direction curve has an impedance peak at 2.3 mm displacement, the aligned withdrawal direction curve should also correspond to the fully engaged pin depth minus 2.3 mm at the same displacement index. This bidirectional measurement and precise alignment eliminates systematic errors found in unidirectional measurements, providing reliable baseline data for subsequent elasticity analysis.

[0056] The digital signal processor performs displacement-by-displacement differencing on the aligned calibration curve pairs. Displacement-by-displacement differencing involves subtracting the corresponding value from the calibration impedance curve for the insertion direction at each displacement index to determine the impedance differential value at that displacement index. Since both measurements undergo the same calibration process, the differential result should ideally be close to zero, indicating good reversibility of the connector's electrical characteristics at that displacement index. The elastic differential curve is a curve formed by connecting the differential values of all displacement points in order of displacement. The magnitude of this curve directly reflects the connector's elastic recovery at different displacement indexes. When the elastic differential curve value is close to zero, the connector deformation at that location is fully reversible elastic deformation. Significant deviations from zero indicate plastic deformation, localized damage, or other irreversible structural changes. For example, in a 5mm insertion / extraction travel test, if the calibration impedance at 3.2mm of displacement in the insertion direction is 52.3 ohms and the calibration impedance at the corresponding position in the extraction direction is 52.8 ohms, the elastic differential value at that point is -0.5 ohms. The entire elastic differential curve contains the differential information of all displacement points, which can accurately identify the health status and potential damage locations of the connector's internal structure.

[0057] The system performs a sliding window variance calculation on the elastic difference curve to generate an adaptive decision threshold. This calculation calculates the statistical variance of the elastic difference values within a sliding window of a preset length, quantifying the degree of data dispersion within that area. The sliding window length is set between 5 and 10 displacement points, based on the spatial scale of the connector's geometric features, to ensure that the window captures local variations while not oversmoothing important details. The system slides the window along the entire elastic difference curve, calculating the variance of the data points within each window position. The variance calculation includes standard statistical steps such as averaging, calculating the sum of squared deviations, and dividing by the number of samples. The sliding window variance sequence is an array of variance values corresponding to all window positions, arranged in sequence. The adaptive elastic threshold sequence is generated by statistically analyzing the sliding window variance sequence. Specifically, the mean and standard deviation of the variance sequence are calculated, and then the mean plus a multiple of the standard deviation is used as the elastic threshold for that position. The threshold multiplier is determined based on the required detection accuracy. A larger coefficient results in a more lenient threshold, while a smaller coefficient results in more sensitive detection. For example, if the variance within a sliding window is 0.15, if the coefficient is set to 2.5 and the standard deviation is 0.08, then the elastic threshold corresponding to this window is 0.15 + 2.5 × 0.08 = 0.35. This adaptive threshold generation method dynamically adjusts the judgment criteria based on the local statistical characteristics of the elastic difference curve, significantly improving the accuracy and reliability of anomaly detection.

[0058] The system performs suprathreshold truncation on the elasticity difference curve based on an adaptive elasticity threshold sequence. This process marks data points in the elasticity difference curve whose absolute values exceed the corresponding elasticity threshold as outliers, while data points whose absolute values fall below the threshold are set to zero. In practice, the system compares the absolute value of each data point in the elasticity difference curve with the elasticity threshold at the corresponding location. If the absolute value exceeds the threshold, the original differential value is retained; otherwise, the value of the point is set to zero. The irreversible differential subsequence consists of all non-zero data points remaining after the suprathreshold truncation process. These points correspond to displacement locations where irreversible changes have occurred within the connector. Segmental integration calculates the integrated area for consecutive non-zero data segments in the irreversible differential subsequence. The system first identifies consecutive non-zero data segments, each corresponding to a potential damage region. The system then numerically integrates the differential values within each data segment. The integrated area reflects the overall severity of the irreversible changes within that region. The impedance damage area factor vector is an array of the integrated area values corresponding to all consecutive data segments, arranged in displacement order. Each element contains three parameters: the starting and ending locations of the damage region, and the integrated area. The location of the irreversible impedance anomaly segment is determined by the position parameter in the impedance damage area factor vector, while the amplitude is quantitatively expressed by the size of the integrated area. For example, in a single test, if there are continuous above-threshold differential values within the displacement range of 1.8 to 2.1 mm, and the integrated area of this segment is 0.85, then this area is marked as an irreversible anomaly segment with a damage position of 1.8-2.1 mm and a damage amplitude of 0.85. This damage localization method based on statistical thresholds and integrated areas can accurately identify microscopic damage within the connector, including various irreversible structural changes such as plating peeling, contact surface scratches, and localized overheating deformation.

[0059] The following describes in detail the detection device of the high-speed connector according to the embodiment of the present invention from the perspective of hardware processing.

[0060] Figure 2FIG2 is a schematic diagram of the structure of a high-speed connector detection device provided by an embodiment of the present invention. The high-speed connector detection device 200 may vary significantly depending on configuration or performance. It may include one or more processors 210 (e.g., one or more processors), memory 220, and one or more storage media 230 (e.g., one or more mass storage devices) storing application programs 233 or data 232. The memory 220 and storage medium 230 may be either transient or persistent storage. The program stored in the storage medium 230 may include one or more modules (not shown), each of which may include a series of instructions for operating on the high-speed connector detection device 200. Furthermore, the processor 210 may be configured to communicate with the storage medium 230, executing the series of instructions stored in the storage medium 230 on the high-speed connector detection device 200 to implement the steps of the high-speed connector detection method described above.

[0061] The high-speed connector detection device 200 may further include one or more power supplies 240, one or more wired or wireless network interfaces 250, one or more input / output interfaces 260, and / or one or more operating systems 231, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. It will be understood by those skilled in the art that Figure 2 The structure of the high-speed connector detection device shown does not constitute a limitation on the high-speed connector detection device provided by the present invention, and may include more or fewer components than shown, or combine certain components, or arrange components differently.

[0062] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions, which, when executed on a computer, cause the computer to execute the steps of the high-speed connector detection method.

[0063] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0064] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0065] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. All equivalent structural transformations made by using the contents of the present invention description and drawings under the inventive concept of the present invention, or direct / indirect application in other related technical fields are included in the patent protection scope of the present invention.

Claims

1. A method for detecting a high-speed connector, characterized in that: include: S1. Perform non-constant speed plugging and unplugging motion control on the high-speed connector. In the initial stage of plugging and unplugging, a pulse sequence carrying a displacement marker is continuously injected into the differential signal pair. The pulses and their echoes are collected to obtain initial reflection data containing the plugging and unplugging displacement-time index. S2. Performing segmented reflection feature analysis on the initial reflection data, determining the boundaries of the gap coupling segment, the initial contact deformation segment, and the full insertion steady-state segment based on changes in scattered energy distribution and echo drift rate, and outputting spatiotemporal boundary information; S3. In the initial deformation section, generating mutually excited pulse pairs according to the spatiotemporal boundary information, performing differential sampling processing on the group delay and reflection coefficient of the mutually excited pulse pairs, and obtaining a thermally induced group delay sequence and a geometrically induced impedance differential sequence, respectively; S4. Generate a dynamic zero reflection anchor point based on the fusion of the thermally induced group delay sequence and the geometrically induced impedance difference sequence, and use the dynamic zero reflection anchor point to perform real-time reference plane migration processing on the corresponding reflection coefficient to obtain a current direction calibration impedance curve; S5. Execute steps S1 to S4 for the insertion process and the removal process respectively to obtain an insertion direction calibration impedance curve and a removal direction calibration impedance curve, perform displacement-time coaxial differential processing on the insertion direction calibration impedance curve and the removal direction calibration impedance curve to generate an elastic differential curve, and determine the position and amplitude of the irreversible impedance abnormal section based on the elastic differential curve.

2. The high-speed connector detection method according to claim 1, characterized in that: The method performs non-constant speed plugging and unplugging motion control on the high-speed connector, continuously injects a pulse sequence carrying a displacement marker into the differential signal pair at the initial plugging and unplugging stage, collects the pulses and their echoes, and obtains initial reflection data containing a plugging and unplugging displacement-time index, including: Perform three-phase speed curve synthesis processing on the plug-in drive, and generate a plug-in path scheduling table with two speed inflection points based on the connector terminal strain model and the allowable acceleration threshold; Performing interpolation and synchronization processing on the real-time displacement count values according to the plug-in and pull-out path scheduling table to obtain an equally spaced, high-resolution virtual displacement grid; injecting a displacement mark pulse sequence carrying orthogonal displacement coding into the differential signal pair according to the virtual displacement grid to obtain a displacement mark pulse sequence; performing displacement-synchronized recursive gated sampling processing on the displacement-marked pulse sequence, intercepting echoes according to the round-trip delay predicted by the virtual displacement grid, and obtaining a displacement segmented echo matrix; An index fusion process is performed based on the displacement segmented echo matrix and the virtual displacement grid to obtain initial reflection data containing plug-in displacement-time indexes.

3. The high-speed connector detection method according to claim 2, characterized in that: The step of performing displacement-synchronized recursive gated sampling processing on the displacement-marked pulse sequence, intercepting echoes according to the round-trip delay predicted by the virtual displacement grid, and obtaining a displacement segmented echo matrix comprises: Calculating an equivalent round-trip delay estimate for each grid point in the virtual displacement grid, and generating a gating delay list based on the equivalent round-trip delay estimate; Setting the sampling window start and end times of the displacement marker pulse sequence point by point according to the gated delay list, and performing recursive self-adjustment processing on the sampling window width so that the sampling window width gradually converges according to the dynamic characteristics of the insertion and removal depth; Double-threshold gating is performed on the echo data intercepted within the sampling window, the main echo and the first-level multipath echo are retained, and the displacement segmented echo matrix is output in grid order.

4. The method for detecting a high-speed connector according to claim 1, wherein: The reflection feature segmentation analysis is performed on the initial reflection data, and the boundaries of the gap coupling segment, the initial contact deformation segment, and the full insertion steady-state segment are determined according to the changes in the scattered energy distribution and the echo drift rate, and the spatiotemporal boundary information is output, including: Performing energy integration processing on the initial reflection data to obtain a scattered energy envelope expanded along the plug-in displacement; Performing a threshold comparison process on the energy gradient according to the scattered energy envelope to obtain an energy transition boundary candidate set; Performing echo time delay extraction processing on the initial reflection data to obtain an echo drift trajectory expanded along the plug-in displacement, performing inflection point detection processing on the first-order derivative of the time delay according to the echo drift trajectory to obtain a drift inflection point boundary candidate set; Intersection fusion processing is performed on the energy transition boundary candidate set and the drift inflection point boundary candidate set to obtain the gap coupling segment boundary, the initial contact deformation segment boundary, and the full insertion steady-state segment boundary arranged in chronological order, and the boundaries marked by the displacement-time coordinates are output as spatiotemporal boundary information.

5. The high-speed connector detection method according to claim 1, characterized in that: In the initial deformation section, mutually excited pulse pairs are generated according to the spatiotemporal boundary information, and differential sampling processing is performed on the group delay and reflection coefficient of the mutually excited pulse pairs to obtain a thermally induced group delay sequence and a geometrically induced impedance differential sequence, respectively, including: In the initial contact deformation stage, frequency domain slope layering processing is performed on the spatiotemporal boundary information, a displacement window sequence arranged along the plugging and unplugging displacement direction is obtained according to the change in the slope of the reflection energy spectrum, and the thermal diffusion association time is calculated for each displacement window; Injecting a mutual excitation pulse pair into the differential signal pair according to the shift window sequence, setting a variable delay matching the duration associated with its thermal diffusion for each shift window, and applying an amplitude inversion mask to the subsequent pulses to obtain a set of mutual excitation pulse pairs with adaptive duration and interleaved polarity; Performing group delay differential sampling processing on the mutually excited pulse pair set, calculating the group delay difference between the front pulse and the rear pulse after variable delay correction, and obtaining a thermally induced group delay sequence; Polarity return-to-zero differential sampling processing is performed on the mutual excitation pulse pair set, and the reflection coefficient difference between the front pulse and the rear pulse under the amplitude inversion mask is calculated to obtain a geometric impedance differential sequence.

6. The high-speed connector detection method according to claim 5, characterized in that: The method includes injecting a mutual excitation pulse pair into the differential signal pair according to the shift window sequence, setting a variable delay matching the duration associated with the heat diffusion of each shift window, and applying an amplitude inversion mask to the subsequent pulse to obtain a set of mutual excitation pulse pairs with adaptive duration and interleaved polarity, including: Calculating the thermal diffusion time constant for each displacement window in the displacement window sequence to generate a thermal diffusion time vector; According to the heat diffusion time vector, a variable delay is configured for the mutual excitation pulse pair so that the time interval between the front pulse and the rear pulse is equal to a preset ratio of the corresponding heat diffusion time constant, and an amplitude inversion mask is applied to the rear pulse to obtain a set of polarity-interleaved mutual excitation pulse pairs; A delay parameter recursive updating process is performed on the polarity-interleaved mutual excitation pulse pair set, the updated delay parameter is written into a delay lookup table vector, and a time-length-adapted and polarity-interleaved mutual excitation pulse pair set is output.

7. The method for detecting a high-speed connector according to claim 1, wherein: The method generates a dynamic zero reflection anchor point by fusing the thermally induced group delay sequence and the geometrically induced impedance difference sequence, and performs real-time reference plane migration processing on the corresponding reflection coefficient using the dynamic zero reflection anchor point to obtain a current direction calibration impedance curve, including: Performing displacement synchronous interpolation processing on the thermally induced group delay sequence and the geometrically induced impedance difference sequence, correcting the sampling time deviation of the two sequences based on cross-correlation analysis, and obtaining synchronously aligned delay vectors and impedance vectors; Calculating first-order gradients for the time delay vector and the impedance vector respectively, generating a temperature gradient weight vector and a geometric gradient weight vector according to the absolute values of the gradients, performing normalization processing on the two weight vectors and multiplying them point by point according to the displacement to obtain a gradient weighting factor sequence; Performing sliding window minimization processing on the impedance vector according to the gradient weighting factor sequence, searching for the minimum weighted absolute reflection value at each bit of the sliding window, and obtaining a dynamic zero reflection anchor point sequence; The reference plane migration process is recursively performed on the original reflection coefficient sequence according to the dynamic zero reflection anchor point sequence, and phase unwrapping correction is synchronously performed at each migration step to obtain a current direction calibration impedance curve.

8. The method for detecting a high-speed connector according to claim 1, wherein: The method further comprises: performing steps S1 to S4 on the insertion process and the removal process respectively to obtain an insertion direction calibration impedance curve and a removal direction calibration impedance curve; performing displacement-time coaxial differential processing on the insertion direction calibration impedance curve and the removal direction calibration impedance curve to generate an elastic differential curve; and determining the position and amplitude of the impedance irreversible abnormal section based on the elastic differential curve. During the insertion process, steps S1 to S4 are executed to obtain an insertion direction calibration impedance curve; during the subsequent removal process, steps S1 to S4 are executed to obtain a removal direction calibration impedance curve, and the two curves are anchor-point aligned according to the displacement index to obtain an aligned calibration curve pair; performing a displacement-by-displacement difference process on the aligned calibration curve pair to obtain an elastic difference curve; Performing sliding window variance calculation processing on the elastic difference curve, and generating an adaptive elastic threshold sequence according to the sliding window variance; performing an above-threshold truncation process on the elastic difference curve according to the adaptive elastic threshold sequence to obtain an irreversible difference subsequence; A segment integration process is performed on the irreversible difference subsequence to obtain an impedance damage area factor vector, and the position and amplitude of the impedance irreversible abnormal segment are determined according to the impedance damage area factor vector.

9. A high-speed connector detection device, characterized in that: The high-speed connector detection device includes: a memory and at least one processor, wherein the memory stores instructions; The at least one processor calls the instructions in the memory to enable the high-speed connector detection device to perform the steps of the high-speed connector detection method according to any one of claims 1 to 8.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by a processor, the steps of the high-speed connector detection method according to any one of claims 1 to 8 are implemented.

Images