A multi-source data fusion-based electronic connector failure precursor identification method

Abstract

The application relates to the technical field of data processing, in particular to a multi-source data fusion-based electronic connector failure precursor identification method, which comprises the following steps: collecting the contact resistance sequence, vibration acceleration data and environmental temperature of the connector, and performing detrending and envelope processing; a vibration coupling deviation calculation model is constructed, the residual error between the actual resistance fluctuation and the theoretical prediction value is calculated based on the contact physics law; a failure precursor index is generated by combining the historical cumulative damage and the static resistance drift trend; and the failure precursor index is compared with the safety warning line to identify the failure precursor. According to the technical scheme, early wear characteristics can be accurately extracted from strong noise, and the maintenance window period is significantly prolonged.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method for identifying early signs of failure in electronic connectors based on multi-source data fusion. Background Technology

[0002] In applications such as high-voltage interlock systems for new energy vehicles, joints for industrial robots, and pitch control systems for wind power, electronic connectors are subjected to long-term high-frequency vibration. Under the combined stress of mechanical vibration and thermal expansion and contraction, invisible micron-level relative motion occurs at the connector contact interface, a phenomenon known as fretting wear. Fretting wear leads to gradual oxidation and corrosion of the contact interface, ultimately causing poor contact or even ablation.

[0003] Currently, mainstream connector health monitoring technologies mainly rely on the static threshold method, which monitors contact resistance in real time and triggers an alarm when the resistance exceeds a fixed threshold. However, under strong vibration conditions, the elastic deformation of the connector causes natural periodic fluctuations in resistance. This elastic noise can easily mask the subtle abnormal signals in the early stages of fretting wear, making it impossible for existing technologies to distinguish between normal fluctuations and abnormal wear.

[0004] Meanwhile, existing solutions lack logical integration of multi-dimensional data, easily monitoring resistance, temperature, and vibration in isolation, ignoring the coupling relationship between physical quantities. In reality, the resistance fluctuations of a healthy connector should follow changes in vibration energy. Once this following logic is broken, it signifies the appearance of a precursor to a fault. Furthermore, alarms from static threshold methods are often delayed. By the time the absolute resistance value exceeds the limit, physical damage is often irreversible, missing the optimal opportunity for preventative maintenance. Summary of the Invention

[0005] To address the technical problem that the weak signals in the early stages of fretting wear of electronic connectors under high-frequency vibration environments are easily masked by vibration noise and difficult to identify, this application provides a method for identifying early signs of failure in electronic connectors based on multi-source data fusion.

[0006] This application provides a method for identifying early signs of failure in electronic connectors based on multi-source data fusion, comprising: acquiring real-time contact resistance sequences at both ends of the connector, triaxial vibration acceleration data of the connector housing, and ambient temperature of the connector surface; preprocessing and time-aligning the contact resistance sequences and triaxial vibration acceleration data to separate high-frequency fluctuation components and vibration energy envelope signals; calculating the damping coupling deviation within the current time window using a preset damping coupling deviation calculation model based on the high-frequency fluctuation components, the vibration energy envelope signals, and the ambient temperature, wherein the damping coupling deviation is used to characterize the difference between the contact interface response to vibration and the healthy elasticity model; generating an early sign of failure index at the current moment based on the damping coupling deviation within a historical time period and the relative rate of change of static resistance at the current moment, combined with a time decay weight; and determining whether the early sign of failure index exceeds a preset safety warning line for multiple consecutive periods to determine whether the connector has a risk of early sign of failure.

[0007] By using high-frequency synchronous acquisition and signal decoupling, we can obtain the dynamic characteristics of vibration and resistance that are of the same source and phase, providing a high-quality data foundation for subsequent correlation analysis. At the same time, through the fusion and analysis of multi-source data throughout the entire process, we can extract weak failure fingerprints from strong vibration noise, significantly advancing the warning time.

[0008] In one embodiment, the preprocessing and time alignment of the contact resistance sequence and the triaxial vibration acceleration data includes: performing detrending processing on the real-time contact resistance sequence to filter out the DC component and retain the high-frequency fluctuation component; aligning the high-frequency fluctuation component with the timestamp of the vibration energy envelope signal through hardware triggering or software interpolation to make the phase synchronization error less than a preset threshold.

[0009] By strictly aligning and preprocessing the time domain, the phase difference between data sources is eliminated, ensuring a one-to-one correspondence between resistance fluctuations and vibration energy in physical time, thereby guaranteeing the rigor of the coupled analysis logic and the accuracy of the calculation results.

[0010] In one embodiment, the vibration damping coupling deviation satisfies the following relationship: ;in, For the first Vibration damping coupling deviation within a time window The number of data points within the sampling window. For the first in the window The resistance fluctuation amplitude at each moment. This is the nominal contact resistance value of the connector. To prevent dividing by zero from being the smallest positive number, The vibration damping sensitivity coefficient, For the first in the window The amplitude of vibrational energy at each moment. As a unit vibration energy standard, The logarithmic correction constant, This is a temperature correction function. For the first The ambient temperature at that moment.

[0011] By constructing a deviation model, normal elastic fluctuations caused by vibration can be removed from the actual signal as theoretical predictions, thereby accurately identifying abnormal wear signals that do not conform to vibration logic and effectively eliminating the interference of vibration noise.

[0012] In one embodiment, the method for acquiring the vibration energy envelope signal includes: performing vector synthesis on the collected triaxial vibration acceleration data to obtain a total vibration acceleration signal; performing envelope detection processing on the total vibration acceleration signal to extract the vibration energy envelope signal, which is used to reflect the trend of vibration energy change over time.

[0013] In one embodiment, the value of the temperature correction function increases as the difference between the ambient temperature and the normal temperature reference value increases; the temperature correction function includes a temperature coefficient, which is used to adjust the correction weight of the ambient temperature on the vibration damping coupling deviation, so as to reflect the accelerating effect of high temperature environment on oxide film instability.

[0014] In one embodiment, the failure precursor index satisfies the following relationship: ;in, For the current moment The failure precursor index, The length of the historical observation window. For the first time in history The vibration damping coupling deviation calculated at any given time. The time decay weighting function is... It is a time constant. It is a balance coefficient between static and dynamic characteristics. This represents the average static resistance at the current moment. The initial resistance fingerprint value for the connector during initial installation.

[0015] By introducing time weighting and static drift correction, the historical cumulative effect of fretting wear is considered, while also taking into account the slow drift trend of static resistance. This effectively filters out accidental impact interference and improves the reliability of decision indicators.

[0016] In one embodiment, the average static resistance value at the current moment is obtained by performing low-pass filtering on the real-time contact resistance sequence to filter out high-frequency fluctuation interference and retaining the DC component that can reflect the long-term drift trend of the resistance as the average static resistance value.

[0017] In one embodiment, determining whether the failure precursor index exceeds a preset safety warning line for multiple consecutive cycles includes: real-time monitoring of the numerical change of the failure precursor index; when the failure precursor index exceeds the safety warning line for a preset number of consecutive calculation cycles, determining that the connector has entered the active period of fretting wear.

[0018] In one embodiment, the vibration damping sensitivity coefficient is obtained by: collecting the resistance fluctuation amplitude and the vibration energy amplitude during the connector's health reference period; fitting the linear relationship between the logarithm of the vibration energy amplitude and the resistance fluctuation amplitude using the least squares method, and obtaining the slope as the vibration damping sensitivity coefficient.

[0019] In one embodiment, the initial resistance fingerprint value is obtained as follows: after the electronic connector is installed and enters a stable operating state, contact resistance data is collected during a break-in period of a preset duration, and the arithmetic mean of the contact resistance data is calculated as the initial resistance fingerprint value and stored.

[0020] The technical solution of this application has the following beneficial technical effects: This application utilizes vibration energy to predict theoretical resistance fluctuations and eliminates them from actual signals. In essence, it constructs an adaptive filter that is sensitive only to abnormal signals that do not conform to the vibration damping following logic. This enables the system to accurately extract weak wear characteristics even under strong vibration conditions, significantly reducing the false alarm rate and achieving accurate perception of the connector's health status.

[0021] Furthermore, by introducing a failure precursor index with time decay weight, the cumulative effect of fretting wear and static drift trend are comprehensively considered, significantly extending the maintenance window and effectively filtering out occasional impact interference. During the latency period when the absolute value of resistance does not exceed the standard, potential hazards can be detected earlier than existing technologies, effectively avoiding ablation accidents. Attached Figure Description

[0022] Figure 1 This is a flowchart of a method for identifying early signs of electronic connector failure based on multi-source data fusion, according to an embodiment of this application.

[0023] Figure 2 This is a characteristic mapping diagram of vibration damping coupling deviation according to an embodiment of this application; Figure 3 This is a schematic diagram comparing the failure precursor index provided in this embodiment of the invention with the prior art. Detailed Implementation

[0024] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0025] Figure 1 This is a flowchart illustrating a method for identifying precursory failures of electronic connectors based on multi-source data fusion, according to an embodiment of this application. Figure 1 As shown, the method for identifying early signs of electronic connector failure based on multi-source data fusion includes steps S101 to S104, which are described in detail below.

[0026] S101 collects real-time contact resistance sequences at both ends of the connector, triaxial vibration acceleration data of the connector housing, and ambient temperature of the connector surface. It also preprocesses and aligns the contact resistance sequence and triaxial vibration acceleration data to separate high-frequency fluctuation components and vibration energy envelope signals.

[0027] In one embodiment, a high-frequency voltage / current sensor with a sampling rate greater than 10 kHz can be used to acquire the real-time contact resistance sequence at both ends of the connector; a MEMS accelerometer mounted on the connector housing can be used to synchronously acquire triaxial vibration acceleration data; and a thermistor can be used to acquire the ambient temperature of the connector surface.

[0028] In this optional embodiment, during the preprocessing stage, the real-time contact resistance sequence can be detrending to subtract the DC component from the original data and retain the high-frequency fluctuation component.

[0029] Furthermore, the collected triaxial vibration acceleration data were vector synthesized using the Euclidean Norm calculation method to obtain the total vibration acceleration signal; then, envelope detection was performed on the total vibration acceleration signal to extract the vibration energy envelope signal, which is used to reflect the trend of vibration energy change over time.

[0030] Meanwhile, the high-frequency fluctuation component is aligned with the timestamp of the vibration energy envelope signal by means of hardware triggering or software interpolation, so that the phase synchronization error is less than a preset threshold, for example, the preset threshold is 0.1ms.

[0031] In this way, pure and time-synchronized dynamic characteristic signals can be obtained, eliminating the analysis errors caused by DC bias and asynchrony, and providing accurate data support for subsequent coupling analysis.

[0032] S102 calculates the vibration damping coupling deviation within the current time window based on high-frequency fluctuation components, vibration energy envelope signals, and ambient temperature using a preset vibration damping coupling deviation calculation model. The vibration damping coupling deviation is used to characterize the difference between the contact interface's response to vibration and the healthy elastic model.

[0033] In one embodiment, Hertzian contact theory from contact physics can be utilized, where, under healthy conditions, the resistance fluctuation amplitude should linearly follow the logarithmic change of vibration energy; however, when fretting wear generates oxide particles, this following logic is broken.

[0034] Based on the high-frequency fluctuation components, vibration energy envelope signal, and ambient temperature, the vibration damping coupling deviation within the current time window is calculated using a pre-defined vibration damping coupling deviation calculation model. The vibration damping coupling deviation satisfies the following relationship:

[0035] in, For the first The vibration damping coupling deviation within a time window is used to characterize the difference between the contact interface response to vibration and the healthy elastic model. This represents the number of data points within the sampling window, for example, 1024; For the first in the window The resistance fluctuation amplitude at the nth moment is obtained by measuring the resistance fluctuation amplitude within the current time window. The original resistance sample value at each time point is subtracted from the arithmetic mean of all resistance sample values ​​within that time window, and the absolute value of the difference is taken to calculate the result. This refers to the nominal contact resistance value of the connector, for example. ; To prevent extremely small positive numbers from being divided by zero, for example, the value is taken as... ; This is the vibration damping sensitivity coefficient, which, for example, is set to 0.5; For the first in the window The vibration energy amplitude at a given moment, for example, is 2g; The unit vibration energy is taken as a reference, for example, with a value of 1g; This is the logarithmic correction constant, which, for example, takes a value of 1; This is a temperature correction function. For the first The ambient temperature at that moment.

[0036] In this optional embodiment, the linear relationship between the logarithm of the vibration energy amplitude and the resistance fluctuation amplitude can be fitted by the least squares method, and the resulting slope can be used as the vibration damping sensitivity coefficient. The value increases with the increase of the difference between the ambient temperature and the normal temperature reference value. The temperature correction function includes a temperature coefficient to adjust the correction weight of the ambient temperature on the vibration damping coupling deviation, so as to reflect the accelerating effect of the high temperature environment on the instability of the oxide film. For example, it is defined as follows: ,in The reference value is the room temperature value. This is the temperature coefficient.

[0037] For example, suppose at a certain moment The connector is in good condition, vibration energy The corresponding theoretical resistance fluctuation should follow a logarithmic relationship, then Then the theoretical term is At this time, the actual resistance fluctuates. After normalization, it is also approximately The difference between the two is close to 0; if fretting wear occurs, the actual resistance fluctuation may suddenly increase, becoming after normalization... At this point, the absolute value of the difference is , The value will increase significantly.

[0038] In this way, it is possible to assess whether the response of the contact interface to vibration conforms to the healthy elasticity model, thereby sensitively capturing nonlinear abnormal fluctuations caused by oxidized particles.

[0039] S103 generates the failure precursor index for the current moment based on the vibration damping coupling deviation over a historical period and the relative rate of change of static resistance at the current moment, combined with the time decay weight.

[0040] In one embodiment, a true precursor to failure manifests as The continuous increase in [something] means that an anomaly at a single moment may be a disturbance, so it is necessary to make a comprehensive judgment by combining historical cumulative damage and static drift trend.

[0041] Specifically, based on the vibration damping coupling deviation over a historical time period and the relative rate of change of static resistance at the current moment, combined with time decay weights, a failure precursor index is generated for the current moment. The failure precursor index satisfies the following relationship:

[0042] in, For the current moment The failure precursor index, The length of the historical observation window, for example, 1000; For the first time in history The vibration damping coupling deviation calculated at any given time. The time decay weighting function is... Using a time constant makes recent anomalies have a larger weight; This is a balance coefficient between static and dynamic characteristics; for example, it takes a value of 10. This represents the average static resistance at the current moment. The initial resistance fingerprint value for the connector during initial installation.

[0043] In this optional embodiment, The method of obtaining the real-time contact resistance sequence is to perform low-pass filtering to remove high-frequency fluctuation interference and retain the DC component that can reflect the long-term drift trend of the resistance, and use it as the average static resistance. For example, the initial value. The method of obtaining the data is as follows: after the electronic connector is installed and enters a stable operating state, the contact resistance data is collected during the break-in period of a preset time, and the arithmetic mean of the contact resistance data is calculated as the initial resistance fingerprint value and stored.

[0044] In this way, by combining dynamic cumulative features and static drift features, misjudgment based on a single indicator can be avoided. It captures early dynamic anomalies and combines the static manifestations of physical damage, thus improving the comprehensiveness of the assessment.

[0045] S104, determine whether the failure precursor index exceeds the preset safety warning line for multiple consecutive cycles, in order to determine whether there is a risk of failure precursor to the connector.

[0046] In one embodiment, the system monitors the changes in the failure precursor index in real time. When the failure precursor index is greater than the safety warning line for a preset number of calculation cycles, for example, 5 cycles, it can be determined that the connector has entered the active period of fretting wear and there is a risk of failure precursor. At this time, a yellow warning signal is issued.

[0047] like Figure 2 The diagram shown is a characteristic mapping of vibration damping coupling deviation according to an embodiment of this application. It can be seen that the data points in the elastic response zone are closely distributed near the logarithmic curve, while the data points in the coupling logic failure zone are significantly deviated from the baseline, showing the distribution pattern of failure precursors.

[0048] like Figure 3 The diagram shown is a comparison between the failure precursor index provided by the embodiment of the present invention and the prior art. It shows that the static resistance monitoring curve of the prior art only rises sharply until the fault critical point, while the curve of the present invention shows an exponential upward trend in the middle of the time axis and crosses the warning line at a time point significantly earlier than that of the prior art, proving that the present invention can obtain a maintenance window period.

[0049] In this way, by making continuous periodic judgments, outlier interference caused by occasional shocks can be effectively filtered out, ensuring the accuracy and reliability of the early warning signal.

[0050] It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the scope of protection of this application. Therefore, the scope of protection of this patent application shall be determined by the appended claims.

Claims

1. A multi-source data fusion-based electronic connector failure precursor identification method, characterized in that, include: The system collects real-time contact resistance sequences at both ends of the connector, triaxial vibration acceleration data of the connector housing, and ambient temperature of the connector surface. It also preprocesses and aligns the contact resistance sequences and triaxial vibration acceleration data to separate high-frequency fluctuation components and vibration energy envelope signals. Based on the high-frequency fluctuation components, the vibration energy envelope signal, and the ambient temperature, the vibration damping coupling deviation within the current time window is calculated using a preset vibration damping coupling deviation calculation model. The vibration damping coupling deviation characterizes the difference between the contact interface's response to vibration and the healthy elastic model. The vibration damping coupling deviation satisfies the following relationship: in, For the first Vibration damping coupling deviation within a time window The number of data points within the sampling window. For the first in the window The resistance fluctuation amplitude at each moment. This is the nominal contact resistance value of the connector. To prevent dividing by zero from being the smallest positive number, The vibration damping sensitivity coefficient, For the first in the window The amplitude of vibrational energy at each moment. As a unit vibration energy standard, The logarithmic correction constant, This is a temperature correction function. For the first The ambient temperature at a given time; the value of the temperature correction function increases as the difference between the ambient temperature and the normal temperature reference value increases; the temperature correction function includes a temperature coefficient, which is used to adjust the correction weight of the ambient temperature on the vibration damping coupling deviation, so as to reflect the accelerating effect of high temperature environment on oxide film instability; the vibration damping sensitivity coefficient is obtained by: collecting the resistance fluctuation amplitude and vibration energy amplitude during the connector's health reference period; fitting the linear relationship between the logarithm of the vibration energy amplitude and the resistance fluctuation amplitude using the least squares method, and the resulting slope is the vibration damping sensitivity coefficient; Based on the vibration damping coupling deviation over the historical time period and the relative rate of change of static resistance at the current moment, combined with time decay weighting, a failure precursor index is generated for the current moment; the failure precursor index satisfies the following relationship: in, For the current moment The failure precursor index, The length of the historical observation window. For the first time in history The vibration damping coupling deviation calculated at any given time. The time decay weighting function is... It is a time constant. It is a balance coefficient between static and dynamic characteristics. This represents the average static resistance at the current moment. Initial resistance fingerprint value for the connector during initial installation; The failure precursor index is determined to exceed a preset safety warning line for multiple consecutive cycles in order to determine whether the connector has a risk of failure precursor.

2. The method for identifying early signs of electronic connector failure based on multi-source data fusion according to claim 1, characterized in that, The preprocessing and time alignment of the contact resistance sequence and the triaxial vibration acceleration data includes: The real-time contact resistance sequence is detrended to filter out the DC component and retain the high-frequency fluctuation component. By using hardware triggering or software interpolation, the timestamps of the high-frequency fluctuation components and the vibration energy envelope signal are aligned, so that the phase synchronization error is less than a preset threshold.

3. The method for identifying precursory failures of electronic connectors based on multi-source data fusion according to claim 1, characterized in that, The methods for obtaining the vibration energy envelope signal include: The collected triaxial vibration acceleration data are vector synthesized to obtain the total vibration acceleration signal; The total vibration acceleration signal is subjected to envelope detection processing to extract the vibration energy envelope signal, which is used to reflect the trend of vibration energy change over time.

4. The method for identifying precursory failures of electronic connectors based on multi-source data fusion according to claim 1, characterized in that, The method for obtaining the average static resistance value at the current moment is as follows: the real-time contact resistance sequence is subjected to low-pass filtering to filter out high-frequency fluctuation interference, and the DC component that can reflect the long-term drift trend of the resistance is retained as the average static resistance value.

5. The method for identifying precursory failures of electronic connectors based on multi-source data fusion according to claim 1, characterized in that, The determination of whether the failure precursor index exceeds the preset safety warning line for multiple consecutive cycles includes: Real-time monitoring of the numerical changes in the failure precursor index; When the calculation period of the failure precursor index is greater than the safety warning line for a consecutive preset number of times, the connector is determined to have entered the active period of fretting wear.

6. The method for identifying early signs of electronic connector failure based on multi-source data fusion according to claim 1, characterized in that, The initial resistance fingerprint value is obtained as follows: After the electronic connector is installed and enters a stable operating state, contact resistance data is collected during a preset break-in period. The arithmetic mean of the contact resistance data is calculated as the initial resistance fingerprint value and stored.

Images