Fault detection method and system for DAC high-speed cable

Abstract

The invention relates to the technical field of cable fault detection, in particular to a fault detection method and system for a DAC high-speed cable, and the method comprises the steps: decomposing test data, obtaining a plurality of component signals, and obtaining the change degree of the component signals according to the fluctuation of each component signal; setting sliding windows, and obtaining the direction anisotropy of adjacent sliding windows according to the change trend of data in the sliding windows; further obtaining the change complexity of each component signal; the attenuation possibility of the time domain data is further obtained; obtaining the variability of the time domain data according to the attenuation possibility of the time domain data and the opening height and the opening width of the eye pattern data; the attenuation degree of the time domain data is obtained; abnormal data are obtained through screening according to the attenuation degree of the time domain data, and the fault of the DAC high-speed cable is detected; according to the method, signal attenuation caused by electromagnetic interference and the like in the data transmission process of the DAC high-speed cable can be accurately analyzed, early warning is performed in time, and signal loss is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of cable fault detection, and particularly relates to a method and system for fault detection of DAC high-speed cables. Background Art

[0002] A DAC high-speed cable, namely a Direct Attach Cable, is a cable assembly with a fixed length, both ends of which are equipped with fixed connectors, the interfaces are non-replaceable, and the module connectors cannot be separated from the copper cable; the DAC high-speed cable does not contain optical lasers and electronic components, so it can only transmit electrical signals instead of optical signals. This characteristic enables the DAC high-speed cable to significantly reduce costs and power consumption in short-distance applications; with the continuous increase in the network bandwidth requirements of data centers, the DAC high-speed cable is widely used in the connection between servers and switches due to its high efficiency, low latency, and economy. However, the DAC cable is prone to physical damage, signal attenuation, and electromagnetic interference during long-term use, resulting in a decline in network performance and link interruption; testing the signal attenuation of the DAC high-speed cable can help confirm the integrity and reliability of the signal during transmission; by detecting the attenuation degree, problems that may lead to an increase in the bit error rate can be identified in a timely manner to ensure the correctness of data transmission.

[0003] In the prior art, the full-automatic SI high-frequency parameter test is an efficient and accurate method for evaluating signal integrity in high-speed circuits; its main purpose is to detect and analyze problems such as attenuation, reflection, crosstalk, and distortion that high-speed signals may suffer during transmission. However, when analyzing the signals obtained from the SI high-frequency parameter test, since the skin effect and dielectric loss cause the insertion loss to grow non-linearly with frequency when the DAC cable transmits high-frequency signals, the non-linear characteristics will distort the signal waveform, resulting in large errors when analyzing the test data. Summary of the Invention

[0004] In order to solve the above problems, the present invention proposes a method and system for fault detection of DAC high-speed cables.

[0005] An embodiment of the present invention provides a method for fault detection of DAC high-speed cables, the method comprising the following steps:

[0006] Obtain test data of the DAC high-speed cable, the test data including time-domain data and eye diagram data;

[0007] Decompose the time-domain data to obtain multiple component signals, obtain the degree of change of the component signals according to the fluctuations of the component signals; adjust to obtain a sliding window according to the degree of change of the component signals; obtain the directional anisotropy of adjacent sliding windows according to the change trend of the data within the sliding window; combine the numerical magnitudes of the data within the sliding window to obtain the change complexity of each component signal; obtain the attenuation possibility of the time-domain data according to the change complexity and frequency of each component signal;

[0008] Obtain the variability of the time-domain data according to the attenuation possibility of the time-domain data, as well as the opening height and opening width of the eye diagram data; obtain the attenuation degree of the time-domain data according to the variability of the time-domain data and the distribution characteristics of the eye diagram data;

[0009] Screen out abnormal data according to the attenuation degree of the time-domain data, and detect the faults of the DAC high-speed cable.

[0010] Optionally, the steps of decomposing the time-domain data to obtain multiple component signals and obtaining the degree of change of the component signals according to the fluctuations of the component signals include the following specific steps:

[0011] Perform EMD decomposition on the time-domain data to obtain multiple component signals; obtain the degree of change of the component signals, and its calculation formula is as follows:

[0012]

[0013] In the formula, p represents the degree of change of the component signal, N represents the number of component signals, s l represents the variance of all amplitudes in the l-th component signal, and sigmoid() represents the sigmoid function.

[0014] Optionally, the steps of adjusting to obtain a sliding window according to the degree of change of the component signals include the following specific steps:

[0015] Preset the initial length of the sliding window, and adjust the length of the sliding window according to the degree of change of the component signals, and its calculation formula is as follows:

[0016]

[0017] In the formula, T represents the length of the sliding window, p represents the degree of change of the component signal, A represents the preset length, represents the ceiling function.

[0018] Optionally, the steps of obtaining the directional anisotropy of adjacent sliding windows according to the change trend of the data within the sliding window include the following specific steps:

[0019] For each component signal, a number of sliding windows are obtained based on the length of the sliding window. The step size of the sliding window is equal to the length, so there is no duplicate data between different segments of the sliding window. The data within any segment of the sliding window is subjected to least squares fitting to obtain a fitting curve. The principal component direction of the fitting curve is obtained through the PCA algorithm and used as the data distribution direction of the sliding window.

[0020] The calculation formula for obtaining the direction anisotropy of adjacent sliding windows is as follows:

[0021] q l(i,i+1) = θ li × θ l(i+1)

[0022] In the formula, q l(i,i+1) represents the direction anisotropy between the i-th sliding window and the (i + 1)-th sliding window of the l-th component signal, θ li represents the data distribution direction of the i-th sliding window of the l-th component signal, and θ l(i+1) represents the data distribution direction of the (i + 1)-th sliding window of the l-th component signal.

[0023] Optionally, the steps for obtaining the change complexity of each component signal are as follows:

[0024]

[0025] In the formula, V l represents the change complexity of the l-th component signal, represents the maximum value of the data means in all sliding windows of the l-th component signal, represents the minimum value of the data means in all sliding windows of the l-th component signal, n represents the number of sliding windows in the component signal, and q l(i,i+1) represents the direction anisotropy between the i-th sliding window and the (i + 1)-th sliding window of the l-th component signal, and exp[] represents the exponential function with the natural constant as the base.

[0026] Optionally, the steps for obtaining the attenuation possibility of the time-domain data based on the change complexity and frequency of each component signal are as follows:

[0027]

[0028] In the formula, w represents the attenuation possibility of the time-domain data, f l represents the frequency of the l-th component signal, V l represents the change complexity of the l-th component signal, N represents the number of component signals, and sigmoid() represents the sigmoid function.

[0029] Optionally, obtaining the variability of the time-domain data based on the attenuation possibility of the time-domain data, and the opening height and opening width of the eye diagram data includes the following specific steps:

[0030]

[0031] In the formula, K represents the variability of the time-domain data, w represents the attenuation possibility of the time-domain data, H represents the opening height of the eye diagram, and E represents the opening width of the eye diagram.

[0032] Optionally, obtaining the attenuation degree of the time-domain data based on the variability of the time-domain data and the distribution characteristics of the eye diagram data includes the following specific steps:

[0033] Obtain the upper vertex and lower vertex of the eye diagram, and two intermediate intersection points; fold the eye diagram in half with the line connecting the upper vertex and lower vertex as the axis of symmetry to obtain the spatial distance between the two intermediate intersection points, and use the template matching algorithm to match the eye diagrams on both sides of the axis of symmetry to obtain the matching degree of the eye diagrams on both sides of the axis of symmetry;

[0034] The calculation method for obtaining the attenuation degree of the time-domain data is as follows:

[0035]

[0036] In the formula, γ represents the attenuation degree of the time-domain data, K represents the variability of the time-domain data, d represents the spatial distance between the two intermediate intersection points, g represents the matching degree of the eye diagrams on both sides of the axis of symmetry, and sigmoid() represents the sigmoid function.

[0037] Optionally, screening out abnormal data according to the attenuation degree of the time-domain data includes the following specific steps:

[0038] Preset a fault threshold. If the attenuation degree of the time-domain data is greater than the fault threshold, the test data is abnormal, and the DAC high-speed cable has a fault.

[0039] Another embodiment of the present invention provides a fault detection system for a DAC high-speed cable, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above method are implemented.

[0040] The beneficial effects of the technical solution of the present invention are as follows: During the analysis of the attenuation characteristics of test data, first, the time-domain data of the test data is decomposed to obtain multiple decomposed signals; then, the change complexity of the component signals is obtained based on the changes of the component signals; further, the attenuation possibility of the time-domain data is obtained; because the time-domain data is greatly affected by interference and the degree of data change is large, the long-time sequence change of the signal is analyzed through the eye diagram data; the variability of the time-domain data is obtained based on the opening height, opening width in the eye diagram data and the attenuation degree of the time-domain data; then, the attenuation degree of the time-domain data is obtained based on the variability of the time-domain data; the fault condition of the tested DAC high-speed cable is obtained based on the attenuation degree of the time-domain data. Through this method, it is possible to accurately analyze the signal attenuation caused by electromagnetic interference and other situations during the data transmission of the DAC high-speed cable, so as to perform timely manual intervention and avoid signal loss during transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0042] Figure 1 is a flowchart of the steps of the fault detection method for the DAC high-speed cable of the present invention;

[0043] Figure 2 is an example diagram of the time-domain data in the test data;

[0044] Figure 3 is an example diagram of the eye diagram data in the test data;

[0045] Figure 4 is a schematic diagram of the distribution characteristics of the eye diagram data. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following, in combination with the drawings and preferred embodiments, details the specific implementation manners, structures, features and effects of the fault detection method for the DAC high-speed cable proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0047] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0048] The following specifically describes the specific solution of the fault detection method for the DAC high-speed cable provided by the present invention in conjunction with the accompanying drawings.

[0049] Please refer to Figure 1 , which shows a flowchart of the steps of the fault detection method for the DAC high-speed cable provided by an embodiment of the present invention. The method includes the following steps:

[0050] Step S001, obtain the test data of the DAC high-speed cable.

[0051] I. Test system architecture design:

[0052] 1. Core device configuration:

[0053] 1) High-speed oscilloscope: bandwidth ≥ 100 GHz, used for eye diagram capture and dynamic signal analysis.

[0054] 2. Auxiliary module:

[0055] 1) Temperature control environmental chamber: maintain a constant temperature (25 ± 0.5 °C) and humidity (≤ 60% RH) to reduce the impact of material Dk / Df fluctuations on the test;

[0056] 2) Automatic fixture system: use low-loss SMA / 2.92 mm connectors with an impedance matching accuracy of ± 2%.

[0057] II. Key steps of experimental layout:

[0058] 1. Fixture calibration and de-embedding processing:

[0059] 1) Use a SOLT (Short-Open-Load-Thru) calibration kit to eliminate fixture parasitic parameters (such as SMA connector capacitance > 1 pF);

[0060] 2) Compensate for fixture losses through the built-in de-embedding algorithm of the VNA (error < 0.05 dB / 40 GHz).

[0061] 2. Fixing and connecting the DAC cable:

[0062] 1) Connect both ends of the DAC cable to the automatic fixture system to ensure uniform distribution of mechanical stress (bending radius ≥ 5 times the wire diameter);

[0063] 2) Let it stand for ≥ 12 hours before testing to eliminate impedance fluctuations caused by temperature drift (impedance tolerance ± 5%).

[0064] 3. Test parameter configuration:

[0065] 1) Frequency domain test: Set the VNA scanning frequency range to 0.1 - 40 GHz, and measure S11 (return loss), S21 (insertion loss), and crosstalk (S31 / S41).

[0066] 2) Time domain test: The TDR sends a 10 ps rising edge pulse to locate the impedance mutation point (resolution ±1 mm).

[0067] 3) Dynamic test: The BERT sends a PRBS31 pattern, and the oscilloscope captures the eye diagram and calculates the eye height (≥50 mV @ 112 Gbps) and the total jitter (≤0.1 UI).

[0068] III. Data acquisition process:

[0069] 1. First, perform system initialization. Start the temperature control environmental chamber and load the preset test script (such as a Python or LabVIEW program), and synchronously connect core devices such as a vector network analyzer (VNA), a high-speed oscilloscope, and a bit error rate tester (BERT). In this stage, strictly monitor the environmental temperature stability (fluctuation ≤ ±0.5 °C) to eliminate the influence of external temperature drift on the test.

[0070] 2. Then, perform calibration and compensation operations. Calibrate the test fixture through a SOLT (Short-Open-Load-Thru) calibration kit to eliminate the parasitic parameters of the fixture (such as connector capacitance and inductance).

[0071] 3. After calibration, enter the S-parameter scanning stage. The VNA performs multi-port synchronous measurement on the DAC cable to obtain full-frequency domain parameters including return loss (S11), insertion loss (S21), and crosstalk (S31 / S41), and obtain time domain data and eye diagram data. The time domain data is as Figure 2 shown, and the eye diagram data is as Figure 3 shown.

[0072] Step S002: Decompose the time domain data to obtain multiple component signals, and obtain the change degree of the component signals according to the fluctuations of the component signals; adjust to obtain a sliding window according to the change degree of the component signals; obtain the directionality of adjacent sliding windows according to the change trend of the data within the sliding window; combine the numerical sizes of the data within the sliding window to obtain the change complexity of each component signal; obtain the attenuation possibility of the time domain data according to the change complexity and frequency of each component signal.

[0073] It should be noted that the DAC cable has stringent performance requirements in high-frequency scenarios. Faults may lead to signal amplitude attenuation, excessive bit error rate, or system communication interruption, especially in scenarios such as data centers and 5G base stations, which will trigger cascading risks. Through fully automated signal integrity (SI) high-frequency parameter testing, combined with frequency-domain, time-domain, and dynamic performance analysis, the physical defects and electrical performance deviations of the cable can be accurately located.

[0074] Furthermore, it should be noted that when testing DAC high-speed cables, the time-domain waveform and eye diagram play important roles together. The time-domain waveform provides detailed information about the change of signal amplitude over time, enabling the analysis of the rise time, fall time, and its stability of the signal. The eye diagram, on the other hand, intuitively shows the quality characteristics of the signal, such as the opening height and opening width, by superimposing time-domain waveforms of multiple cycles. The opening height reflects the noise tolerance ability of the signal, while the opening width indicates the time jitter of the signal.

[0075] Furthermore, it should be noted that the rise time and fall time of the time-domain waveform reflect the time required for the signal to jump from a low level to a high level and from a high level back to a low level. Signal attenuation usually manifests as a gradual decrease in the amplitude of the time-domain waveform; as the signal attenuates, not only will the amplitude of the time-domain waveform decrease, but it may also become distorted; and it will lead to an increase in the jitter of the time-domain waveform; while the eye diagram is a graph formed by superimposing multiple repeated signal waveforms, which can intuitively reflect the quality of the signal, especially characteristics such as attenuation and jitter during transmission; however, due to electromagnetic coupling between wire pairs causing near-end crosstalk (NEXT) and far-end crosstalk (FEXT), it will cause random fluctuations in signal amplitude and an increase in bit error rate; the fluctuating signal acts as noise, and when the signal attenuates, the influence of the noise will relatively increase, manifested as an increase in the noise floor in the eye diagram. As the signal amplitude decreases, the relative intensity of noise and interference will increase, resulting in a more blurred eye opening in the eye diagram, which means a decline in signal quality and an increase in bit error rate. Based on this, by analyzing the change relationship between time-domain data and the eye diagram, the faults of the DAC high-speed cable can be obtained to judge its quality level.

[0076] Furthermore, it should be noted that in the time-domain signal, the greater the degree of fluctuation of the received signal, the greater the interference generated by the signal when the DAC high-speed cable is transmitting. Therefore, the time-domain signal is first described to obtain the degree of interference of the signal; and through empirical mode decomposition (EMD) of the time-domain data, the component signals obtained can represent the change complexity of the original signal. If the change complexity of the original signal is higher, the number of component signals obtained by decomposition will be more, and the degree of signal change will be higher.

[0077] Specifically, perform EMD decomposition on the time-domain data to obtain multiple component signals; obtain the degree of change of the component signals, and its calculation formula is as follows:

[0078]

[0079] In the formula, p represents the degree of change of the component signal, N represents the number of component signals, and s l represents the variance of all amplitudes in the l-th component signal, and sigmoid() represents the sigmoid function, which is used for normalization processing in this embodiment.

[0080] It should be noted that s l reflects the fluctuation situation through the variance of the amplitude of the component signal, and further characterizes the degree of change. The larger its value, the greater the degree of change of the component signal. Because in the EMD algorithm, the decomposition of the signal is based on frequency. However, if the degree of change of a single component signal is larger, it means that the original signal is more affected, and the corresponding degree of change of the original signal will also be larger. Here, the degree of change p of the component signal reflects the overall degree of change of all component signals.

[0081] It should be further noted that the degree of change of the above-mentioned component signal describes the overall characteristics, and then analyzes a single component signal; because when testing the DAC high-speed cable, an electrical signal is transmitted to the cable, and then the fluctuation changes of the echo signal are monitored to judge the quality of the cable. If signal attenuation occurs, it is manifested as conductor loss and dielectric loss of the cable, and the attenuation of high-frequency components is more significant, resulting in an overall decrease in the signal amplitude; therefore, analyze the change of a single component signal.

[0082] It should be further noted that a single component signal represents the detailed changes of the original time-domain signal. If there is attenuation in the original signal, it will show that the high-frequency component signal attenuates; in order to capture the changes of the signal at different times, it is necessary to set a sliding window for analysis.

[0083] Specifically, preset the initial length of the sliding window. In this embodiment, the initial length is described as A = 5; then adjust the length of the sliding window according to the degree of change of the component signal, and its calculation formula is as follows:

[0084]

[0085] In the formula, T represents the length of the sliding window, p represents the degree of change of the component signal, A represents the preset length, represents the ceiling function.

[0086] Further, several sliding windows are obtained for each component signal based on the length of the sliding window. The step size of the sliding window is equal to its length, so there is no duplicate data between different segments of the sliding window. The data within any segment of the sliding window is subjected to least squares fitting to obtain a fitting curve. Least squares fitting is a well-known prior art and will not be elaborated here. The principal component direction of the fitting curve is obtained through the PCA algorithm and used as the data distribution direction of the sliding window. Then, the calculation formula for the directional anisotropy of adjacent sliding windows is as follows:

[0087] q l(i,i+1) = θ li × θ l(i+1)

[0088] In the formula, q l(i,i+1) represents the directional anisotropy between the i-th sliding window and the (i + 1)-th sliding window of the l-th component signal, θ li represents the data distribution direction of the i-th sliding window of the l-th component signal, and θ l(i+1) represents the data distribution direction of the (i + 1)-th sliding window of the l-th component signal.

[0089] Further, based on the directional anisotropy of adjacent sliding windows and combined with the numerical magnitude of the data within the sliding window, the change complexity of each component signal is obtained, and its calculation formula is as follows:

[0090]

[0091] In the formula, V l represents the change complexity of the l-th component signal, represents the maximum value of the data mean in all sliding windows of the l-th component signal, represents the minimum value of the data mean in all sliding windows of the l-th component signal, n represents the number of sliding windows in the component signal, and q l(i,i+1) represents the directional anisotropy between the i-th sliding window and the (i + 1)-th sliding window of the l-th component signal. exp[] represents the exponential function with the natural constant as the base. In this embodiment, the exp[-x] model is used to present the inverse proportional relationship and normalization process. x is the input of the model, and the implementer can set the inverse proportional function and normalization function according to the actual situation.

[0092] It should be noted that represents the amplitude change of the l-th component signal. The greater the degree of difference, the greater the degree of change that occurs during the transmission of the time-domain signal. Therefore, the possible signal attenuation is more severe. The directional anisotropy of the sliding window represents the change in signal amplitude. The greater the directional anisotropy, it also means that the data distribution direction trends of consecutive sliding windows are similar, indicating that the signal presents a state of continuous attenuation or growth. Here, q l(i,i+1)The value can be positive or negative. When it is negative, it indicates that the signal shows a decaying trend.

[0093] Furthermore, based on the change complexity and frequency of each component signal, the decay possibility of the time-domain data is obtained. The calculation formula is as follows:

[0094]

[0095] In the formula, w represents the decay possibility of the time-domain data, f l represents the frequency of the l-th component signal, V l represents the change complexity of the l-th component signal, N represents the number of component signals, and sigmoid() represents the sigmoid function, which is used for normalization processing in this embodiment.

[0096] It should be noted that when signal decay occurs, the high-frequency components decay more significantly. Therefore, the frequency of the component signal is used as a reference weight here. If the signal change complexity of the high-frequency signal is greater, it means that the possibility of signal decay of the high-frequency signal is greater, and thus the possibility of signal decay of the original test data is greater.

[0097] So far, the decay possibility of the time-domain data has been obtained.

[0098] Step S003: According to the decay possibility of the time-domain data, as well as the opening height and opening width of the eye diagram data, obtain the variability of the time-domain data; according to the variability of the time-domain data and the distribution characteristics of the eye diagram data, obtain the decay degree of the time-domain data.

[0099] It should be noted that according to the above analysis, the possibility of signal decay is obtained. However, when testing the DAC high-speed cable, environmental noise, ground interference, crosstalk between signal lines, etc. will all cause the noise level to rise, resulting in a decrease in the time stability of the signal, manifested as irregularity of the signal edges in the time-domain waveform, showing random fluctuations or spikes, which affect the clarity of the signal. Then, it is inaccurate to judge signal decay only based on the change of the time-domain waveform; and the eye diagram is a very effective tool for analyzing the integrity and quality of high-speed digital signals, especially in evaluating whether the signal has decayed; by superimposing multiple cycles of the signal, the eye diagram can intuitively display the high and low levels, opening height and width of the signal, and can display multiple signal characteristics at the same time, and observe the influence of decay on multiple parameters simultaneously;

[0100] Further, it should be noted that the time-domain waveform is a graph representing the variation of signal amplitude over time, and the eye diagram is a graph generated by superimposing multiple cycles of the time-domain waveform. The eye diagram overlaps the signal waveforms that repeat multiple times, with time as the horizontal axis and amplitude as the vertical axis. The characteristics of the time-domain waveform (such as rise time, fall time, amplitude, noise) directly affect the shape and characteristics of the eye diagram. For example, the faster the rise and fall speeds of the time-domain waveform, the larger the opening of the eye diagram usually is.

[0101] Specifically, obtain the opening height H and opening width E of the eye diagram; then, in combination with the attenuation possibility of the time-domain data, obtain the variability of the time-domain data, and its calculation formula is as follows:

[0102]

[0103] In the formula, K represents the variability of the time-domain data, w represents the attenuation possibility of the time-domain data, H represents the opening height of the eye diagram, and E represents the opening width of the eye diagram.

[0104] It should be noted that the greater the attenuation possibility of the time-domain data, the greater the abnormal change of the signal in the time domain; and the decrease of the opening height and the opening width indicates that the amplitude of the signal decreases, which is caused by attenuation. Therefore, the greater the variability of the time-domain data.

[0105] Further, it should be noted that when there is no attenuation in the signal, the opening of the eye diagram should be symmetric and the shape should be regular; if there is attenuation in the signal, the opening of the eye diagram will be asymmetric and tilt to one side; and the opening of the eye diagram may become blurred or even completely closed. Based on the above characteristics, obtain the attenuation degree of the time-domain data.

[0106] Specifically, as Figure 4 shown, first obtain the upper vertex and lower vertex of the eye diagram, as well as two intermediate intersection points; with the line connecting the upper vertex and the lower fixed point as the axis of symmetry, fold the eye diagram in half to obtain the spatial distance d between the two intermediate intersection points; then, use the template matching algorithm to match the eye diagrams on both sides of the axis of symmetry to obtain the matching degree g of the eye diagrams on both sides of the axis of symmetry; then, the calculation formula for obtaining the attenuation degree of the time-domain data is as follows:

[0107]

[0108] In the formula, γ represents the attenuation degree of the time-domain data, K represents the variability of the time-domain data, d represents the spatial distance between the two intermediate intersection points, g represents the matching degree of the eye diagrams on both sides of the axis of symmetry, and sigmoid() represents the sigmoid function, which is used for normalization processing in this embodiment.

[0109] So far, the attenuation degree of the time-domain data has been obtained.

[0110] Step S004: Screen out abnormal data based on the attenuation degree of the time-domain data, and detect the faults of the DAC high-speed cable.

[0111] It should be noted that after obtaining the attenuation degree of the time-domain data, it is further determined whether there is a signal attenuation fault when testing the DAC high-speed cable according to the attenuation degree of the signal.

[0112] Specifically, a fault threshold is preset. In this embodiment, the fault threshold is described using 0.6. If the attenuation degree of the time-domain data is greater than the fault threshold, it indicates that there is a large degree of attenuation in the test data at this time, and the main reason for the signal attenuation is electromagnetic interference. Therefore, it is necessary to promptly eliminate the signal interference to avoid signal loss during signal transmission; then the test data is abnormal and the DAC high-speed cable has a fault.

[0113] Through the above steps, a fault detection method for a DAC high-speed cable is completed.

[0114] Another embodiment of the present invention provides a fault detection system for a DAC high-speed cable, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a fault detection method for a DAC high-speed cable in steps S001 to S004.

[0115] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A fault detection method for a DAC high-speed cable, characterized in that, The method includes the following steps: Obtain the test data of the DAC high-speed cable, where the test data includes time-domain data and eye diagram data; Decompose the time-domain data to obtain multiple component signals, obtain the degree of change of the component signals according to the fluctuations of each component signal; adjust to obtain a sliding window according to the degree of change of the component signals; obtain the direction anisotropy of adjacent sliding windows according to the change trend of the data within the sliding window; combine the numerical magnitudes of the data within the sliding window to obtain the change complexity of each component signal; obtain the attenuation possibility of the time-domain data according to the change complexity and frequency of each component signal; Obtain the variability of the time-domain data according to the attenuation possibility of the time-domain data, as well as the opening height and opening width of the eye diagram data; obtain the attenuation degree of the time-domain data according to the variability of the time-domain data and the distribution characteristics of the eye diagram data; Screen out abnormal data according to the attenuation degree of the time-domain data and detect the faults of the DAC high-speed cable.

2. The fault detection method of a DAC high-speed cable according to claim 1, characterized in that The specific steps included in decomposing the time-domain data to obtain multiple component signals and obtaining the degree of change of the component signals according to the fluctuations of each component signal are as follows: Perform EMD decomposition on the time-domain data to obtain multiple component signals; obtain the degree of change of the component signals, and its calculation formula is as follows: where p represents the degree of change of the component signal, N represents the number of component signals, s l represents the variance of all amplitudes in the l-th component signal, and sigmoid() represents the sigmoid function.

3. The fault detection method for a DAC high-speed cable according to claim 1, characterized in that The specific steps included in adjusting to obtain a sliding window according to the degree of change of the component signals are as follows: Preset the initial length of the sliding window, and adjust the length of the sliding window according to the degree of change of the component signals, and its calculation formula is as follows: Wherein, T represents the length of the sliding window, p represents the degree of change of the component signal, and A represents the preset length, represents the ceiling function.

4. A fault detection method for a DAC high-speed cable according to claim 1, characterized in that The specific steps included in obtaining the direction anisotropy of adjacent sliding windows according to the change trend of the data within the sliding window are as follows: Based on the length of the sliding window, obtain several segments of sliding windows for each component signal. The step size of the sliding window is equal to the length, so there is no repeated data between different segments of sliding windows; perform least squares fitting on the data within any segment of the sliding window to obtain a fitting curve; obtain the principal component direction of the fitting curve through the PCA algorithm and use it as the data distribution direction of the sliding window; The calculation formula for obtaining the direction anisotropy of adjacent sliding windows is as follows: q l(i,i+1) = θ li × θ l(i+1) where q l(i,i+1) represents the anisotropy between the i-th sliding window and the (i + 1)-th sliding window of the l-th component signal, and θ li represents the data distribution direction of the i-th sliding window of the l-th component signal, and θ l(i+1) represents the data distribution direction of the (i + 1)-th sliding window of the l-th component signal.

5. The fault detection method of a DAC high-speed cable according to claim 1, characterized in that, The specific steps included in obtaining the change complexity of each component signal are as follows: where, V l represents the variation complexity of the l-th component signal, represents the maximum value of the data means in all sliding windows of the l-th component signal, represents the minimum value of the data means in all sliding windows of the l-th component signal, n represents the number of sliding windows in the component signal, q l(i,i+1) represents the direction anisotropy between the i-th sliding window and the (i + 1)-th sliding window of the l-th component signal, and exp[] represents the exponential function with the natural constant as the base.

6. The fault detection method for a DAC high-speed cable according to claim 1, wherein, The specific steps included in obtaining the attenuation possibility of the time-domain data according to the change complexity and frequency of each component signal are as follows: where w represents the attenuation possibility of time-domain data, and f l represents the frequency of the l-th component signal, V l represents the change complexity of the l-th component signal, N represents the number of component signals, and sigmoid() represents the sigmoid function.

7. A fault detection method for a DAC high-speed cable according to claim 1, characterized in that, The specific steps included in obtaining the variability of the time-domain data according to the attenuation possibility of the time-domain data, as well as the opening height and opening width of the eye diagram data are as follows: In the formula, K represents the variability of the time-domain data, w represents the attenuation possibility of the time-domain data, H represents the opening height of the eye diagram, and E represents the opening width of the eye diagram.

8. A fault detection method for a DAC high-speed cable according to claim 1, characterized in that, The specific steps included in obtaining the attenuation degree of the time-domain data according to the variability of the time-domain data and the distribution characteristics of the eye diagram data are as follows: Obtain the upper vertex and lower vertex of the eye diagram, as well as two intermediate intersection points; fold the eye diagram in half with the line connecting the upper vertex and the lower vertex as the axis of symmetry to obtain the spatial distance between the two intermediate intersection points, and perform template matching on the eye diagrams on both sides of the axis of symmetry through the template matching algorithm to obtain the matching degree of the eye diagrams on both sides of the axis of symmetry; The calculation method for obtaining the attenuation degree of the time-domain data is as follows: Wherein, γ represents the attenuation degree of the time-domain data, K represents the variability of the time-domain data, d represents the spatial distance between two intermediate crossover points, g represents the matching degree of the eye diagrams on both sides of the symmetry axis, and sigmoid() represents the sigmoid function.

9. The fault detection method for a DAC high-speed cable according to claim 1, characterized in that The abnormal data screened according to the attenuation degree of the time-domain data includes the following specific steps: Preset a fault threshold. If the attenuation degree of the time-domain data is greater than the fault threshold, the test data is abnormal, indicating that there is a fault in the DAC high-speed cable.

10. A fault detection system for a DAC high-speed cable, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the computer program is executed by a processor, it implements the steps of a fault detection method for a DAC high-speed cable as described in any one of claims 1-9.

Images