A wire harness connectivity detection system based on high frequency signal reflection

Abstract

The present application relates to the technical field of electrical performance test, and more particularly to a wire harness connectivity detection system based on high-frequency signal reflection, comprising: a pulse transceiving unit for transmitting high-frequency step signals and receiving echoes; a single-point dense sampling unit for repeatedly sampling at the theoretical echo delay window time of a crimping area to obtain a voltage amplitude sequence; a kurtosis calculation unit for determining an excess kurtosis based on the voltage amplitude sequence; a baseline management unit for retrieving an ideal reference waveform according to the information of the wire harness under test; a logarithmic residual integral unit for constructing a logarithmic ratio residual curve based on the actual echo and the ideal reference waveform, and integrating the logarithmic ratio residual curve within the theoretical echo delay window to obtain an integral area; a peak-to-average ratio calculation unit for calculating the peak-to-average ratio of the logarithmic ratio residual curve; and a decision output unit for outputting a connectivity detection conclusion based on the excess kurtosis, the integral area and the peak-to-average ratio. The present application effectively identifies wire harness micro-disconnection defects and distinguishes external electromagnetic interference.

Description

Technical Field

[0001] This invention relates to the field of electrical performance testing technology, and in particular to a wire harness connectivity detection system based on high-frequency signal reflection. Background Technology

[0002] High-voltage wiring harnesses in new energy vehicles are a crucial component of the vehicle's electrical system. They typically consist of numerous stranded copper wires connected by crimp terminals. Due to the long-term exposure to complex conditions such as vibration, high temperatures, and electrochemical corrosion, the copper wires at the crimp terminals are at risk of partial breakage. To address this, the industry commonly employs the time-domain reflectometry (TD-RESP) method to test the continuity of the wiring harness. The basic principle of TRP is to inject a fast-edge step signal into the wiring harness under test. Utilizing the characteristic of impedance discontinuities generating echo reflections, the impedance distribution within the wiring harness is inferred by acquiring and analyzing the time-domain waveform of the echo signal. This method offers advantages such as non-destructive testing, location capability, and fast response speed, and has been widely applied in the field of cable testing, resulting in a relatively mature testing instrument system and engineering application standards.

[0003] However, time-domain reflectometry (TD-RESP) has significant limitations in detecting partial strand breakage defects in high-voltage wiring harnesses for new energy vehicles. High-voltage wiring harnesses contain numerous copper wires. When only a small portion of these wires break, the remaining wires maintain normal DC conduction, resulting in minimal overall DC resistance change and normal macroscopic electrical performance. In this case, the impedance change caused by the breakage is minimal, and the corresponding change in the echo reflection coefficient is drowned out by system sampling noise. Existing TDP testing equipment cannot identify effective strand breakage characteristics from the time-domain waveform, leading to missed detections of some strand breakage defects. Furthermore, existing TDP equipment faces external electromagnetic interference in actual industrial testing environments. Interference signals superimposed on the echo waveform are easily misinterpreted as internal wiring harness defects, leading to false alarms. The lack of effective methods to distinguish the weak echo anomalies caused by partial strand breakage from background noise and external electromagnetic interference restricts the application of TDP in the refined quality inspection of high-voltage wiring harnesses.

[0004] Chinese Patent Publication No. CN120352808A discloses a continuity testing system and method for engine wiring harnesses. The system acquires the wiring harness layout diagram, deploys matrix relays to form a programmable switch matrix, performs a logic self-test, searches for the optimal continuity path based on a hierarchical partitioning model and a DFS algorithm, identifies abnormal wiring harness segments by overlapping and mapping the path maps of each partition, and finally determines the anomaly and provides guidance for physical repair through analog signal injection and waveform analysis. However, this system and method suffer from the following problems: they rely on analog signal injection and path continuity verification, but weak impedance changes (without change in DC resistance) caused by partial strand breakage cannot be effectively detected by conventional continuity testing or waveform analysis; furthermore, their anti-interference capability is insufficient, and in complex electromagnetic environments, external interference can easily cause waveform distortion at the receiving end, leading to misjudgments. Summary of the Invention

[0005] To address this, the present invention provides a wire harness connectivity detection system based on high-frequency signal reflection, which overcomes the problems in the prior art where the time-domain reflection method cannot identify the weak impedance changes caused by partial strand breakage and cannot distinguish between strand breakage defects and external electromagnetic interference.

[0006] To achieve the above objectives, the present invention provides a wire harness connectivity detection system based on high-frequency signal reflection, comprising: The pulse transceiver unit is used to transmit high-frequency step signals to the wire harness under test and receive echo signals. A single-point dense sampling unit, which is connected to the pulse transceiver unit, is used to determine the theoretical echo delay window corresponding to the crimping area based on the nominal characteristic impedance and nominal length of the wire harness under test. The echo voltage amplitude is repeatedly sampled a preset number of times at the time corresponding to the theoretical echo delay window to obtain the voltage amplitude sequence. A kurtosis calculation unit, which is connected to the single-point dense sampling unit, is used to construct an amplitude probability density histogram based on the voltage amplitude sequence and determine the over-value kurtosis according to the amplitude probability density histogram. The baseline management unit contains a database of qualified sample reflection waveforms indexed by wire harness model identifier and ambient temperature. It is used to retrieve and output the corresponding ideal reference waveform based on the model identifier of the wire harness under test and the current ambient temperature. The logarithmic residual integration unit is connected to the pulse transceiver unit, the single-point dense sampling unit, and the baseline management unit, respectively. It is used to construct a logarithmic ratio residual curve based on the actual echo waveform and the ideal reference waveform, and to perform time integration on the logarithmic ratio residual curve within the theoretical echo delay window to obtain the integration area. The peak-to-average power ratio (PAPR) calculation unit, which is connected to the logarithmic residual integration unit, is used to calculate the ratio of the maximum absolute value of the amplitude of the logarithmic ratio residual curve to the absolute value of the amplitude mean within the theoretical echo delay window, thereby obtaining the PAPR. The decision output unit is connected to the kurtosis calculation unit, the logarithmic residual integration unit, and the peak-to-average ratio (PARR) calculation unit, respectively, and is used to output a connectivity detection conclusion based on the comparison results of the excess kurtosis, the integral area, and the PARR with the corresponding preset threshold.

[0007] Furthermore, the pulse transceiver unit includes a fast-edge pulse generator and a high-speed analog-to-digital converter; The fast-edge pulse generator generates a step signal and transmits it to the wire harness under test. The high-speed analog-to-digital converter receives the echo signal and converts it into digital waveform data. The fast-edge pulse generator and the high-speed analog-to-digital converter are connected by a directional coupler to form a closed-loop signal path.

[0008] Furthermore, the single-point dense sampling unit determines the theoretical echo delay window corresponding to the crimping area based on the nominal characteristic impedance and nominal length of the tested wire harness. At the time corresponding to the theoretical echo delay window, the echo voltage amplitude is repeatedly sampled several times. Each sampling only obtains the voltage amplitude at that time, thus obtaining a voltage amplitude sequence.

[0009] Furthermore, the baseline management unit pre-stores the measured reflection waveform of the first qualified sample in the same batch as the batch baseline, and the measured reflection waveform of the tested wire harness after its connectivity is confirmed as the historical baseline.

[0010] Furthermore, based on the determination result of the existence of the historical baseline, the baseline management unit uses the historical baseline as the initial ideal reference waveform; based on the determination result that the historical baseline does not exist and the model identifier of the tested wire harness has a corresponding record in the qualified sample reflection waveform database, the baseline management unit uses the corresponding waveform retrieved from the qualified sample reflection waveform database as the initial ideal reference waveform.

[0011] Furthermore, the baseline management unit uses the batch baseline as the initial ideal reference waveform based on the judgment result that the historical baseline does not exist and the model identifier of the tested wire harness does not have a corresponding record in the qualified sample reflection waveform database.

[0012] Furthermore, the baseline management unit corrects the initial ideal reference waveform based on the difference between the current ambient temperature and the ambient temperature when the selected initial ideal reference waveform was stored, thereby obtaining the ideal reference waveform.

[0013] Further, the rise time of the high-frequency step signal emitted by the fast-edge pulse generator is less than the minimum resolvable time interval corresponding to the system background noise.

[0014] Further, the logarithmic residual integration unit takes the logarithm of the ratio of the absolute value of the amplitude of the actual echo waveform at each time point plus the anti-zero constant to the absolute value of the amplitude of the ideal reference waveform at the corresponding time point plus the anti-zero constant as the logarithmic ratio residual value at this time point, and the logarithmic ratio residual curve is continuously formed by the logarithmic ratio residual values at each time point.

[0015] Further, the decision output unit pre-stores the kurtosis preset threshold corresponding to the overvalue kurtosis, the integral preset threshold corresponding to the integral area, and the peak-to-average ratio preset threshold corresponding to the peak-to-average ratio respectively; If the overvalue kurtosis is less than or equal to the kurtosis preset threshold, output that the connectivity is qualified; If the overvalue kurtosis is greater than the kurtosis preset threshold and the integral area is greater than or equal to the negative integral preset threshold, output the determination of external electromagnetic interference; If the overvalue kurtosis is greater than the kurtosis preset threshold, the integral area is less than the negative integral preset threshold, and the peak-to-average ratio is greater than the peak-to-average ratio preset threshold, output a micro-break strand warning; If the overvalue kurtosis is greater than the kurtosis preset threshold, the integral area is less than the negative integral preset threshold, and the peak-to-average ratio is less than or equal to the peak-to-average ratio preset threshold, output the determination of uniform offset of the harness impedance.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows. By jointly judging three characteristic quantities, namely, the amplitude statistical distribution characteristics obtained by single-point dense sampling, the window integral area of the logarithmic ratio residual curve, and the peak-to-average ratio, the present invention realizes the identification of partial break strand defects at the crimp terminals of high-voltage harnesses, and solves the problem that the impedance change caused by micro-break strands in the existing time domain reflectometry method is submerged in the system background noise and cannot be detected.

[0017] The present invention focuses the repeated sampling resources on the theoretical echo delay window moment corresponding to the crimping area for single-point dense sampling, so that the non-Gaussian characteristics of the amplitude probability density distribution can be statistically manifested from a large number of samplings, rather than relying on the signal-to-noise ratio of single sampling, thereby improving the detection sensitivity of weak break strand characteristics without improving the hardware index.

[0018] By introducing the logarithmic ratio residual curve, the present invention non-linearly amplifies the slight amplitude difference between the actual echo waveform and the ideal reference waveform in the logarithmic domain into a resolvable offset amount, and solves the problem that the break strand characteristics are submerged by the amplifier drift and low-frequency interference when directly taking the difference in the linear domain.

[0019] This invention uses the integral area and peak-to-average power ratio in combination to distinguish and determine three sources of abnormality: micro-strand breakage, uniform offset of wire harness impedance, and external electromagnetic interference. This makes the detection conclusion output by the system have a clear qualitative significance of the fault, rather than just a binary conclusion of whether it is qualified or not.

[0020] This invention establishes a three-level baseline selection mechanism in the baseline management unit, prioritizing historical baselines, qualified sample reflection waveform databases, and batch baselines, and applies temperature compensation correction to the selected baselines. This ensures that the ideal reference waveform can accurately reflect the expected echo amplitude of the tested wire harness under qualified conditions in different testing scenarios and at different ambient temperatures, thus guaranteeing the validity of the baseline comparison results. Attached Figure Description

[0021] Figure 1 This is a connection block diagram of the wire harness connectivity detection system based on high-frequency signal reflection according to the present invention; Figure 2 This is a logic block diagram of the single-point dense sampling and over-value kurtosis calculation of the harness connectivity detection system based on high-frequency signal reflection of the present invention. Figure 3 This is a logic block diagram of the baseline management and peak-to-average ratio calculation of the harness connectivity detection system based on high-frequency signal reflection of the present invention. Figure 4 This is a logic block diagram of the decision output unit of the wire harness connectivity detection system based on high-frequency signal reflection of the present invention. Detailed Implementation

[0022] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0023] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0024] It should be noted that in the description of this invention, the terms "upper," "lower," "left," "right," "inner," and "outer," etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is merely for ease of description and does not indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0025] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0026] Please see Figure 1 The diagram shown is a link block diagram of the wire harness connectivity detection system based on high-frequency signal reflection according to the present invention. The present invention provides a wire harness connectivity detection system based on high-frequency signal reflection, comprising: The pulse transceiver unit is used to transmit high-frequency step signals to the wire harness under test and receive echo signals. Specifically, the pulse transceiver unit includes a fast-edge pulse generator and a high-speed analog-to-digital converter; The fast-edge pulse generator generates a step signal and transmits it to the wire harness under test. The high-speed analog-to-digital converter receives the echo signal and converts it into digital waveform data. The fast-edge pulse generator and the high-speed analog-to-digital converter are connected by a directional coupler to form a closed-loop signal path.

[0027] Specifically, the rise time of the high-frequency step signal emitted by the fast-edge pulse generator is less than the minimum resolvable time interval corresponding to the system's background noise.

[0028] In one specific embodiment, the pulse transceiver unit comprises a fast-edge pulse generator, a high-speed analog-to-digital converter (ADC), and a directional coupler. The fast-edge pulse generator preferably employs a commercially available fast-edge pulse source based on a step recovery diode (SRD) or gallium arsenide field-effect transistor (GaAs FET) drive circuit. The rise time of the high-frequency step signal generated by this generator is less than the minimum resolvable time interval corresponding to the system's noise floor; preferably, the rise time is no greater than 500 ps. The high-speed ADC preferably employs a commercially available high-speed ADC chip with a sampling rate of no less than 1 GSa / s, which converts the received analog echo signal into digital waveform data and transmits it to the subsequent processing unit. The signal output terminal of the fast-edge pulse generator and the signal input terminal of the high-speed ADC are both connected to the corresponding ports of the directional coupler. The directional coupler isolates the transmit path from the receive path, forming a closed-loop signal path: the transmitted signal is injected into the tested wiring harness through the transmit port of the directional coupler, and the echo signal from the tested wiring harness returns through the same connection port and is then directionally exported to the high-speed ADC through the receive port of the directional coupler. The transmitted signal and the echo signal do not interfere with each other on the physical path.

[0029] Understandably, the shorter the rise time, the richer the high-frequency components contained in the step signal, the smaller the skin depth of the signal propagating in the tested wire harness, and the more concentrated the current is on the surface layer of the copper conductor. For high-voltage stranded wire harnesses, this means that the contribution weight of the outer copper wire to the signal transmission impedance increases significantly with the increase of signal frequency. Therefore, using a step signal with a sufficiently short rise time is a prerequisite for the system to have sufficient sensitivity to the small impedance changes caused by the breakage of the outer copper wire at the crimp terminal. The introduction of a directional coupler is the key to ensuring the isolation between the transmitted and received signals. If crosstalk of the transmitted signal enters the receiving path, it will be directly superimposed on the weak broken-strand echo signal, making it impossible for subsequent signal processing to effectively extract the broken-strand characteristics.

[0030] It is understandable that high-voltage stranded wire harnesses are composed of numerous twisted and crimped copper wires, with the conductive cross-section of the copper wires uniformly distributed across the entire conductor cross-section under DC or low-frequency conditions. However, as the signal frequency increases, the skin effect of the electromagnetic field causes the current to gradually concentrate towards the conductor surface, with the skin depth inversely proportional to the square root of the signal frequency. When the rising edge of a step signal is sufficiently short, the skin depth corresponding to the high-frequency components in the signal can be comparable to or even smaller than the diameter of a single copper wire. In this case, the current is almost entirely concentrated in the outermost layer of copper wire in the stranded conductor. Under these conditions, if a portion of the outer copper wire breaks, its impact on the signal transmission impedance has a much greater weight than the proportion of that portion of the copper wire in the total cross-sectional area. Therefore, the impedance change caused by the broken strand is structurally amplified. This is the physical reason why this system can detect a small number of broken copper wires through high-frequency echo signals even when the overall DC resistance of the wire harness remains almost constant. The directional coupler isolates the transmitting and receiving paths on the same physical interface, allowing the system to complete signal injection and echo reception with only one wiring port. This simplifies the connection structure with the wire harness under test and avoids the strong transmitting signal from blocking the weak echo signal.

[0031] A single-point dense sampling unit, which is connected to the pulse transceiver unit, is used to determine the theoretical echo delay window corresponding to the crimping area based on the nominal characteristic impedance and nominal length of the wire harness under test. The echo voltage amplitude is repeatedly sampled a preset number of times at the time corresponding to the theoretical echo delay window to obtain the voltage amplitude sequence. Specifically, the single-point dense sampling unit determines the theoretical echo delay window corresponding to the crimping area based on the nominal characteristic impedance and nominal length of the tested wire harness. At the time corresponding to the theoretical echo delay window, the echo voltage amplitude is repeatedly sampled several times. Each sampling only obtains the voltage amplitude at that time, thus obtaining a voltage amplitude sequence.

[0032] Please continue reading. Figure 2 As shown, it is a logic block diagram of single-point dense sampling and over-value kurtosis calculation of the wire harness connectivity detection system based on high-frequency signal reflection of the present invention. In one specific embodiment, the single-point dense sampling unit calculates the theoretical echo delay of a step signal propagating from the transmitting end to the crimping area and back, based on the nominal characteristic impedance and nominal length of the wire harness under test. Using this theoretical echo delay as the center, a certain margin is extended to both sides to determine the theoretical echo delay window corresponding to the crimping area. Those skilled in the art will understand that the theoretical echo delay is calculated by multiplying the nominal length of the wire harness under test by 2 and then dividing by the signal propagation speed in the wire harness. The signal propagation speed is determined by the nominal characteristic impedance and capacitance per unit length of the wire harness under test. Preferably, the width of the theoretical echo delay window covers the propagation time corresponding to the physical length of the crimping terminal, and a margin not less than the signal rise time is added at both ends to ensure that the echo characteristics of the crimping area completely fall within the window range.

[0033] After determining the theoretical echo delay window, the single-point dense sampling unit locks the sampling clock phase to the time corresponding to that window and repeatedly triggers sampling of the echo signal output by the pulse transceiver unit at that fixed time. Each trigger only records the echo voltage amplitude at that moment, without recording the complete time-domain waveform curve. The voltage amplitudes obtained from all sampling iterations constitute a voltage amplitude sequence. The number of repeated sampling iterations is preferably no less than 100,000, and the sampling process is completed within a preset time window, preferably not exceeding 100 ms. The above-mentioned single-point repeated sampling function is based on the equivalent time sampling technique, which is a mature existing technology and has been widely used in commercial time-domain reflectometers.

[0034] Understandably, the setting of the theoretical echo delay window width needs to consider two aspects: if the window is too wide, it will include the echo signal of the wire harness segment outside the crimped area in the analysis range, introducing irrelevant interference; if the window is too narrow, the echo characteristics of the crimped area may be truncated due to the deviation between the actual length and the nominal length of the wire harness. Therefore, adding a margin based on the rise time at both ends is a necessary measure to ensure the integrity and effectiveness of the window coverage.

[0035] Understandably, the traditional time-domain reflectometry method relies on point-by-point reading and waveform morphology analysis of the complete time-domain waveform to determine the connectivity of the wire harness. The detection sensitivity of this method is limited by the signal-to-noise ratio of a single sample. When the change in echo amplitude caused by a broken strand is submerged in the sampling noise, the waveform morphology hardly undergoes any discernible change, and the defect cannot be identified. The technical solution of this invention concentrates sampling resources on a single moment corresponding to the crimping area, accumulating a large number of repeated sampling values ​​in a very short time, and constructing a statistical distribution of the echo voltage amplitude at that moment. For intact wire harnesses, the reflection coefficient of the crimping area is a fixed quantity, and the voltage amplitude fluctuations obtained from repeated sampling only originate from system thermal noise and quantization noise, and their statistical distribution follows a Gaussian distribution. However, for wire harnesses with partially broken strands, several micro-contact points are formed at the crimping terminals. The contact resistance of each micro-contact point is determined by the contact area, clamping force, and oxide film state, and its distribution follows a log-normal distribution. After multiple such random contact points are connected in parallel, the probability distribution of the equivalent contact impedance exhibits obvious asymmetric thick-tail characteristics, causing the repeated sampling distribution of the echo voltage amplitude at that moment to deviate from a Gaussian shape to a positively skewed thick-tailed shape. This change in statistical distribution cannot be observed from any single sampling value; it can only be revealed through high-order statistical analysis of the overall distribution after accumulating a sufficient number of repeated samples. Therefore, single-point dense sampling is a prerequisite for the subsequent kurtosis calculation unit to extract the broken strand features. The more sampling times, the closer the amplitude probability density histogram is to the true distribution, the more accurate the estimation of the overvalued kurtosis, and the more reliable the identification of some broken strands.

[0036] A kurtosis calculation unit, which is connected to the single-point dense sampling unit, is used to construct an amplitude probability density histogram based on the voltage amplitude sequence and determine the over-value kurtosis according to the amplitude probability density histogram. In one specific embodiment, the kurtosis calculation unit receives a voltage amplitude sequence from a single-point dense sampling unit, divides its amplitude range into several intervals at equal intervals, counts the frequency of amplitude occurrence in each interval, and obtains an amplitude probability density histogram after normalization. The number of intervals is determined based on the total number of samplings of the voltage amplitude sequence, so that the histogram achieves a balance between resolution and statistical stability.

[0037] Specifically, the kurtosis calculation unit in this embodiment of the invention further includes: The probability density construction submodule is used to divide the amplitude range of the voltage amplitude sequence into several intervals at equal intervals, count the frequency of amplitude occurrence in each interval and normalize it to obtain an amplitude probability density histogram. The excessive kurtosis calculation submodule is used to determine the excessive kurtosis K based on the amplitude probability density histogram. exThe excessive kurtosis is the ratio of the fourth central moment of the voltage amplitude sequence to the square of the variance, minus the theoretical kurtosis value of the standard Gaussian distribution. The fourth central moment reflects the contribution weight of extreme values ​​that deviate significantly from the mean in the amplitude sequence, and the square of the variance serves as a dimension normalization. Subtracting the theoretical kurtosis value of the standard Gaussian distribution makes the excessive kurtosis benchmark value corresponding to the intact wire bundle zero. ex It is a dimensionless parameter.

[0038] Understandably, the fourth central moment in the numerator of the formula is highly sensitive to extreme values ​​in the amplitude sequence that deviate significantly from the mean. This is because the deviation term is raised to the fourth power, and the contribution weight of the tail-like samples far from the mean to the numerator is much greater than their proportion in the sample. The denominator, the square of the variance, serves to normalize the dimensions and amplitude scale, making K... ex The value is independent of the absolute magnitude of the signal amplitude, reflecting only the thickness of the tail of the distribution pattern. Subtracting the theoretical kurtosis value of the standard Gaussian distribution reduces the excessive kurtosis of the standard Gaussian distribution to zero, ensuring that the normal output value of a healthy wire harness is theoretically zero. The positive deviation value directly quantifies the degree to which the distribution tail exceeds the Gaussian distribution, facilitating threshold setting. The preset kurtosis threshold is determined by adding the mean of the excessive kurtosis to the preset multiple of the standard deviation of the excessive kurtosis after sampling and statistically analyzing the no-load background noise of qualified wire harnesses of the same model.

[0039] Understandably, excessive kurtosis is a fourth-order statistic of the probability distribution, specifically used to measure the thickness of the distribution tails relative to a standard Gaussian distribution. For intact wire harnesses, the high-frequency echo reflection coefficient of the crimped area is a stable and deterministic quantity. The voltage amplitude fluctuations obtained by repeated sampling at this moment are entirely due to the thermal noise of the front-end amplifier and the quantization noise of the analog-to-digital converter. These two types of noise are physically independent and have bounded amplitudes. Their superposition distribution is constrained by the central limit theorem, closely approximating a Gaussian distribution, and the excessive kurtosis is close to zero. For wire harnesses with partially broken strands, the number of effective micro-contact points at the crimped terminals decreases due to copper wire breakage, increasing the current density borne by the remaining contact points. The contact resistance of each micro-contact point is determined by the contact area, the thickness of the surface oxide film, and the local clamping force. These parameters exhibit significant randomness at the microscale, and their statistical distribution follows a log-normal distribution. When multiple random micro-contact points are connected in parallel, the probability distribution of the equivalent contact impedance retains the positively skewed and thick-tailed characteristics of the log-normal distribution. That is, the probability density is concentrated near the mean, but the right tail of the distribution extends significantly longer than the left, making the probability of high impedance extreme values ​​much higher than expected by a Gaussian distribution. This asymmetric thick-tailed morphology directly modulates the successive sampling values ​​of the high-frequency reflection coefficient in the crimping region, causing the probability distribution of the voltage amplitude sequence to also exhibit a thick-tailed skewness, resulting in a measurable positive deviation in the super-value kurtosis. Since the fourth-order statistic is much more sensitive to the tail of the distribution than the first-order mean and second-order variance, even if the change in the mean of the reflection amplitude caused by the broken strand is extremely small and completely submerged in noise, the super-value kurtosis can still capture this tail structure change from the overall distribution pattern of a large number of repeated samples, thus achieving statistical identification of the micro-broken strand characteristics.

[0040] The baseline management unit contains a database of qualified sample reflection waveforms indexed by wire harness model identifier and ambient temperature. It is used to retrieve and output the corresponding ideal reference waveform based on the model identifier of the wire harness under test and the current ambient temperature. Please continue reading. Figure 3 As shown, it is a logic block diagram of the baseline management and peak-to-average ratio calculation of the harness connectivity detection system based on high-frequency signal reflection of the present invention. Specifically, the baseline management unit pre-stores the measured reflection waveform of the first qualified sample in the same batch as the batch baseline, and the measured reflection waveform of the tested wire harness after its connectivity is confirmed as the historical baseline.

[0041] Specifically, the baseline management unit uses the historical baseline as the initial ideal reference waveform based on the judgment result that the historical baseline exists; the baseline management unit uses the corresponding waveform retrieved from the qualified sample reflection waveform database as the initial ideal reference waveform based on the judgment result that the historical baseline does not exist and the model identifier of the tested wire harness has a corresponding record in the qualified sample reflection waveform database.

[0042] Specifically, the baseline management unit uses the batch baseline as the initial ideal reference waveform based on the judgment result that the historical baseline does not exist and the model identifier of the tested wire harness does not have a corresponding record in the qualified sample reflection waveform database.

[0043] Specifically, the baseline management unit corrects the initial ideal reference waveform based on the difference between the current ambient temperature and the ambient temperature at which the selected initial ideal reference waveform was stored, thereby obtaining the ideal reference waveform.

[0044] In one specific embodiment, the baseline management unit's non-volatile memory pre-stores a database of reflected waveforms from qualified samples. The database is organized and stored using the harness model identifier and the ambient temperature at the time of acquisition as dual index fields. Each record contains complete measured reflected waveform data for the corresponding model of qualified sample, along with the ambient temperature value at the time of acquisition. During the product model development phase, the waveform data in the database is written to samples that have passed full-item electrical performance certification after actual measurement and acquisition. During acquisition, the sample connection status must be consistent with the actual testing station.

[0045] The baseline management unit maintains two types of dynamic baselines: one is the batch baseline, which is the measured reflection waveform of the first sample in the same production batch that passes the appearance and size inspection at the testing station. The batch baseline is automatically updated as the batch changes. The other is the historical baseline, which is the measured reflection waveform automatically stored by the system when the tested wire harness outputs a continuity qualified conclusion in a previous test. The historical baseline is stored separately with the unique code of the tested wire harness as an index.

[0046] The baseline management unit determines the initial ideal reference waveform according to the following priorities: if the historical baseline of the tested wire harness exists, the historical baseline is used as the initial ideal reference waveform; if the historical baseline does not exist, and the model identifier of the tested wire harness has a corresponding record in the qualified sample reflection waveform database, the corresponding waveform retrieved from the database is used as the initial ideal reference waveform; if the historical baseline does not exist and there is no corresponding record in the qualified sample reflection waveform database, the batch baseline is used as the initial ideal reference waveform.

[0047] In one specific embodiment, the baseline management unit includes a temperature correction submodule, which is used to perform temperature correction on the initial ideal reference waveform to obtain the ideal reference waveform, denoted as Video(t). The temperature correction is based on the temperature coefficient of resistance of copper conductors. According to the difference between the current ambient temperature and the ambient temperature when the selected initial ideal reference waveform was stored, the amplitude of the initial ideal reference waveform Vinit(t) at each time point is proportionally calculated to eliminate the influence of temperature changes on the characteristic impedance of the harness and the amplitude of the echo reflection.

[0048] It is understandable that the characteristic impedance of the wire harness is proportional to the square root of the resistance per unit length, which in turn is proportional to the resistivity of copper. The resistivity of copper changes with temperature, described by a linear temperature coefficient of resistance. Therefore, when there is a deviation between the ambient temperature during detection and the ambient temperature during baseline acquisition, the characteristic impedance of the wire harness changes accordingly, and the amplitude of the echo reflection changes. If a direct comparison is performed without correction, a temperature-related systematic bias will be introduced into the logarithmic residual curve, interfering with the subsequent integral area calculation. After introducing the above temperature correction, the ideal reference waveform is converted to the expected amplitude under the current temperature conditions, eliminating the influence of temperature drift on the baseline comparison results.

[0049] Understandably, the decision of the logarithmic residual integral unit depends on the minute difference between the actual echo waveform and the ideal reference waveform. The closer the ideal reference waveform is to the physical state of the tested wire harness, the smaller the systematic bias introduced by individual wire harness differences and station connection differences in the logarithmic ratio residual curve, and the higher the detection sensitivity of the strand breakage feature. The historical baseline comes from the measured waveform of the tested wire harness itself in a qualified state. It naturally contains all individual information such as the geometric dimension deviation, crimping process deviation, and connection state of the testing station of the wire harness. It has the strongest comparability with the current measured waveform and therefore has the highest priority. The waveforms in the qualified sample reflection waveform database come from qualified samples of the same model. They reflect the common characteristics of the wire harness of this model in the standard state and are suitable for the first test or when the historical baseline is missing. Their comparability is second best. The batch baseline comes from the first sample of the same batch. It reflects the characteristics of the wire harness under the current production process state. It is used as a fallback baseline in the first test of a new model wire harness that is not recorded in the model database. It ensures that the system can provide an effective comparison benchmark under any circumstances and avoids the inability to detect due to the lack of a baseline. Temperature correction, based on the three-level baseline mechanism mentioned above, further eliminates the impact of ambient temperature changes on baseline effectiveness, ensuring that the baseline can accurately reflect the expected echo amplitude of the tested wire harness under qualified conditions under different testing temperature conditions.

[0050] The logarithmic residual integration unit is connected to the pulse transceiver unit, the single-point dense sampling unit, and the baseline management unit, respectively. It is used to construct a logarithmic ratio residual curve based on the actual echo waveform and the ideal reference waveform, and to perform time integration on the logarithmic ratio residual curve within the theoretical echo delay window to obtain the integration area. Specifically, the logarithmic residual integration unit uses the logarithmic ratio residual value at each time point as the logarithmic ratio residual value of the actual echo waveform at each time point after adding a zero constant to the absolute value of the amplitude of the ideal reference waveform at the corresponding time point. The logarithmic ratio residual value at each time point is used to continuously form the logarithmic ratio residual curve.

[0051] In one specific embodiment, the logarithmic residual integration unit obtains the actual echo waveform digital data of the tested wire harness from the pulse transceiver unit, obtains the ideal reference waveform Video(t) after temperature correction from the baseline management unit, and obtains the start and end times of the theoretical echo delay window from the single-point dense sampling unit.

[0052] Specifically, the logarithmic residual integration unit described in this embodiment of the invention further includes: The logarithmic ratio residual construction submodule is used to take the logarithmic ratio residual value R(t) at the corresponding time point as the logarithmic ratio residual value R(t) based on the absolute value of the amplitude of the actual echo waveform at each time point plus a zero-prevention constant and the absolute value of the amplitude of the ideal reference waveform Video(t) at the corresponding time point. The logarithmic ratio residual value at each time point is continuously used to construct the logarithmic ratio residual curve. The integral area calculation submodule is used to perform time integration on the logarithmic ratio residual curve R(t) within the theoretical echo delay window to obtain the integral area S. S is a signed number. A negative S indicates that the actual echo energy of the tested wire harness within the theoretical echo delay window has a net attenuation relative to the ideal reference waveform Video(t). A positive S or S close to zero indicates that the actual echo energy is equivalent to or higher than the ideal reference waveform Video(t). In the mathematical implementation, the above integral area S is approximated by a discrete summation method, that is, by multiplying the R(t) values ​​at each discrete time point within the window by the sampling time interval and then summing them.

[0053] Understandably, the introduction of the zero-prevention constant is a numerical stability protection measure. Its value must be much smaller than the minimum resolvable voltage amplitude of the system to ensure that, within the region where the ideal reference waveform Video(t) is normally non-zero, the effect of the zero-prevention constant on the logarithmic ratio residual value is negligible and does not change the physical meaning of R(t). R(t) is in dB: the larger the negative absolute value of S, the deeper the attenuation of the actual echo energy in the crimped area relative to the qualified reference, and the more severe the breakage.

[0054] Understandably, given that the kurtosis calculation unit has already determined the statistical distribution anomaly through excessive kurtosis, further verification and quantification of the anomaly's nature from an energy perspective are necessary. In the linear domain, the minute difference between the actual echo amplitude and the ideal reference waveform is directly subtracted, and the result is on the same order of magnitude as the DC drift and low-frequency interference of the system amplifier, thus obscuring the difference characteristics. However, after taking the logarithm of the amplitude ratio, the slope of the logarithmic function is extremely large near the independent variable 1, causing the minute deviation in the amplitude ratio to be nonlinearly amplified into a recognizable bias in the logarithmic domain. This amplification is effective in both the amplitude attenuation and amplitude increase directions, thus the logarithmic ratio residual curve can reflect the subtle amplitude deviation of the actual echo waveform relative to the ideal reference waveform at various time points. By integrating the logarithmic ratio residual curve within the theoretical echo delay window and accumulating and superimposing the logarithmic ratio residual values ​​at all times within the window on the time axis, the minute bias at a single moment forms a statistically significant area quantity after integration, further improving the detection sensitivity of weak segmentation characteristics. Strictly limiting the integration range to within the theoretical echo delay window is to eliminate the interference contribution of other impedance discontinuities outside the crimped area of ​​the tested wire harness to the integration area, so that the integration area S specifically reflects the echo energy state of the crimped terminal area, ensuring the relevance of subsequent decision-making.

[0055] The peak-to-average power ratio (PAPR) calculation unit, which is connected to the logarithmic residual integration unit, is used to calculate the ratio of the maximum absolute value of the amplitude of the logarithmic ratio residual curve to the absolute value of the amplitude mean within the theoretical echo delay window, thereby obtaining the PAPR. In one specific embodiment, the peak-to-average power ratio (PAPR) calculation unit obtains the complete logarithmic ratio residual curve R(t) and the start and end times of the theoretical echo delay window from the logarithmic residual integration unit, and calculates the PAPR C.

[0056] Specifically, the peak-to-average power ratio (PAPR) calculation unit in this embodiment of the invention further includes: The maximum amplitude determination submodule is used to determine the maximum absolute value of the amplitude of the logarithmic ratio residual curve R(t) within the theoretical echo delay window. The mean amplitude determination submodule is used to determine the arithmetic mean of the absolute values ​​of the amplitudes of the logarithmic ratio residual curve R(t) within the theoretical echo delay window. The peak-to-average ratio (PAPR) calculation submodule is used to determine the PAPR C as the ratio of the maximum absolute value of the amplitude to the arithmetic mean of the absolute values ​​of the amplitude, where C is a dimensionless parameter, C≥1, and C=1 corresponds to the uniform case where the absolute values ​​of the logarithmic ratio residual amplitudes are completely equal at each time point within the window.

[0057] The preset peak-to-average power ratio (PAPR) threshold is determined based on statistical measurements of known micro-fractured strand samples and known uniform impedance offset samples of wire harnesses.

[0058] Understandably, the numerator is set to the maximum absolute value of the amplitude to capture local extreme deviations in the logarithmic ratio residual curve within the window, regardless of whether the deviation is positive or negative. The denominator is set to the mean of the absolute values ​​of the amplitudes, rather than the algebraic mean, to avoid the situation where positive and negative residual values ​​cancel each other out, causing the mean to approach zero and the peak-to-average ratio (PAR) to become meaningless. Dividing both by their absolute values ​​ensures that C specifically reflects the spatial concentration of the residual curve within the window, regardless of the positive or negative direction of the residuals. The calibration of the preset PAR threshold needs to cover the typical models and typical strand breakage ratio ranges of the tested wire harnesses to ensure that the threshold has distinguishing power on wire harnesses of different specifications.

[0059] Understandably, micro-strand breakage occurs locally at the crimp terminal, and its impact on the echo signal is concentrated in the time domain near the specific moment corresponding to the crimp terminal. This manifests as one or more large-amplitude local negative peaks on the logarithmic ratio residual curve, while the residual value at other moments within the window is close to zero. Uniform impedance shift, on the other hand, is caused by factors such as bending, localized temperature rise, or insulation layer compression. Its effect is more uniformly distributed along the length of the wire harness, and the corresponding logarithmic ratio residual curve shows a uniform overall downward shift throughout the entire theoretical echo delay window, without any prominent local peaks. Since the integral area S in both cases may exhibit a negative bias, the value of S alone cannot distinguish between the two. Introducing the peak-to-average power ratio (PAPR) reveals that the residual curve corresponding to micro-strand breakage has a large PAPR due to the prominent local peaks, while the residual curve corresponding to uniform impedance shift has a PAPR close to 1 due to the similar amplitude at each point. The two cases show a significant difference in PAPR values, thus effectively distinguishing them. In addition, although external transient electromagnetic interference may also produce local spikes on the logarithmic ratio residual curve, its direction is random, with both positive and negative spikes, and the interference is not limited to the theoretical echo delay window. It differs from the negative local spikes caused by micro-fractures in terms of directionality and location specificity. This difference has been partially addressed by the joint decision of the kurtosis calculation unit and the integral area. On this basis, the peak-to-average power ratio further provides supplementary morphological constraints, so that the three characteristic quantities together constitute a complete description of the state of the measured wire harness.

[0060] The decision output unit is connected to the kurtosis calculation unit, the logarithmic residual integration unit, and the peak-to-average ratio (PARR) calculation unit, respectively, and is used to output a connectivity detection conclusion based on the comparison results of the excess kurtosis, the integral area, and the PARR with the corresponding preset threshold.

[0061] Please continue reading. Figure 4 The diagram shown is a logic block diagram of the decision output unit of the wire harness connectivity detection system based on high-frequency signal reflection of the present invention. Specifically, the decision output unit pre-stores a kurtosis preset threshold corresponding to the excessive kurtosis, an integral preset threshold corresponding to the integral area, and a peak-to-average ratio preset threshold corresponding to the peak-to-average ratio. If the super kurtosis is less than or equal to the kurtosis preset threshold, output that the connectivity is qualified; If the super kurtosis is greater than the kurtosis preset threshold and the integral area is greater than or equal to the negative integral preset threshold, output an external electromagnetic interference determination; If the super kurtosis is greater than the kurtosis preset threshold, the integral area is less than the negative integral preset threshold, and the peak-to-average ratio is greater than the peak-to-average ratio preset threshold, output a micro broken strand warning; If the super kurtosis is greater than the kurtosis preset threshold, the integral area is less than the negative integral preset threshold, and the peak-to-average ratio is less than or equal to the peak-to-average ratio preset threshold, output a harness impedance uniform offset determination.

[0062] In a specific embodiment, the decision output unit obtains the super kurtosis K from the kurtosis calculation unit ex , obtains the integral area S from the logarithmic residual integral unit, obtains the peak-to-average ratio C from the peak-to-average ratio calculation unit, and compares the above three characteristic quantities with the pre-stored kurtosis preset threshold ε1, integral preset threshold ε2, and peak-to-average ratio preset threshold ε3 respectively, and outputs the connectivity detection conclusion: First branch: If K ex ≤ε1, output "connectivity qualified"; Second branch: If K ex >ε1 and S≥-ε2, output "external electromagnetic interference determination"; Third branch: If K ex >ε1, S<-ε2, and C>ε3, output "micro broken strand warning"; Fourth branch: If K ex >ε1, S<-ε2, and C≤ε3, output "harness impedance uniform offset determination".

[0063] Among them, ε1 is the kurtosis preset threshold, dimensionless, which is determined by taking the mean of the background super kurtosis plus the standard deviation of the background super kurtosis multiplied by a preset multiple after sampling and statistics of the no-load background noise of qualified harnesses of the same model; ε2 is the integral preset threshold, taking a positive number, -ε2 represents the negative determination boundary of the integral area, which is determined based on the statistical distribution of the integral area of qualified harnesses of the same model under normal working conditions; ε3 is the peak-to-average ratio preset threshold, dimensionless, which is determined based on the actual measurement and statistics of known micro broken strand samples and known harness uniform impedance offset samples.

[0064] Understandably, excessive kurtosis is the first criterion for judgment. Its physical meaning is whether the statistical distribution of the echo voltage amplitude at the crimped terminal deviates from Gaussian characteristics. Deviation indicates the presence of an abnormal source with increased randomness at that location. If there is no deviation, broken strands and interference are directly excluded, resulting in a qualified conclusion and avoiding unnecessary subsequent analysis of normal harnesses. The integral area is the second criterion for judgment. Its physical meaning is whether the echo energy in the crimped area shows a net attenuation relative to the qualified benchmark. Net attenuation means an increase in the high-frequency impedance of the crimped area, which is a sign of broken strands or structural changes in impedance. If the excessive kurtosis is abnormal but the integral area does not show a negative net attenuation, it indicates that the disturbance in the statistical distribution comes from external energy injection rather than an increase in internal losses, and is judged as external electromagnetic interference. Peak-to-average power ratio (PAPR) is the third criterion for judgment. Its physical meaning is whether the energy of the logarithmic ratio residual curve in the crimped area is concentrated or uniformly distributed on the time axis. Concentrated distribution corresponds to the local impedance change caused by strand breakage, while uniform distribution corresponds to the overall impedance drift of the entire wire harness caused by bending or temperature rise. Under the premise that the integral area has been confirmed to have negative net attenuation, the size of PAPR can distinguish between these two impedance changes with different properties, so that the detection conclusion has a clear fault characterization significance and provides a basis for subsequent maintenance and handling.

[0065] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A wire harness connectivity detection system based on high-frequency signal reflection, characterized in that, include: The pulse transceiver unit is used to transmit high-frequency step signals to the wire harness under test and receive echo signals. A single-point dense sampling unit, which is connected to the pulse transceiver unit, is used to determine the theoretical echo delay window corresponding to the crimping area based on the nominal characteristic impedance and nominal length of the wire harness under test. The echo voltage amplitude is repeatedly sampled a preset number of times at the time corresponding to the theoretical echo delay window to obtain the voltage amplitude sequence. A kurtosis calculation unit, which is connected to the single-point dense sampling unit, is used to construct an amplitude probability density histogram based on the voltage amplitude sequence and determine the over-value kurtosis according to the amplitude probability density histogram. The baseline management unit contains a database of qualified sample reflection waveforms indexed by wire harness model identifier and ambient temperature. It is used to retrieve and output the corresponding ideal reference waveform based on the model identifier of the wire harness under test and the current ambient temperature. The logarithmic residual integration unit is connected to the pulse transceiver unit, the single-point dense sampling unit, and the baseline management unit, respectively. It is used to construct a logarithmic ratio residual curve based on the actual echo waveform and the ideal reference waveform, and to perform time integration on the logarithmic ratio residual curve within the theoretical echo delay window to obtain the integration area. The peak-to-average power ratio (PAPR) calculation unit, which is connected to the logarithmic residual integration unit, is used to calculate the ratio of the maximum absolute value of the amplitude of the logarithmic ratio residual curve to the absolute value of the amplitude mean within the theoretical echo delay window, thereby obtaining the PAPR. The decision output unit is connected to the kurtosis calculation unit, the logarithmic residual integration unit, and the peak-to-average ratio (PARR) calculation unit, respectively, and is used to output a connectivity detection conclusion based on the comparison results of the excess kurtosis, the integral area, and the PARR with the corresponding preset threshold.

2. The wire harness connectivity detection system according to claim 1, characterized in that, The pulse transceiver unit includes a fast-edge pulse generator and a high-speed analog-to-digital converter; The fast-edge pulse generator generates a step signal and transmits it to the wire harness under test. The high-speed analog-to-digital converter receives the echo signal and converts it into digital waveform data. The fast-edge pulse generator and the high-speed analog-to-digital converter are connected by a directional coupler to form a closed-loop signal path.

3. The wire harness connectivity detection system according to claim 1, characterized in that, The single-point dense sampling unit determines the theoretical echo delay window corresponding to the crimping area based on the nominal characteristic impedance and nominal length of the tested wire harness. At the time corresponding to the theoretical echo delay window, the echo voltage amplitude is repeatedly sampled several times. Each sampling only obtains the voltage amplitude at that time, thus obtaining a voltage amplitude sequence.

4. The wire harness connectivity detection system according to claim 1, characterized in that, The baseline management unit pre-stores the measured reflection waveform of the first qualified sample in the same batch as the batch baseline, and the measured reflection waveform of the tested wire harness after its connectivity is confirmed as the historical baseline.

5. The wire harness connectivity detection system according to claim 4, characterized in that, Based on the determination of the existence of the historical baseline, the baseline management unit uses the historical baseline as the initial ideal reference waveform; Based on the judgment result that the historical baseline does not exist and the model identifier of the tested wire harness has a corresponding record in the qualified sample reflection waveform database, the baseline management unit uses the corresponding waveform retrieved from the qualified sample reflection waveform database as the initial ideal reference waveform.

6. The wire harness connectivity detection system according to claim 5, characterized in that, Based on the judgment result that the historical baseline does not exist and there is no corresponding record of the model identification of the measured wire harness in the qualified sample reflection waveform database, the baseline management unit uses the batch baseline as the initial ideal reference waveform.

7. The wire harness connectivity detection system according to claim 6, characterized in that, The baseline management unit corrects the initial ideal reference waveform according to the difference between the current ambient temperature and the ambient temperature when the selected initial ideal reference waveform is stored, and obtains the ideal reference waveform.

8. The wire harness connectivity detection system according to claim 2, characterized in that, The rise time of the high-frequency step signal emitted by the fast-edge pulse generator is less than the minimum resolvable time interval corresponding to the system background noise.

9. The wire harness connectivity detection system according to claim 1, characterized in that, The logarithmic residual integration unit uses the logarithm of the ratio of the absolute value of the amplitude of the actual echo waveform at each time point plus the anti-zero constant to the absolute value of the amplitude of the ideal reference waveform at the corresponding time point plus the anti-zero constant as the logarithmic ratio residual value at this time point, and continuously constitutes the logarithmic ratio residual curve from the logarithmic ratio residual values at each time point.

10. The wire harness connectivity detection system according to claim 1, characterized in that, The decision output unit pre-stores the kurtosis preset threshold corresponding to the excessive kurtosis, the integral preset threshold corresponding to the integral area, and the peak-to-average ratio preset threshold corresponding to the peak-to-average ratio respectively; If the excessive kurtosis is less than or equal to the kurtosis preset threshold, it outputs that the connectivity is qualified; If the excessive kurtosis is greater than the kurtosis preset threshold and the integral area is greater than or equal to the negative integral preset threshold, it outputs the determination of external electromagnetic interference; If the excessive kurtosis is greater than the kurtosis preset threshold, the integral area is less than the negative integral preset threshold, and the peak-to-average ratio is greater than the peak-to-average ratio preset threshold, it outputs the micro-breakage warning; If the excessive kurtosis is greater than the kurtosis preset threshold, the integral area is less than the negative integral preset threshold, and the peak-to-average ratio is less than or equal to the peak-to-average ratio preset threshold, it outputs the determination of uniform impedance offset of the wire harness.

Images