A cable connection reliability test method for a locomotive brake system

Abstract

The application relates to the technical field of electrical testing, and discloses a cable connection reliability testing method for a locomotive braking system, which comprises the following steps: S1, reference modeling: performing diagnostic grouping on a connection point, collecting an initial electrical fingerprint of the connection point, and establishing a health reference database; S2, excitation and monitoring: applying dynamic excitation to a cable to be tested, synchronously collecting a transient resistance signal in real time, and calculating a resistance fluctuation spectrum; S3, fault diagnosis: performing correlation analysis on the resistance fluctuation spectrum and dynamic excitation characteristics, combining the reference database, and diagnosing potential connection defects; and S4, evaluation and reporting: performing quantitative evaluation on the health state of the connection point according to the diagnosis result, and finally generating a comprehensive diagnosis report. The application adopts electrical response monitoring and health index evaluation technology under dynamic excitation, accurately identifies intermittent defects, and realizes quantitative evaluation and prediction of the cable connection reliability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electrical testing, in particular to a cable connection reliability test method for a locomotive braking system. BACKGROUND

[0002] The locomotive braking system is one of the core systems to ensure the safety of train operation, and its reliability requirement is extremely strict. As the "nerve network" of signal transmission and power distribution within the braking system, the reliability of the connection points of the cable assembly is directly related to the stable operation of the entire system. However, the locomotive is subjected to severe mechanical vibration, impact and large amplitude temperature periodic change during operation, which poses a severe challenge to the long-term reliability of the cable connection points in such harsh service environment. Therefore, developing a test method that can accurately and prospectively evaluate the reliability of the cable connection is of great significance to ensure the safety of the locomotive braking system.

[0003] In the field of electrical testing, the existing technical solutions for cable assemblies mainly focus on verifying their basic electrical continuity and insulation performance. This usually includes using a multimeter or micro-ohmmeter to measure the loop resistance of the connection points to confirm their continuity; and using an insulation resistance tester to measure the insulation performance between different conductors to prevent short circuits. These testing methods have been widely used in production and routine maintenance, aiming to ensure that the cable assemblies meet the basic electrical design specifications after leaving the factory or maintenance.

[0004] Although the existing technology can ensure the basic electrical functions of the cable assembly to a certain extent, there are still some deficiencies: the existing test method is essentially a static test, and the test condition is that the connection points are in a static and stress-free state. The reason why this detection method has limitations is that a large number of connection reliability problems, such as fretting corrosion, material fatigue of crimping points, and virtual welding cracks of welding points, do not exhibit obvious resistance abnormalities in a static state, and their defect characteristics are only exposed instantaneously when subjected to specific mechanical vibration or thermal stress excitation. Therefore, static testing cannot effectively detect such intermittent and hidden defects related to dynamic stress. In addition, the results of such tests are usually binary determinations of "pass" and "fail", which lack the ability to quantitatively evaluate the health status of the connection points. Even if the resistance of a connection point has deteriorated several times from the initial value, as long as it is still within the tolerance range, it will be judged as qualified. This method cannot reveal the degradation trend of the connection points, and therefore cannot provide data support for predictive maintenance. Finally, when detecting resistance out-of-tolerance faults, static testing cannot provide diagnostic information about the root cause of the fault, making it difficult to distinguish whether the fault is caused by corrosion, loosening or other factors, increasing the difficulty and time cost of fault troubleshooting. SUMMARY

[0005] The purpose of the present application is to provide a cable connection reliability test method for a locomotive braking system, which solves the problem that the prior art uses a static test method, which cannot effectively detect intermittent and initial degradation connection defects that are exposed only under dynamic stress, and lacks the ability to quantitatively evaluate the health status of the connection points and to perform in-depth diagnosis of the fault source.

[0006] To achieve the above purpose, the present application is implemented by the following technical solutions: a cable connection reliability test method for a locomotive braking system, the method comprising the following steps:

[0007] S1, reference modeling step: first, a plurality of connection points on the cable assembly to be tested are diagnosed and grouped according to predetermined rules such as function or physical location. Then, under static and no external excitation conditions, the initial electrical fingerprints of the plurality of connection points are collected. Specifically, the collection process includes:

[0008] The reference loop resistance values of each connection point are measured by the Kelvin four-wire method.

[0009] The loop resistance of each connection point is continuously collected for a short time at a predetermined high sampling frequency to obtain its time domain signal, and the statistical characteristics of the signal constitute the resistance fluctuation base, which is used to represent the inherent electrical noise level of the connection point.

[0010] For the connection points of the diagnostic grouping of communication cables, the time domain reflection (TDR) technology is used to apply a pulse signal and measure its reflection waveform, thereby obtaining the reference time domain reflection impedance waveform along the cable distance distribution.

[0011] After the collection is completed, the initial electrical fingerprints such as the reference loop resistance, resistance fluctuation base and reference time domain reflection impedance waveform are digitally processed, and a corresponding data record is created for each connection point in the database. The digitized fingerprint data is stored in association with the physical identification and diagnostic grouping identification of the connection point, thereby establishing a health reference database.

[0012] S2, dynamic excitation and response monitoring step: the cable assembly to be tested is placed in a composite environmental test chamber, and dynamic excitation is applied thereto. The dynamic excitation can be a wideband sweep vibration excitation covering the common frequency range of locomotive operation, or a temperature cycle excitation following a predetermined temperature change curve, or a combination of the two. During the whole process of applying dynamic excitation, the resistance of each connection point is continuously and high-speed sampled by a test instrument at a predetermined sampling frequency to obtain its instantaneous resistance signal in the time domain. Based on the signal, the following processing is performed to obtain the resistance fluctuation spectrum:

[0013] From the collected instantaneous resistance signal, the short-term moving average of the signal is subtracted to obtain a zero-mean resistance fluctuation signal.

[0014] The resistance fluctuation signal is subjected to fast Fourier transform, and its power spectrum density is calculated, which is the resistance fluctuation spectrum. Its calculation formula is:

[0015]

[0016] In the formula, R is the resistance fluctuation spectrum; ΔR is the fluctuation amount of the instantaneous resistance signal; T is the signal sampling time window; f is the frequency; j is the imaginary unit; t is the time; e is the base of natural logarithm; ∫ is the definite integral operator, indicating integration of the time variable in the interval from t1 to t2; π is the circular constant; t1 is the lower limit of the integration interval; t2 is the upper limit of the integration interval; and dt is the differential of the time t.

[0017] In an optional technical solution, the step further includes a closed-loop control process: when the broadband excitation is applied, the resistance fluctuation spectrum of each connection point is monitored in real time. When the energy of the resistance fluctuation spectrum of a connection point at a specific frequency point is found to exceed a preset threshold, the system identifies the frequency point as a potential mechanical resonance frequency of the connection point. Subsequently, the system automatically adjusts the output of the excitation source, stops the broadband scanning, and instead performs focused resonance dwelling or high-precision frequency sweeping in the narrowband range around the potential mechanical resonance frequency point, to apply stress amplification excitation.

[0018] S3. Fault diagnosis step: the step processes and analyzes the monitored data to diagnose potential connection defects. The step includes: first, performing correlation analysis. The analysis includes at least one of the following modes:

[0019] The frequency at which a significant peak value appears in the resistance fluctuation spectrum is compared with the vibration excitation frequency as the dynamic excitation. If the frequency values of the two are highly consistent or there is a frequency multiplication relationship, it is judged that the connection point has a mechanical connection defect caused by vibration.

[0020] ​​​​​​​​​​​​​​​​For two connection points in different diagnostic groups (defined as the first connection point and the second connection point), the resistance fluctuation signals thereof are acquired respectively, and a coherence function between the two signals is calculated. By judging whether the value of the coherence function at a specific frequency (such as a potential mechanical resonance frequency) tends to 1, it is determined whether there is a fault propagation path between the two connection points caused by vibration transmission. The formula for calculating the coherence function is:

[0021] ;

[0022] wherein, is the coherence function; is the auto-power spectral density of the resistance fluctuation signal of the first connection point; is the auto-power spectral density of the resistance fluctuation signal of the second connection point; is the cross-power spectral density of the resistance fluctuation signals of the first and second connection points; is the frequency.

[0023] Secondly, the real-time monitored data is compared with the health benchmark database established in step S1 for diagnosis. The diagnosis includes at least one of the following ways:

[0024] The average value of the instantaneous resistance signal measured during the dynamic test is compared with the benchmark loop resistance of the connection point stored in the database. If a significant and unrecoverable numerical increase occurs, it is identified that there is an irreversible resistance drift defect.

[0025] The total energy or peak energy of the resistance fluctuation spectrum calculated during the dynamic test is compared with the energy of the resistance fluctuation base of the connection point stored in the database. If the former is significantly higher than the latter, it is identified that the instability of the connection is increased.

[0026] The time-domain reflected impedance waveform acquired in real time during the dynamic test is compared with the benchmark time-domain reflected impedance waveform of the connection point stored in the database. By analyzing the difference waveform between the two, the impedance discontinuity defect on the connection point or cable body is identified and located.

[0027] S4. Evaluation and reporting step: After the above diagnosis is completed, this step makes a final quantitative evaluation of the health status of the connection points. This evaluation is achieved by calculating a comprehensive connection health index (CHI). The calculation of this index combines two types of features: the first is the data-driven feature directly extracted from the test results, such as the normalized resistance fluctuation frequency spectrum peak energy, irreversible resistance drift amount, etc.; the second is the physical model inference feature calculated based on the pre-set physical failure model (such as material fatigue S-N curve or fretting wear model), such as cumulative fatigue damage degree or equivalent wear accumulation. By weighted sum of the two types of normalized features, the final connection health index is obtained.

[0028] Secondly, a comprehensive diagnostic report is generated. The generation process specifically includes the following operations:

[0029] Data association and formatting: structurally associate the unique identifier of each connection point with its calculated connection health index, the list of potential mechanical resonance frequencies identified in the fault diagnosis step, and the specific defect type diagnosed at the data level.

[0030] Visualization topology generation: based on the coherence function analysis results between the connection point pairs calculated in the fault diagnosis step, a visualization topology graph representing the fault propagation path is generated. In this topology graph, nodes are used to represent each connection point, and when the coherence function value between two nodes corresponding to the connection points exceeds the pre-set judgment threshold, a connection line is generated between the two nodes, which represents the existence of a fault propagation path between them.

[0031] Report file synthesis: integrate the data associated and formatted above, and the generated visualization topology graph, into a single, structured data file or document to form the comprehensive diagnostic report.

[0032] In summary, the present application includes at least one of the following beneficial technical effects:

[0033] 1. The present application applies a vibration and temperature compound dynamic excitation simulating the real working environment of the cable assembly to be tested, and acquires high-frequency instantaneous resistance and other electrical parameters during this process, and then processes the dynamic response through spectrum analysis and other means. This technical means can effectively excite and capture intermittent and initial degradation connection defects that cannot be detected by traditional static test methods, thereby identifying potential connection reliability risks at an extremely early stage of failure, improving the depth and predictability of the test.

[0034] 2. This invention establishes a health benchmark database and weights and fuses multiple features extracted from dynamic testing, such as resistance drift and fluctuation energy, to calculate a connection health index that characterizes the health status of connection points. This method transforms the previous binary judgment of "qualified / unqualified" into a precise and traceable quantitative assessment, providing objective data support for cable assembly condition monitoring, lifespan prediction, and the formulation of predictive maintenance strategies.

[0035] 3. This invention, by analyzing the correlation between the resistance fluctuation spectrum and dynamic excitation characteristics, can infer the connection between defects and physical causes such as mechanical resonance; by analyzing the signal coherence between connection points, a fault propagation path topology map can be constructed. This comprehensive diagnostic capability enables technicians to quickly understand fault modes and accurately locate defect positions, improving the efficiency of fault diagnosis and repair. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0037] Figure 2 This is a flowchart illustrating the baseline modeling steps of the present invention;

[0038] Figure 3 This is a flowchart illustrating the dynamic excitation and response monitoring steps of the present invention.

[0039] Figure 4 This is a flowchart illustrating the fault diagnosis steps of the present invention.

[0040] Figure 5 This is a flowchart illustrating the evaluation and reporting steps of this invention. Detailed Implementation

[0041] The following is in conjunction with the appendix Figure 1 -Appendix Figure 5 The present invention will be further described in detail below.

[0042] like Figure 1 As shown, Figure 1 This is a flowchart illustrating a cable connection reliability testing method for a locomotive braking system according to an embodiment of the present invention. The present invention provides a cable connection reliability testing method for a locomotive braking system, which may include the following steps:

[0043] S1. Benchmark modeling steps: Diagnosely group multiple connection points of the cable assembly under test, collect initial electrical fingerprints of multiple connection points, and establish a health benchmark database containing the initial electrical fingerprints.

[0044] S2, dynamic excitation and response monitoring step: dynamic excitation is applied to the cable assembly under test, and in the process, the instantaneous resistance signals of multiple connection points are collected in real time, and the resistance fluctuation spectrum representing the frequency domain distribution of the fluctuation energy is calculated based on the instantaneous resistance signals;

[0045] S3, fault diagnosis step: correlation analysis is performed between the resistance fluctuation spectrum and the characteristics of the dynamic excitation, and the health benchmark database is combined to diagnose whether there is a potential connection defect in the multiple connection points;

[0046] S4, evaluation and reporting step: based on the results of the fault diagnosis step, the health status of the multiple connection points is quantitatively evaluated, and a comprehensive diagnostic report is generated.

[0047] In this embodiment, the test system for performing the above method, the composition includes the object under test, the dynamic excitation system, the response monitoring system and the central processing and control unit.

[0048] The object under test is a set of locomotive brake system cable assembly. The assembly includes power cables that provide power for brake control units, signal cables that transmit control commands, and data buses for on-board network communication. These cables are interconnected through multiple connectors, including D-Sub series connectors and M12 series circular connectors. The "connection points" in the test method refer to the paired contact points of the conductive pins and sockets in these connectors, or the crimp points of the wires and connector terminals.

[0049] The dynamic excitation system is used to apply controlled mechanical and environmental stress to the object under test. The system includes a six-degree-of-freedom vibration table and a high-low temperature and humidity combined environmental test chamber. The vibration table is used to apply vibration excitation covering a frequency range of 5Hz to 2000Hz. The environmental test chamber is used to apply temperature cycle excitation following a pre-set temperature curve, with a temperature range of -40°C to +125°C. The cable assembly under test is mounted on the table surface of the vibration table and placed inside the environmental test chamber.

[0050] The response monitoring system is used to collect the electrical parameter responses of the object under test during the test. The system includes a multi-channel high-precision data acquisition (DAQ) system, a Kelvin four-wire test fixture corresponding to each connection point, and a time domain reflectometer (TDR) module integrated into the data acquisition system. The sampling resolution of the data acquisition system is 24 bits, and the maximum synchronous sampling frequency of each channel is 100kS / s.

[0051] The central processing and control unit is an industrial computer running test control and data analysis software. This unit communicates with the controller of the dynamic excitation system via GPIB or Ethernet interface to precisely control the application of vibration and temperature excitation. Simultaneously, this unit connects to the response monitoring system via PCIe or USB interface for configuring acquisition parameters, starting or stopping data acquisition, and receiving the acquired raw data.

[0052] During the execution of the method, the various systems work collaboratively. In step S1, the central processing and control unit (CPU) controls the response monitoring system to acquire initial electrical fingerprints at each connection point without dynamic excitation. In step S2, the CPU controls the dynamic excitation system to apply excitation based on a preset test profile, and simultaneously instructs the response monitoring system to acquire real-time, continuous instantaneous resistance signals at each connection point. All acquired data is transmitted to the CPU, where its internal software executes all computational tasks defined in steps S3 and S4, including data processing, analysis, diagnosis, and report generation.

[0053] like Figure 2 As shown, Figure 2 This is a flowchart illustrating the baseline modeling step S1 according to an embodiment of the present invention. This step is performed before the test begins, when the object under test is in a static, unexcited state, and is controlled by the central processing and control unit. Specifically, it includes the following operations.

[0054] First, diagnostic grouping is performed on multiple connection points in the cable assembly under test. This grouping operation is performed in the software of the central processing and control unit, based on a pre-loaded cable assembly configuration file. For example, all connection points of the power cables used to power the brake control unit are grouped into the "Power Group".

[0055] All connection points used to transmit control logic level signals are grouped into a "signal group";

[0056] All connection points that are part of a data bus such as CAN or Ethernet are grouped into a "communication group".

[0057] Each connection point is assigned a unique connection point identifier and a diagnostic group identifier.

[0058] Subsequently, initial electrical fingerprints were collected for each connection point. The first fingerprint was the reference loop resistance. The central processing and control unit command response monitoring system was connected to each connection point via a Kelvin four-wire test fixture. A constant DC test current of 10mA was applied, and the voltage drop across the connection point was measured. The reference loop resistance value was then calculated using Ohm's law. This value is recorded and associated with the corresponding connection point identifier.

[0059] The second fingerprint is the resistance fluctuation floor. Under the condition that the Kelvin four-wire connection is maintained and no external dynamic excitation is applied, the central processing and control unit instructs the data acquisition system to continuously sample the loop resistance of each connection point at a sampling frequency of 50 kS / s for a period of 2 seconds, obtaining a high-resolution resistance time-domain signal. The root mean square (RMS) value of this time-domain signal is calculated, and this value is taken as the resistance fluctuation floor representing the inherent electrical noise level of the connection point.

[0060] The third fingerprint is the reference time-domain reflected impedance waveform. This acquisition is only performed for the connection points that are diagnostically grouped as "communication group". The central processing and control unit instructs the time-domain reflectometer (TDR) module to connect to the data bus connection point under test through the corresponding probe. The TDR module injects a step voltage pulse with a rise time of 50 ps at one end of the cable and synchronously acquires the reflected signal voltage returned within a preset time window (for example, 200 ns), thereby generating a reference time-domain reflected impedance waveform reflecting the impedance change along the cable distance .

[0061] Finally, the health reference database is established. The central processing and control unit creates a data table in its memory. The data table contains at least the following fields: connection point identifier, diagnostic grouping identifier, reference loop resistance, resistance fluctuation floor, reference time-domain reflected impedance waveform. The initial electrical fingerprint data of each connection point acquired and calculated in the previous steps are digitized (for example, floating point values, time-domain waveform data arrays) and filled into the record row associated with the corresponding connection point identifier in the data table. After the data filling of all connection points is completed, the health reference database for subsequent comparative analysis is established.

[0062] As shown in Figure 3 , the flowchart of the dynamic excitation and response monitoring step S2 according to an embodiment of the present application is shown in Figure 3 . This step is performed after the reference modeling is completed, and the dynamic excitation is applied to the object under test and the electrical response is synchronously acquired through the precise coordination of the central processing and control unit.

[0063] At the beginning of this step, the central processing and control unit sends control instructions to the dynamic excitation system according to the preset test specification. For example, it instructs the six-degree-of-freedom vibration table to apply a broadband random vibration, and the vibration power spectral density is set to a specific profile in the frequency range of 5 Hz to 2000 Hz.

[0064] At the same time, it instructs the high-low temperature and humidity combined environmental test chamber to start a temperature cycle, for example, from room temperature to -40°C at a rate of 5°C / min, keep at this low temperature point for 30 minutes, then rise to +125°C at the same rate, keep at this high temperature point for 30 minutes, to form a complete cycle.

[0065] During the whole process of dynamic excitation application, the central processing and control unit synchronously instructs the response monitoring system to start data acquisition. The data acquisition system continuously acquires the instantaneous resistance signals of all monitored connection points at a synchronous sampling frequency of 100 kS / s, and transmits the acquired raw digital signals to the central processing and control unit in the form of data blocks in real time.

[0066] After receiving each data block, the central processing and control unit immediately processes it to calculate the resistance fluctuation spectrum. First, from the raw instantaneous resistance signal, the moving average value of the signal in a short time window is calculated, and the moving average value is subtracted from the raw signal to obtain a zero-mean resistance fluctuation signal .

[0067] Subsequently, a window function (for example, a Hanning window) is applied to the data block of the resistance fluctuation signal to suppress spectral leakage, and then a fast Fourier transform is performed to calculate the power spectral density as the resistance fluctuation spectrum . The formula for calculating the power spectral density is:

[0068] ;

[0069] wherein is the resistance fluctuation spectrum; is the fluctuation amount of the instantaneous resistance signal; is the signal sampling time window; is the frequency; is the imaginary unit; is the time; is the base of the natural logarithm; is the definite integral operator, indicating integration with respect to the time variable in the interval from to ; is the ratio of the circumference to the diameter; is the lower limit of the integration interval; is the upper limit of the integration interval; is the differential with respect to time .

[0070] In one specific scheme of the embodiment, this step further includes a closed-loop control process. While calculating the resistance fluctuation spectrum in real time, the central processing and control unit continuously monitors the peak value of the spectrum. When the calculated power spectral density value at a certain frequency point exceeds a preset threshold (for example, 10 times the resistance fluctuation base energy measured in step S1), the central processing and control unit will adjust the frequency It was identified as a potential mechanical resonance frequency.

[0071] After identifying the potential mechanical resonance frequency, the central processing and control unit immediately issues new control commands to the dynamic excitation system, automatically adjusting the excitation application method. For example, it pauses the current broadband random vibration and instead instructs the vibration table to operate at that frequency. A constant-frequency sinusoidal vibration lasting 60 seconds is applied. This operation achieves focused stress amplification excitation at the weak frequency of the connection point, thereby accelerating the exposure of potential defects.

[0072] like Figure 4 As shown, Figure 4 This is a flowchart illustrating fault diagnosis step S3 according to an embodiment of the present invention. This step is automatically executed by the central processing and control unit after receiving dynamic response data, and its purpose is to identify and characterize potential connection defects by analyzing the data. This step is mainly achieved through two methods: correlation analysis and database comparison diagnosis.

[0073] Correlation analysis aims to establish causal relationships between dynamic excitation and electrical response;

[0074] The first correlation analysis method involves comparing the peak value of the spectrum with the excitation frequency; the central processing and control unit calculates the resistance fluctuation spectrum from step S2. In the process, the frequency corresponding to the highest energy peak is extracted. Simultaneously, it obtains the frequency of the currently applied vibration excitation from the controller of the dynamic excitation system. .when When the frequency tolerance is less than a preset frequency tolerance (e.g., 1Hz), the system determines that there is a mechanical connection defect at the connection point caused directly by the current vibration frequency, and records this diagnosis result along with the corresponding connection point identifier.

[0075] The second type of correlation analysis is coherence analysis between connection points. This analysis is used to determine whether a fault propagates through different connection points. The central processing and control unit selects a pair of connection points, for example, the first connection point. Second connection point And acquire their resistance fluctuation signals collected within the same time window. and Subsequently, the autopower spectral density of the two signals was calculated. and and cross-power spectral density Based on these spectral densities, the coherence function is calculated. The calculation formula is as follows:

[0076] ;

[0077] wherein, is the coherence function; is the auto-power spectral density of the resistance fluctuation signal at the first connection point; is the auto-power spectral density of the resistance fluctuation signal at the second connection point; is the cross-power spectral density of the resistance fluctuation signals at the first and second connection points; is the frequency.

[0078] If the value of the coherence function exceeds a high threshold value (e.g. 0.9) at a certain frequency of interest (e.g. an identified resonance frequency), it is determined that a fault propagation path exists between the connection points and .

[0079] Database comparison diagnosis aims at identifying the degradation of the electrical properties of the connection points. The central processing and control unit accesses the health reference database established in step S1 and compares the real-time monitoring data with the stored reference values.

[0080] The first comparison diagnosis is the identification of irreversible resistance drift. The central processing and control unit calculates the average value of the instantaneous resistance signal over a long time window (e.g. 10 seconds) . This value is compared with the reference loop resistance of the connection point stored in the database . If the percentage difference between continuously exceeds a preset threshold value (e.g. 5%), an irreversible resistance drift defect is identified.

[0081] The second comparison diagnosis is the identification of connection instability. The central processing and control unit obtains the total energy of the resistance fluctuation spectrum by integrating the real-time calculated resistance fluctuation spectrum . This energy is compared with the resistance fluctuation base energy of the connection point stored in the database . If the ratio of exceeds a preset multiple (e.g. 100), it is determined that the connection instability of the connection point has significantly increased.

[0082] The third comparison diagnosis is the identification and positioning of impedance discontinuity defects, which is only applicable to the connection points of the "communication group". The central processing and control unit instructs the TDR module to obtain the real-time time-domain reflection impedance waveform and retrieves the corresponding reference waveform from the database. By calculating the difference waveform , analyze whether there is a peak value exceeding the preset impedance threshold in the difference waveform. If there is, it is determined that an impedance discontinuity defect occurs, and according to the time point at which the peak value occurs and the known propagation rate of the signal in the cable, the physical position of the defect point from the test port is calculated.

[0083] As shown in Figure 5 , Figure 5 is a flowchart of the evaluation and reporting step S4 according to an embodiment of the present application. This step is executed by the central processing and control unit after the fault diagnosis is completed, aiming to quantitatively evaluate the health state of the connection point and generate a final comprehensive diagnostic report.

[0084] This step first performs quantitative evaluation by calculating a connection health index (CHI). The calculation process of the index is as follows: first, a set of features is extracted and quantified from the diagnostic results of step S3. The set of features includes data-driven features and physical model inferred features. Examples of data-driven features include: the normalized irreversible resistance drift amount, i.e. ; and the normalized resistance fluctuation energy ratio, i.e. .

[0085] An example of physical model inferred features is the cumulative fatigue damage degree. Its calculation is based on a simplified Miner damage accumulation rule. The central processing and control unit identifies the number of stress cycles according to the amplitude of the resistance fluctuation signal and the frequency of the vibration excitation, and finds the allowable fatigue life under the corresponding stress level according to the preset material S-N curve (stress-life curve). The calculation formula of the cumulative fatigue damage degree is as follows:

[0086] ;

[0087] In the formula, is the cumulative fatigue damage degree; is the actual number of stress cycles at the th stress level; is the allowable fatigue life at the th stress level.

[0088] All extracted features, including data-driven features and physical model inferred features, are normalized to the interval of 0 to 1 to obtain a set of normalized features . Then, the connection health index (CHI) is calculated by a weighted fusion algorithm, and the calculation formula is as follows:

[0089] ;

[0090] In the formula, To connect the health index, its value range is [0, 1]; For the first The normalized eigenvalues; For the first Preset weights corresponding to each feature; This represents the total number of features used in the calculation; For summation index.

[0091] After calculating the health index of all connection points, this step then generates a comprehensive diagnostic report. This generation process is automatically executed by the central processing and control unit and includes the following operations: First, data association and formatting. A structured data record is created for each connection point's unique identifier, containing its calculated CHI value, descriptions of all defects diagnosed in step S3, and all identified potential mechanical resonance frequencies.

[0092] Second, a visual topology map is generated. A graphics generation module within the central processing and control unit draws a fault propagation path topology map based on the coherence analysis results from step S3. In this map, each connection point is represented as a node, and the node's color is rendered according to its CHI value (e.g., green indicates health, red indicates danger). If the coherence function value between any two connection points exceeds a preset threshold at a certain frequency, a line is drawn between these two nodes to visually represent the fault propagation relationship between them.

[0093] Third, the report file is synthesized. The central processing and control unit embeds the aforementioned formatted structured data records (presented in tabular form) and the generated visual topology map (presented in image form) into a preset report template, ultimately synthesizing a single, easy-to-view comprehensive diagnostic report file in PDF format. This file is the final output of the method of this invention.

[0094] The above-mentioned technical solutions for generating comprehensive diagnostic reports are all existing technologies, and will not be elaborated further in this article.

[0095] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for testing the reliability of cable connections in a locomotive braking system, characterized in that, The method includes the following steps: S1. Benchmark modeling steps: Diagnosely group multiple connection points of the cable assembly under test, collect the initial electrical fingerprints of the multiple connection points, and establish a health benchmark database containing the initial electrical fingerprints. S2. Dynamic excitation and response monitoring steps: Apply dynamic excitation to the cable assembly under test, and during this process, collect the instantaneous resistance signals of the multiple connection points in real time, and calculate the resistance fluctuation spectrum that characterizes the frequency domain distribution of its fluctuation energy based on the instantaneous resistance signals. S3. Fault diagnosis steps: Perform correlation analysis between the resistance fluctuation spectrum and the characteristics of the dynamic excitation, and combine with the health benchmark database to diagnose whether there are potential connection defects in the multiple connection points; S4. Assessment and Reporting Steps: Based on the results of the fault diagnosis steps, the health status of the multiple connection points is quantitatively assessed, and a comprehensive diagnostic report is generated. The fault diagnosis step, which involves performing correlation analysis between the resistance fluctuation spectrum and the characteristics of the dynamic excitation, includes at least one of the following analysis methods: The peak frequency of the resistance fluctuation spectrum is compared with the vibration excitation frequency that serves as the dynamic excitation. If the two frequencies are the same, it is determined that there is a mechanical connection defect. For the first and second connection points in different diagnostic groups, calculate the coherence function between their respective resistance fluctuation signals, and based on the value of the coherence function, determine whether there is a fault propagation path between the first and second connection points. In the fault diagnosis step, the health benchmark database is used to diagnose whether there are potential connection defects among the multiple connection points, specifically including at least one of the following diagnostic methods: The instantaneous resistance signal is compared with the reference loop resistance in the health benchmark database to identify irreversible resistance drift defects; The energy of the resistance fluctuation spectrum is compared with the energy of the resistance fluctuation base in the health benchmark database to identify an increase in connection instability; The real-time acquired time-domain reflection impedance waveform is compared with the reference time-domain reflection impedance waveform in the health benchmark database to identify and locate impedance discontinuity defects. The assessment and reporting steps include: The health status of the multiple connection points is quantitatively assessed by calculating a connection health index, wherein the connection health index is obtained by weighted fusion of data-driven features extracted from test results and physical model inference features based on physical failure model inference. Based on the connection health index and the results of the fault diagnosis steps, a comprehensive diagnostic report is generated, which includes the connection health index of each connection point, a list of potential mechanical resonance frequencies, and a visualized topology map of the fault propagation path.

2. The method for testing the reliability of cable connections in a locomotive braking system according to claim 1, characterized in that, The benchmark modeling step includes the following steps: Collecting the initial electrical fingerprints of the multiple connection points. The reference loop resistance of the multiple connection points was measured using the Kelvin four-wire method. Without external excitation, the loop resistance of the multiple connection points is sampled for a short time at a preset high sampling frequency to obtain a resistance fluctuation substrate characterizing its inherent electrical noise level. For the connection points of the diagnostically grouped communication cables, the reference time-domain reflected impedance waveform along the cable distance is obtained using time-domain reflectometry.

3. The method for testing the reliability of cable connections in a locomotive braking system according to claim 2, characterized in that, The benchmark modeling step, which involves establishing a health benchmark database containing the initial electrical fingerprint, includes the following steps: The reference loop resistance, resistance fluctuation base, and reference time-domain reflection impedance waveforms of each connection point are digitally processed. Create a data record for each connection point in the database; The digitized initial electrical fingerprint is associated with and stored with the diagnostic group identifier corresponding to the connection point to form the health benchmark database.

4. The method for testing the reliability of cable connections in a locomotive braking system according to claim 3, characterized in that, The dynamic excitation and response monitoring step includes the step of applying dynamic excitation to the cable assembly under test, which includes: The cable assembly under test is placed in a composite environmental test chamber and subjected to at least one of the following: a broadband sweep frequency vibration excitation covering a preset frequency range, or a temperature cycling excitation following a preset temperature curve.

5. The method for testing the reliability of cable connections in a locomotive braking system according to claim 4, characterized in that, The dynamic stimulus and response monitoring step further includes: During the application of the dynamic excitation, the resistance fluctuation spectrum is monitored in real time. When the energy at a certain frequency point exceeds a preset threshold, that frequency point is identified as a potential mechanical resonance frequency. The dynamic excitation is automatically controlled to apply focused stress amplification excitation at or near the potential mechanical resonance frequency.

6. The method for testing the reliability of cable connections in a locomotive braking system according to claim 5, characterized in that, In the dynamic excitation and response monitoring step, the step of acquiring the instantaneous resistance signals of the multiple connection points in real time and calculating the resistance fluctuation spectrum characterizing the frequency domain distribution of its fluctuation energy based on the instantaneous resistance signals includes: The instantaneous resistance signal of the connection point is obtained by continuously sampling the connection point at a preset sampling frequency using a testing instrument. The resistance fluctuation signal is obtained by subtracting the short-time moving average value from the instantaneous resistance signal. ; For the resistance fluctuation signal Perform a Fast Fourier Transform to calculate the power spectral density, which is the spectrum of the resistance fluctuation. The calculation formula is: ； In the formula, The resistance fluctuation spectrum; The fluctuation amount of the instantaneous resistance signal; For signal sampling time window; For frequency; The imaginary unit; For time; is the base of the natural logarithm; The definite integral operator is used to represent the integral of time with respect to the time variable. arrive Integrate within the interval; Pi; This is the lower limit of the integration interval; This represents the upper limit of the integration interval; For time The differential.

7. The method for testing the reliability of cable connections in a locomotive braking system according to claim 1, characterized in that, The coherence function is calculated as follows: ； In the formula, The coherence function is defined as follows; The power spectral density of the resistance fluctuation signal at the first connection point; The power spectral density of the resistance fluctuation signal at the second connection point; The cross-power spectral density of the resistance fluctuation signals at the first connection point and the second connection point; For frequency.

Images