A method and system for testing the damage of automobile electrical appliances under extreme temperature impact

Abstract

The application discloses a kind of automobile electric appliance extreme temperature impact damage test method and system, it is related to automobile electric appliance reliability test technology.Imply-40 ℃~125 ℃ cyclic temperature impact is applied using temperature change test box, and wide-frequency impedance spectrum is collected by impedance analyzer at the end of residence;Equivalent circuit fitting is extracted electrical parameter, and damage feature vector is constructed after normalization processing, and weighted operation obtains dynamic damage index.Failure is determined using the double criterion that damage index is greater than or equal to 0.7 and sensitive frequency point impedance is 3 times the original value, and the cycle number is recorded.Statistics batch average, minimum life after batch testing;When damage index is to 0.4, linear extrapolation is used to predict the remaining life of single product, and the result is corrected by three verifications.The matching system includes temperature change test box, impedance analyzer, digital signal processor, damage evaluator and life prediction unit.The application realizes damage dynamic monitoring and life prediction, and makes up for the deficiency of traditional post-detection.

Description

Technical Field

[0001] The present invention belongs to the technical field of reliability testing of automotive electrical appliances, and particularly relates to a method and system for testing extreme temperature shock damage of automotive electrical appliances, which is especially applicable to the dynamic evaluation of thermal fatigue damage, critical failure determination, and remaining life prediction of automotive-grade electronic components in an extreme temperature cycling environment. Background Art

[0002] In the context of the continuous improvement of automotive electronicization, automotive electrical appliances such as engine control units, body controllers, and sensors need to withstand extreme temperature shock environments during service. For example, the temperature in the engine compartment frequently alternates between -40°C and +125°C. Such a drastic temperature cycle can cause cumulative thermal fatigue damage to internal solder joints, wire bonding interfaces, dielectric materials, etc. in the electrical appliances, ultimately leading to functional failures. Currently, the reliability testing of automotive electrical appliances in the industry mainly refers to vehicle-grade component verification standards such as AEC-Q100 and AEC-Q200, and uses a programmable temperature change test chamber to conduct a specified number of temperature cycle tests on samples, and then checks whether the samples are失效 after the test. The above method can only determine whether the sample is qualified or unqualified, and cannot dynamically monitor the cumulative degree of damage during the test.

[0003] In the prior art, for example, the Chinese invention patent "Overload Testing Device for Photoelectric Components under High-Speed Impact and Its Damage Evaluation Method" (Application No.: 202510713309.3) discloses an overload testing device for photoelectric components under high-speed impact and a damage evaluation method. This technology applies a high-speed impact force to the test body, uses a load measurement component to measure the impact load transmitted to the photoelectric component, takes out the photoelectric component after the impact and observes its damage situation on the damage detection component, and obtains the corresponding relationship between the load and the damage after repeating the test multiple times. This technical solution solves the technical problem that directly applying a high-speed impact force to the photoelectric component is likely to cause multiple damages, and provides a reference for the selection and anti-overload design of photoelectric components in the guidance system. However, this technical solution still belongs to the ex-post evaluation mode of impacting first and then detecting, and cannot real-time monitor the evolution process of damage during the impact, and can only establish a static corresponding relationship between the load and the damage, and cannot dynamically predict the remaining life of the sample. At the same time, this technical solution is aimed at the high-speed single-impact scenario and is not applicable to the evaluation of thermal fatigue cumulative damage of automotive electrical appliances under cyclic temperature impact.

[0004] In view of this, the present invention proposes a method and system for testing extreme temperature shock damage of automotive electrical appliances. Summary of the Invention

[0005] In view of the shortcomings of the prior art, the purpose of this invention is to provide a method and system for testing extreme temperature shock damage of automotive electrical appliances that can dynamically and quantitatively evaluate the thermal fatigue damage process of automotive electrical appliances under extreme temperature shock environment, and can realize batch reliability statistics and individual remaining life prediction.

[0006] To achieve the above objectives, the present invention adopts the following technical solution; A method for testing extreme temperature shock damage to automotive electrical components includes the following steps: S100. Place the automotive electrical appliance under test in a programmable temperature change test chamber, and control the temperature of the test chamber to cycle between the lowest temperature and the highest temperature. At the end of each temperature dwell stage, use an impedance analyzer connected in parallel with the port of the electrical appliance under test to scan at multiple characteristic frequency points in a preset frequency band and collect the impedance spectrum sequence that changes with the number of cycles. S200. Perform equivalent circuit fitting on the impedance spectrum for each cycle number, extract at least two key electrical parameters, and calculate the normalized rate of change of each key electrical parameter relative to the initial value, and combine them into a cumulative damage feature vector. S300. Calculate the dynamic damage accumulation index based on the cumulative damage feature vector and monitor the impedance modulus at at least one characteristic frequency point; when the dynamic damage accumulation index reaches the first threshold, or the impedance modulus at the characteristic frequency point exceeds a predetermined percentage of the initial value, determine that the electrical appliance under test has reached the critical failure point, trigger the test termination and record the number of critical failure cycles of the sample. S400. Repeat steps S100 to S300 to obtain the critical failure cycle counts of multiple samples in the same batch. Use the critical failure cycle counts of multiple samples to perform batch lifetime statistics and output the batch reliability index. For a single test sample in the batch, after the dynamic damage accumulation index exceeds the preset early warning threshold, use the measured dynamic damage accumulation index trajectory of the sample to perform linear extrapolation fitting, solve for the predicted remaining lifetime of the sample, and output the prediction results after verification.

[0007] As a further aspect of the present invention, the specific steps for acquiring the impedance spectrum sequence that varies with the number of cycles include: The automotive electrical appliance under test is fixed on the test fixture, and a full-band frequency sweep measurement is performed on the electrical port using an impedance analyzer at room temperature. The initial impedance spectrum is recorded as the reference value. Set the temperature cycling parameters: the lowest temperature is -40℃, the highest temperature is 125℃, the temperature change rate is 30℃ / min, and the dwell time at each extreme point is 30 minutes. At the end of each temperature dwell phase, the impedance analyzer is triggered to select 15 to 20 characteristic frequency points in the logarithmic coordinate system, covering the frequency band from 1 kHz to 10 MHz for frequency sweep measurement, and the impedance magnitude and phase angle at each frequency point are recorded to form complex impedance information. The collected data is organized according to the number of temperature shock cycles to form an impedance spectrum sequence that varies with the number of cycles and is then stored.

[0008] As a further aspect of the present invention, the specific steps for extracting the cumulative damage feature vector include: Construct an equivalent circuit topology, wherein the equivalent circuit topology is as follows: the contact resistor, the constant phase angle element, and the polarization resistor are connected in parallel, and then connected in series with the lead inductor. The Levenberg-Marquardt nonlinear least squares algorithm is used to fit the measured impedance spectrum to the equivalent circuit topology to obtain the key electrical parameters under the current cycle, including contact resistance, constant phase angle element amplitude parameters, polarization resistance and lead inductance. Using the contact resistance, constant phase angle element amplitude parameters, polarization resistance, and lead inductance at the initial cycle as a reference, calculate the normalized rate of change of each parameter; The four normalized rates of change are combined into a cumulative damage feature vector, and the rate of change sequence of each parameter is filtered using a three-point moving average.

[0009] As a further aspect of the present invention, the specific steps for determining the critical failure point include: The dynamic damage accumulation index is calculated by weighted summation of the four normalized rates of change in the cumulative damage feature vector. The sum of the weight coefficients of each rate of change is one and none of them are less than zero. The weight coefficients are determined by pre-experiment calibration. Calculate the cumulative rate of change of impedance modulus at each test frequency point, find the frequency point that makes the cumulative rate of change reach a local maximum as a candidate sensitive characteristic frequency, calculate and update in each loop, and finally confirm the sensitive characteristic frequency after three consecutive loops. A dual-criteria failure determination condition is set. Criterion 1 is that the dynamic damage accumulation index reaches 0.7, and criterion 2 is that the impedance modulus at the sensitive characteristic frequency point reaches or exceeds 3 times the initial value. When either criterion is met for the first time, the electrical appliance under test is determined to have reached the critical failure point. Once the critical failure point is determined, the test is stopped immediately, and the current cycle count is recorded as the critical failure cycle count.

[0010] As a further aspect of the present invention, the specific method for batch lifetime statistics is as follows: at least three samples in the same batch are tested to obtain the critical failure cycle number of each sample, the arithmetic mean of the critical failure cycle number of all samples is calculated as the average lifetime, the minimum value among them is taken as the minimum lifetime, and the average lifetime and the minimum lifetime are output as batch reliability indicators.

[0011] As a further aspect of the present invention, the specific steps for obtaining the predicted remaining lifetime of the sample by linear extrapolation fitting using the measured dynamic damage accumulation index trajectory of the sample include: Real-time monitoring of the dynamic damage accumulation index; when the index first reaches 0.4, the number of cycles at this point is recorded as the warning start cycle number. Based on the number of cycles from the warning start point, dynamic damage accumulation index data points from three consecutive cycles—the cycle before the warning start point, the current cycle from the warning start point, and the cycle after the warning start point—are selected. The damage growth rate is calculated using the first and last point difference method, which is to subtract the damage index from the cycle before the warning start point from the damage index of the cycle after the warning start point, and then divide by 2. Divide 0.3 by the growth rate to get the remaining number of cycles required from the warning start point to failure, and round it to the nearest integer as the predicted remaining lifetime output.

[0012] As a further aspect of the present invention, the specific steps for verifying the prediction results and then outputting them include: The first verification step: Check whether the damage growth rate reaches or exceeds 0.01. If it does not reach 0.01, no predicted value will be output, and a message indicating that the damage growth is slow and continued monitoring is recommended will be output. The second verification step is to check the rationality of the linear assumption, that is, to calculate the deviation between the actual damage index of the current cycle at the warning starting point and the linear interpolation result. If the deviation exceeds 0.05, the additional data fluctuation will be large when outputting the predicted remaining lifespan, and the prediction result will be for reference only. The third verification step: Check whether the predicted remaining lifetime has reached or exceeded 5 cycles. If it is less than 5 cycles, output a prompt message indicating that it is about to fail and recommends checking immediately, and output the predicted value at the same time. Based on the combined results of the three verifications, the final output format is determined.

[0013] As a further aspect of the present invention, if the test sample has triggered the criterion that the impedance modulus at the sensitive characteristic frequency point changes to more than three times the initial value before the dynamic damage accumulation index reaches 0.4, it is determined to be a sudden local failure, and the individual remaining life prediction step is skipped, and only the number of critical failure cycles is recorded.

[0014] Another aspect of this application provides an automotive electrical extreme temperature shock damage testing system, the system comprising: The programmable rapid temperature change test chamber is used to house the automotive electrical components under test and apply periodic extreme temperature shocks according to a preset temperature cycling curve. The high-precision impedance analyzer is connected in parallel with the external port of the automotive electrical appliance under test via a test cable. It is used to perform wideband sweep frequency measurement on the electrical port at the end of each temperature dwell phase and output the impedance magnitude and phase angle at each frequency point. A digital signal processor, connected to the output of an impedance analyzer, is used to receive impedance spectrum sequences, perform equivalent circuit fitting, extract key electrical parameters, calculate normalized rate of change, and generate cumulative damage feature vectors. The damage evaluator, connected to the output of the digital signal processor, is used to calculate the dynamic damage accumulation index, identify sensitive characteristic frequencies, perform dual-criteria failure determination, and send a test termination signal to the test chamber when the determination reaches the critical failure point. The life prediction unit is connected to the output of the damage evaluator and is used to receive the critical failure cycle count and dynamic damage accumulation index trajectory to perform batch life statistics and individual remaining life linear extrapolation prediction.

[0015] As a further aspect of the present invention, the lifetime prediction unit is also equipped with a report generation module, which is used to automatically generate a test report containing the critical failure cycle number of each sample, the average lifetime of the batch, the minimum lifetime of the batch, and the predicted remaining lifetime of the individual sample after the testing of at least three samples is completed.

[0016] In summary, due to the adoption of the above technical solution, the beneficial technical effects of the invention are as follows: This invention represents a technological leap from post-event assessment to dynamic monitoring. Unlike existing technologies that rely on post-event assessment followed by testing, this invention simultaneously acquires broadband impedance spectra during temperature shock cycling. It then uses equivalent circuit fitting to extract key electrical parameters such as contact resistance, amplitude parameters of constant phase angle elements, polarization resistance, and lead inductance. The normalized rate of change of each parameter relative to its initial value is calculated, constructing a cumulative damage feature vector that evolves with the number of cycles. This feature vector, presented as a four-dimensional normalized rate of change, quantitatively characterizes the cumulative process of physical damage, including the propagation of microcracks in solder joints, interface layer degradation, and dielectric material aging within electrical components. This enables dynamic quantitative assessment of the thermal fatigue damage process, filling a technological gap where existing technologies cannot monitor the degree of damage accumulation during testing.

[0017] This invention improves the accuracy and timeliness of failure assessment. Unlike existing technologies that rely on single threshold judgments or post-event inspections, this invention employs a dual-criteria fusion strategy combining a dynamic damage accumulation index and the impedance modulus at sensitive characteristic frequencies. Firstly, a comprehensive damage index is obtained by weighted fusion of four normalized rates of change; when this index reaches 0.7, it is determined to be an overall degradation failure. Secondly, sensitive characteristic frequencies are automatically identified by calculating the cumulative rate of change of impedance modulus at each frequency; when the impedance modulus at that frequency abruptly changes to three times its initial value, it is determined to be a sudden local failure. The two criteria are independently monitored and mutually verified, enabling both the capture of overall degradation trends and timely response to sudden local failures, effectively avoiding misjudgments or omissions that might occur with a single criterion.

[0018] This invention achieves a two-tiered lifetime prediction system combining batch statistical assessment and precise individual prediction. Unlike existing technologies that can only establish a static correspondence between load and damage, this invention, after obtaining the critical failure cycles of multiple samples, calculates the batch average lifetime and minimum lifetime through simple arithmetic statistics, providing a reliability benchmark for product selection and quality control. Simultaneously, for a single sample under test, prediction is initiated when the dynamic damage accumulation index reaches 0.4. Three consecutive cycles of data points near the warning starting point are selected, and the damage growth rate is calculated using the first-to-last point difference method. The remaining lifetime is predicted through linear extrapolation, and the reliability of the prediction results is ensured through three verifications: the reasonableness of the growth rate, the reasonableness of the linear assumption, and the reasonableness of the remaining lifetime. The entire prediction process involves only four arithmetic operations (addition, subtraction, multiplication, and division), requiring no iterative solution. It has low computational load, is simple to implement, and can provide real-time decision-making support for preventative maintenance and health management.

[0019] The system boasts high integration and automated testing processes. This invention also provides a complete automotive electrical extreme temperature shock damage testing system, integrating a programmable rapid temperature change test chamber, a high-precision impedance analyzer, a digital signal processor, a damage evaluator, and a life prediction unit. It forms a complete closed loop from stress application, data acquisition, feature extraction, failure determination to life prediction, achieving automation and standardization of the testing process, reducing errors caused by human intervention, and improving testing efficiency and result repeatability.

[0020] In summary, this invention overcomes the technical shortcomings of existing technologies that cannot dynamically monitor the degree of damage accumulation during testing and cannot predict the remaining life of samples in real time. It provides a dynamic, accurate, and predictable complete technical solution for the thermal fatigue reliability assessment of automotive electrical appliances, and has good application prospects and promotional value. Attached Figure Description

[0021] Figure 1 Flowchart of a test method for extreme temperature shock damage to automotive electrical components; Figure 2A flowchart of an extreme temperature shock damage test method S100 for automotive electrical components; Figure 3 A flowchart of an extreme temperature shock damage test method S200 for automotive electrical components; Figure 4 S300 flowchart of an extreme temperature shock damage test method for automotive electrical appliances; Figure 5 This is a flowchart of S400, a test method for extreme temperature shock damage to automotive electrical components. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0024] The core technical logic of this invention: This invention addresses the difficulty in quantifying and assessing thermal fatigue damage of automotive electrical components under extreme temperature shock conditions by proposing a dynamic damage assessment and life prediction method based on broadband impedance spectrum monitoring and equivalent circuit modeling.

[0025] The core logic is that temperature shock causes cumulative thermal fatigue damage to the solder joints, interface layers, and dielectric materials inside electrical appliances, and this damage is directly reflected in the port impedance characteristics of the appliances. By simultaneously acquiring broadband impedance spectra during temperature cycling and extracting key electrical parameters using equivalent circuit models, macroscopic electrical measurements can be transformed into microscopic physical damage characteristics.

[0026] Specifically, the technical logic of this invention comprises four levels: The first layer involves stress application and data acquisition. A programmable rapid temperature change test chamber is used to apply controllable, periodic extreme temperature shock stress to the electrical appliance under test, simulating the harsh environment faced by automotive electrical appliances in actual use. At the end of each temperature dwell phase, a high-precision impedance analyzer is used to synchronously acquire the broadband impedance spectrum of the electrical appliance port, obtaining raw data that changes with the number of cycles.

[0027] The second layer involves feature extraction and damage quantification. An equivalent circuit model with clear physical meaning is constructed, and the measured impedance spectrum is fitted to this model to extract four key electrical parameters: contact resistance, amplitude parameters of constant phase angle elements, polarization resistance, and lead inductance. By calculating the normalized rate of change of each parameter relative to its initial value, individual differences among samples are eliminated, and a cumulative damage feature vector is formed.

[0028] The third layer involves failure assessment and data recording. The cumulative damage feature vector is weighted and fused to calculate the dynamic damage accumulation index; simultaneously, the most sensitive characteristic frequency in the impedance spectrum is identified. A dual-criteria fusion assessment strategy is employed—when the dynamic damage accumulation index reaches 0.7, or the impedance modulus at the sensitive characteristic frequency point abruptly changes to three times its initial value, the electrical appliance is determined to have reached the critical failure point, and the number of critical failure cycles is recorded.

[0029] The fourth layer involves batch statistics and individual prediction. At least three samples from the same batch are tested to obtain the batch average lifetime and minimum lifetime as reliability indicators. For a single sample to be tested, prediction is initiated when the dynamic damage accumulation index reaches 0.4. Three consecutive data points near the warning starting point are selected, and the damage growth rate is calculated using linear extrapolation to predict the remaining lifetime. The reliability of the prediction results is ensured through three verifications.

[0030] The above four layers of logic, from stress application to data acquisition, from feature extraction to failure determination, and from batch statistics to individual prediction, form a complete closed-loop technical solution, realizing quantitative mapping from macroscopic electrical measurement to microscopic physical damage, and dual-level life prediction from batch statistical assessment to precise individual prediction.

[0031] Example 1: refer to Figure 1 As one embodiment of the present invention, this embodiment also provides a method for testing extreme temperature shock damage of automotive electrical appliances. This method aims to achieve dynamic quantitative assessment and remaining life prediction of the thermal fatigue damage process of automotive electrical appliances through the synergistic effect of periodic temperature shock excitation and broadband impedance spectrum monitoring.

[0032] Specifically, it includes: S100. Apply stress and acquire raw impedance spectrum data; S200. Extract the cumulative damage feature vector from the impedance spectrum; S300. Determine the critical failure point based on the damage feature vector and record the failure data; S400. Utilize failure data and damage evolution trajectories for statistical evaluation and lifetime prediction.

[0033] In a method for testing extreme temperature shock damage to automotive electrical components, the purpose of S100 is to apply controllable and periodic extreme temperature shock stress to the automotive electrical components under test, and to synchronously acquire their port broadband impedance spectrum at a specific phase point in each temperature cycle, providing raw data for subsequent damage feature extraction.

[0034] Specific steps are as follows Figure 2 As shown: S110. Installation and Initial Calibration: The automotive electrical component under test is fixed to the test fixture, ensuring a reliable connection between the electrical port and the external measurement circuit. The fixture is then placed inside the working chamber of the programmable rapid temperature change test chamber. At room temperature (25℃), a full-band frequency sweep measurement is performed on the electrical port using an impedance analyzer, and the initial impedance spectrum is recorded as the reference value for subsequent normalization calculations.

[0035] S120. Set temperature cycle parameters: The temperature cycling parameters are set via the test chamber controller. In this embodiment, the minimum temperature is set to -40℃, the maximum temperature to +125℃, the temperature change rate to 30℃ / min, and the dwell time at each extreme point to 30 minutes. These parameters cover the typical extreme environments faced by automotive-grade electronic components in the engine compartment and exterior of the vehicle body. The test chamber operates automatically according to the set curve, with the temperature periodically switching between the minimum and maximum temperatures, thus creating accelerated thermal fatigue excitation for the electrical components.

[0036] S130. Synchronous acquisition of impedance spectrum: At the end of each temperature dwell phase (i.e., when the appliance maintains a constant temperature at its lowest or highest point), a high-precision impedance analyzer connected in parallel with the appliance's ports is triggered to perform measurements. Data acquisition is performed at the end of each dwell phase to avoid transient interference from temperature inhomogeneities during temperature changes on the impedance measurement.

[0037] The impedance analyzer selects 15 to 20 characteristic frequency points in a logarithmic coordinate system to perform continuous frequency sweep measurements covering the 1kHz to 10MHz frequency band.

[0038] The reason for using a logarithmic coordinate system instead of a nonlinear coordinate system is that the changes in electrical parameters in electrochemical impedance spectroscopy (EIS) often exhibit different sensitivity characteristics in the low-frequency and high-frequency ranges. The logarithmic distribution can uniformly acquire characteristic information over a wider frequency range.

[0039] At each frequency point, the impedance analyzer simultaneously records two key quantities: Impedance magnitude reflects the total impedance of an electrical port at that frequency. Phase angle reflects the phase difference between voltage and current.

[0040] The impedance magnitude and phase angle together constitute the complex impedance information at this frequency point, expressed as: (1); This is the impedance magnitude; Phase angle; It is an imaginary number; For the first Each characteristic frequency point The value is between 15 and 20; The logic behind the above expression lies in integrating the impedance magnitude and phase angle measured by the impedance analyzer into a polar coordinate expression for the complex impedance, thereby providing a complete description of the electrical port at a given frequency. AC electrical response characteristics under the following conditions; for the first Sub-temperature shock cycle Record each frequency point impedance magnitude at the following and phase angle The initial state (when no temperature shock has been applied) is denoted as... .

[0041] The expression establishes a mapping relationship between the measured data and the complex domain, enabling subsequent least squares fitting to utilize information from both the magnitude and phase angle dimensions, thereby improving the accuracy and robustness of parameter extraction.

[0042] S140. Organize and store data: The collected data were categorized according to the number of temperature shock cycles. Organize and form according to the number of cycles Changing impedance spectrum sequence The sequence is then transmitted to a digital signal processor for storage.

[0043] During the process: Indicates the initial, unimpacted state; This indicates the number of temperature shock cycles completed, with the number of cycles per cycle being [number missing]. This corresponds to a complete temperature shock cycle (i.e., the entire process of cooling from room temperature to the lowest temperature, staying at the lowest temperature, heating up to the highest temperature, and staying at the highest temperature).

[0044] The logic behind the above data organization method can be understood as: based on the number of loops... Using a time-axis index, the evolution of electrical damage with increasing temperature shocks can be clearly tracked, providing structured raw data for subsequent cumulative damage feature extraction and life prediction.

[0045] The temperature shock application and impedance spectrum acquisition of the tested automotive electrical components were completed through the above steps S110-S140.

[0046] In a test method for extreme temperature shock damage of automotive electrical components, the purpose of S200 is to transform the original impedance spectrum data into a low-dimensional cumulative damage feature vector with clear physical meaning, thereby realizing a quantitative mapping from macroscopic electrical measurements to microscopic physical damage.

[0047] Specific steps are as follows Figure 3 As shown: S210. Construct the equivalent circuit topology: Based on the port characteristics of the automotive electrical appliances under test, an equivalent circuit topology with clear physical meaning is constructed.

[0048] The equivalent circuit topology used in this embodiment is: a contact resistor that is independent of frequency. A constant phase angle element representing dielectric loss. A polarization resistor representing the interface effect. After the three are connected in parallel, they are then connected to an inductor representing the lead inductance. Series, i.e. Topology.

[0049] The described topology can effectively describe the changes in the electrical behavior of internal solder joints, lead frames, dielectric materials, and interface layers of electrical appliances during thermal fatigue. Specifically: Contact resistance : Represents the contact resistance of electrical terminals and the ohmic resistance of internal leads and solder joints. It is most sensitive to microcracks in solder joints and increases as the cracks expand. Constant phase angle element : Represents the polarization loss of dielectric materials. Its amplitude parameter changes with the aging of dielectric materials and is used to evaluate the degradation of insulation performance; polarization resistor : Represents the charge transfer resistance of the interface layer, which decreases as the interface deteriorates; Lead inductor : Represents the parasitic inductance formed by the package leads and PCB traces. It is relatively stable and mainly serves as an auxiliary reference.

[0050] S220. Equivalent circuit fitting: For each loop count The impedance spectrum measured below The digital signal processor uses the Levenberg-Marquardt nonlinear least squares algorithm to fit the measured impedance spectrum to the equivalent circuit topology described above. The fitting calculation yields four key electrical parameters for the current cycle: contact resistance, amplitude parameters of constant phase angle elements, polarization resistance, and lead inductance.

[0051] The fitting process solves the following least-squares problem by finding a set of circuit parameters that minimizes the sum of squared errors between the theoretical impedance value calculated by the model and the measured impedance value: (2); in, For the first Frequency point in the next cycle The measured complex impedance at the point is calculated using formula (1) in step S130; The theoretical complex impedance calculated for the equivalent circuit model under given parameters; This represents the total number of characteristic frequency points, with a value between 15 and 20. The theoretical impedance expression for the equivalent circuit model: For the implementation of this embodiment Topology, theoretical complex impedance The calculation requires knowing the impedance expressions for each component.

[0052] First, constant phase angle element Complex impedance The expression is: (3); in, Angular frequency; The imaginary unit; This is the amplitude parameter of a constant phase angle element, and its value reflects the degree of polarization of the dielectric material; The diffusion index has a range of values ​​of 100. ,when Time phase angle element Degenerates into an ideal capacitor when It degenerates into a pure resistor. In this embodiment... Take a fixed empirical value of 0.85; Secondly, contact resistance Constant phase angle element Polarization resistance The equivalent impedance after the three are connected in parallel for: (4); Finally, connect it to the lead inductor. By connecting them in series, the overall theoretical impedance is obtained. : (5); In formula (5), in the low-frequency range, the current mainly flows through the polarization resistor. and constant phase angle element Impedance characteristics are dominated by interface and dielectric behavior; in the mid-to-high frequency range, contact resistance... The series effect gradually becomes apparent; at extremely high frequencies, the lead inductance... The inductive effect becomes dominant. The contribution of each parameter to the total impedance has different sensitivities in different frequency ranges, which is the physical basis for broadband impedance spectroscopy to extract multiple independent parameters simultaneously.

[0053] The Levenberg-Marquardt algorithm combines the advantages of the Gauss-Newton method and the gradient descent method, and can efficiently solve nonlinear least squares problems. It has wide applications in the field of impedance spectrum fitting.

[0054] S230. Calculate the normalized rate of change: Read initial Initial contact resistance corresponding to the state Initial constant phase angle element amplitude parameters Initial polarization resistance Initial lead inductance As a benchmark. In order to eliminate the differences in the initial absolute values ​​between different samples and make the damage characteristics comparable, it is necessary to calculate the rate of change of each parameter relative to the initial value.

[0055] The formula for calculating the normalized rate of change is as follows: (6); (7); (8); (9); in, This represents the current contact resistance. This refers to the amplitude parameters of the current constant phase angle element; This is the current polarization resistance; For the current lead inductance; This represents the rate of change of contact resistance. The rate of change of a constant phase angle element; This represents the rate of change of polarization resistance. This represents the rate of change of lead inductance. The normalized rate of change is calculated as the proportion of the change in the current parameter value relative to the initial value. A positive value indicates that the parameter is increasing, and a negative value indicates that the parameter is decreasing. Using normalized rate of change eliminates the influence of individual sample differences on damage characteristics, making the damage levels comparable between different samples within the same batch and between different batches. In practical applications, this means that two samples with different initial resistance values ​​are considered to be in the same damage state as long as their relative rates of change are the same.

[0056] S240. Combining eigenvectors and smoothing: The four normalized rates of change mentioned above are combined into a four-dimensional vector, which is defined as the cumulative damage feature vector. : (10); The evolution trajectory of the cumulative damage feature vector with the number of cycles quantitatively describes the accumulation process of internal physical damage in electrical appliances. For example, The continuous increase in size indicates that the weld crack is constantly expanding; A continuous decrease indicates that the interface layer is constantly deteriorating.

[0057] These constitute four independent rate-of-change sequences, each consisting of a set of discrete data points with the number of cycles on the x-axis and the rate of change of the parameter on the y-axis.

[0058] To reduce the impact of single-measurement noise on the feature vector, a three-point moving average is used to filter the rate-of-change sequence of each parameter: In actual measurement processes, due to factors such as the inherent noise of the impedance analyzer, the potential slight displacement of the test cable during temperature cycling, and vibration of the test chamber, random measurement noise inevitably gets mixed into each rate of change sequence. This noise manifests as non-physical jitter in the rate of change between adjacent cycles, and if left unaddressed, it will interfere with subsequent failure determination and lifespan prediction.

[0059] To reduce the impact of single-measurement noise on the feature vector, the rate of change sequence of each parameter is filtered using a three-point moving average. For ease of consistent description, let's call it... For any parameter ( The normalized rate of change at any given iteration is the filtered rate of change. Calculation formula: (11); in, This represents the normalized rate of change of the parameter in the previous iteration; This represents the normalized rate of change of the parameter in the current loop; This represents the normalized rate of change of the parameter in the next iteration; The filtering formula takes the arithmetic mean of the current point's value and the values ​​of its two adjacent points to smooth out random measurement noise. In the filtered rate of change sequence, the value of each point incorporates the results of one measurement before and one before it, thus effectively suppressing high-frequency random noise while preserving the low-frequency trend of damage evolution.

[0060] edge point ( The initial point (and the last point) is not smoothed because it lacks adjacent points on one side; the original value is retained directly. Since this point is the reference point for normalization calculations, the original value must be retained to ensure the accuracy of all rate of change calculations. For the last measurement point (i.e., the latest cycle), the original value is also retained directly because there is no data for subsequent points, ensuring that the latest damage status can be obtained in real time.

[0061] Through steps S210 to S240, the transformation from the original impedance spectrum to the cumulative damage feature vector was completed. This feature vector, in the form of a four-dimensional normalized rate of change, quantitatively characterizes the evolution of key electrical parameters inside the electrical appliance with the number of temperature shocks, providing a quantitative basis for subsequent failure determination and life prediction.

[0062] In a test method for extreme temperature shock damage of automotive electrical appliances, the purpose of S300 is to construct a comprehensive damage index based on the cumulative damage feature vector and combine it with impedance modulus monitoring at sensitive frequency points to achieve proactive and accurate determination of critical failure points.

[0063] Specific steps are as follows Figure 4 As shown: S310. Calculate the dynamic damage accumulation index : The cumulative damage feature vector Input the damage evaluator and calculate a dimensionless dynamic damage accumulation index using a weighted summation method. This index integrates damage information from four dimensions into a comprehensive indicator, facilitating subsequent threshold determination. (12); Different electrical parameters contribute differently to the overall damage state of electrical appliances, therefore requiring different weighting coefficients. For the weighting coefficients, satisfying And all of them are not less than zero.

[0064] The specific values ​​of each weighting coefficient were determined through preliminary experimental calibration. The calibration method is as follows: Samples from the same batch were subjected to a complete temperature shock test until failure. Simultaneously, corresponding physical failure indicators (such as weld crack length and interface layer resistance changes) were measured offline. Then, multiple linear regression was used to calculate the contribution coefficients of each normalized rate of change to the failure indicators. Finally, the coefficients were normalized to obtain the weighting coefficients. This calibration process ensures that the weighting coefficients match the actual failure modes.

[0065] This embodiment targets the failure mode primarily characterized by solder joint fatigue. All of them are not less than zero; among them, the contact resistance change rate has the highest weight because the microcracks in the solder joint directly lead to a significant increase in contact resistance.

[0066] S320. Frequency of identifying sensitive features: For impedance spectrum sequence Perform frequency domain analysis for each test frequency. Calculate the cumulative rate of change of its impedance magnitude from the initial state to the current cycle number. : (13); The formula represents the ratio of the change in the current impedance magnitude relative to the initial value. The larger the cumulative rate of change, the more sensitive that frequency point is to damage.

[0067] Iterate through all test frequency points to find the one that makes The frequency points that reach local maxima are defined as the sensitive feature frequencies. Since the total number of characteristic frequency points is only 15 to 20, the frequency resolution is limited. In order to capture changes in sensitive frequencies in a timely manner, this embodiment uses each cycle for calculation and updating to ensure that no damage feature information is missed.

[0068] The specific identification steps are as follows: Step 1: Immediately after each loop is completed, calculate the cumulative rate of change at each frequency point. ; Step 2: Find the loop that makes The frequency points that reach local maxima are used as current candidate sensitive frequencies; Step 3: Record the candidate sensitivity frequency and compare it with the candidate results from the previous two loops; Step 4: If the frequency remains unchanged for three consecutive cycles, it is then finally identified as the sensitive characteristic frequency. ; Step 5: If the candidate frequency drifts, then take the frequency that has appeared most frequently up to the current loop as the sensitive feature frequency. .

[0069] The purpose of this multi-cycle verification mechanism is to avoid misidentification caused by fluctuations in a single measurement. By employing stability checks for three consecutive cycles, random frequency drift caused by measurement noise, temperature fluctuations, etc., can be effectively filtered out, ensuring that the identified sensitive frequencies are representative and stable.

[0070] Subsequently, the impedance magnitude at this frequency was monitored in real time. This is used for subsequent dual-criteria invalidation determination.

[0071] S330. Dual-criteria failure determination: The reason for setting two independent failure criteria and adopting dual-criteria logic is that the comprehensive damage index can reflect the overall degradation trend, but may not be able to capture sudden local failures; while the sudden change in the impedance modulus of a single frequency point corresponds to local physical failures; the two complement each other and can improve the reliability of failure determination.

[0072] Criterion 1: The dynamic damage accumulation index reaches the first threshold. When the first threshold... When the value reaches 0.7, the appliance has usually entered the accelerated degradation stage, with less than 30% of the total lifespan remaining. Further testing is unnecessary, and the damage state at this threshold point has good repeatability.

[0073] Criterion 2: The percentage by which the impedance magnitude at the sensitive characteristic frequency point exceeds the initial value. ; First, define the percentage threshold. This indicates the percentage by which the current impedance magnitude is allowed to exceed the initial value, expressed as a percentage. This indicates that the current impedance magnitude is allowed to increase by 200% compared to the initial value.

[0074] Based on this, This indicates the number of times the current impedance magnitude is allowed to reach the initial value.

[0075] In this criterion The meaning is: the current impedance magnitude has reached the initial value. times. Specifically, when hour, This means that the current impedance magnitude is allowed to reach three times the initial value. This sudden change in magnitude usually corresponds to the complete break of a conductive path inside the electrical appliance.

[0076] This embodiment takes The corresponding conditions are: (14); Right now: (15); Two criteria are monitored independently to determine the following conditions in real time: (16); When either of the above two conditions is met for the first time, the one that occurs first shall be used to determine that the electrical appliance under test has reached the critical failure point.

[0077] S340. Record the number of failure cycles: Once the damage assessor determines that the critical failure point has been reached, it immediately sends an emergency interrupt command to the temperature change test chamber controller to stop the temperature cycle test. The number of cycles at this point is automatically recorded, which is the critical failure cycle number for the sample.

[0078] If the test chamber runs continuously for more than the preset maximum number of cycles without triggering failure, the system automatically terminates the test, marks the sample as having an ultra-long lifespan, and excludes it from the fitting of lifespan models for subsequent batches. This design avoids ultra-long lifespan samples consuming excessive test resources.

[0079] Through the above S310 to S340, the critical failure point of the tested electrical appliance is actively and accurately determined, and key failure data is recorded, providing a basis for subsequent batch reliability assessment and individual remaining life prediction.

[0080] In a method for testing extreme temperature shock damage to automotive electrical components, S400 aims to establish a statistical lifetime model based on the failure data of batch samples and extrapolate the damage evolution trajectory of individual samples, thereby achieving dual outputs of batch reliability assessment and individual remaining life prediction.

[0081] Specific steps are as follows Figure 5 As shown: S410. Batch lifespan statistics: Repeat steps S100 to S300 to test at least three samples of the same batch and model. During the test of each sample, when the dual criterion condition in S330 is met for the first time, the system automatically records the critical failure cycle count for that sample. After the test is completed, the critical failure cycle counts for the three samples in the batch are obtained and recorded as the failure counts for the first sample, the second sample, and the third sample, respectively.

[0082] To assess the overall reliability level of this batch of products, two simple and practical statistical indicators need to be calculated.

[0083] The first metric is mean lifetime, which is the arithmetic mean of the number of critical failure cycles for the three samples. Mean lifetime represents the typical lifetime level of this batch of products and can be used as a benchmark for comparison with similar products.

[0084] The second metric is minimum lifetime, which is the minimum number of critical failure cycles among the three samples. This metric reflects the worst-case scenario for this batch of products and is suitable for conservative estimations and determining the lower limit of reliability.

[0085] The reason only three samples are required, rather than more, is that the core of this method lies in the online prediction of individual samples; batch statistics are only used to provide a reference baseline. Three samples are sufficient to eliminate the influence of chance, while avoiding excessive testing costs.

[0086] In practical applications, if higher precision is required, the number of samples can be increased to 5 or 10, and the formula for calculating the average lifespan can be adjusted accordingly.

[0087] S420. Determine the starting point for individual prediction: For each individual sample in the batch, its dynamic damage accumulation index is monitored in real time during the test. The dynamic damage accumulation index is defined and calculated by formula (12). It is a comprehensive index of cumulative damage from the initial state to the current cycle. The value range starts from 0 and gradually increases with the increase of the number of temperature shock cycles.

[0088] In S330, when the index reaches 0.7, the sample is considered to have reached the critical failure point and the test is terminated. However, to achieve remaining lifetime prediction, extrapolation must begin before the sample fails. Therefore, an early warning threshold below 0.7 needs to be set as the starting point for prediction.

[0089] In this embodiment, the early warning threshold is set to 0.4. This value is based on extensive preliminary experimental observations, which show that when the dynamic damage accumulation index reaches 0.4, the sample's damage evolution has transitioned from an initial slow growth phase to a stable expansion phase. During the stable expansion phase, the damage growth rate remains essentially constant, and the damage accumulation index exhibits a good linear relationship with the number of cycles. Linear extrapolation at this point provides good prediction accuracy and stability.

[0090] Meanwhile, the threshold of 0.4 leaves a margin of 0.3 from the failure threshold of 0.7, meaning that the sample has a sufficiently long remaining lifespan for prediction, providing meaningful early warnings for users. If the threshold is set too low (e.g., 0.2), the sample is still in the early stages of damage, with unclear linear characteristics and large extrapolation errors; if the threshold is set too high (e.g., 0.6), the remaining lifespan is too short, and the warning is not very meaningful.

[0091] When the dynamic damage accumulation index first reaches 0.4, the number of cycles at this point is automatically recorded and designated as the warning start cycle number. Warning start point cycle count It represents the number of cycles a sample undergoes from its initial state until the damage index reaches 0.4, and serves as a time reference point for individual prediction.

[0092] In one possible implementation, if the sample triggers criterion two in S330 (i.e., a sudden change in the impedance modulus at the sensitive characteristic frequency point) before the dynamic damage accumulation index reaches 0.4, it indicates that the sample has experienced a sudden local failure. This failure mode cannot be predicted in advance by damage accumulation trends. In this case, the subsequent prediction start point recording work in S420 and the remaining lifetime prediction steps in S430 and S440 are skipped. Only the critical failure cycle number of the sample is recorded as required by S340, and the batch sample testing procedure continues.

[0093] S430. Linear extrapolation prediction of remaining lifetime: Once the prediction starting point is determined, the dynamic damage accumulation index data near the warning starting point needs to be used to estimate the damage growth rate, and then extrapolate the remaining number of cycles required to reach the failure threshold.

[0094] Due to the dynamic damage accumulation index during the stable expansion phase The relationship between the temperature shock cycle number and the actual temperature is approximately linear; therefore, a linear model can be used for fitting. The linear model takes the following form: (17); in, The growth rate represents the increase in the damage index with each additional temperature shock cycle, and is a key parameter for predicting remaining lifespan. The intercept represents the offset of the linear model on the vertical axis.

[0095] To calculate the growth rate It is necessary to select data points corresponding to three consecutive temperature shock cycles near the warning starting point. These three data points are based on the number of cycles from the warning starting point. The selection is based on the following criteria: Point A: The previous cycle before the warning starting point, with a cycle count of [number missing]. The corresponding damage index value is ; Point B: Warning starting point for this cycle, number of cycles is [number missing]. The corresponding damage index value is ; Point C: The next loop after the warning start point, the number of loops is... The corresponding damage index value is ; The number of cycles and damage index values ​​corresponding to the above three data points have been recorded and stored by the system during previous testing.

[0096] The reason for using three points instead of two is that the intermediate point B can be used to verify whether the linear hypothesis holds. The reason for using three points instead of more is that in real-time testing, only one subsequent point is available after the warning starting point; at the same time, three points are sufficient to estimate the slope, without waiting for more data and causing prediction lag.

[0097] Using these three data points, the average growth rate is calculated using the difference between the first and last data points. The specific method is as follows: subtract the damage index of point A from the damage index of point C, and then divide by the difference in the number of cycles between point C and point A. (18); Gain growth rate Then, the first threshold can be calculated. The required remaining number of cycles is 0.3 for the damage increment from the warning starting point (damage index of 0.4) to the failure threshold (0.7). This damage increment of 0.3 is divided by the growth rate. This means obtaining the number of cycles required from the start of the warning to its failure. : (19); Finally, predict remaining lifespan. This refers to the number of cycles required from the start of the warning to its failure: (20); The output is based on the number of temperature shock cycles, rounded down to the nearest integer. If the calculated number is 8.3 cycles, it is rounded down to 8 (a conservative estimate); if the user requires a more conservative warning, it can be rounded up to 9. This embodiment uses rounding down for a more conservative output.

[0098] S440. Prediction Result Verification and Output: To ensure the reliability of the prediction results, the growth rate calculated in S430 needs to be verified before output. and predicting remaining lifespan Three verifications were conducted. These three verifications examined the reasonableness of the growth rate, the reasonableness of the linear assumption, and the reasonableness of the remaining lifespan, checking the reliability of the prediction results item by item.

[0099] The first verification is the rationality of the growth rate: growth rate It is the core parameter for predicting remaining lifespan, and its value must meet two basic conditions.

[0100] The first condition is that the growth rate must be positive, that is... This indicates that the damage index increases with the number of temperature shock cycles, consistent with the physical law of thermal fatigue damage accumulation. If the calculated aa is zero or negative, it means that the damage index does not show an increasing trend in the three data points near the warning starting point. This may be due to excessive measurement noise or the sample having entered an abnormal state. In this case, linear extrapolation is not applicable, and the system does not output a predicted value.

[0101] The second condition is that the growth rate cannot be too low. In this embodiment, the lower limit of the growth rate is set at 0.01, that is... The physical meaning of this threshold is that the damage index increases by at least 1% for every 100 temperature shock cycles. If the growth rate is less than 0.01, it means that the damage growth is extremely slow, and the error of linear extrapolation will be amplified.

[0102] When the growth rate is less than 0.01, the system does not output a specific predicted remaining lifetime value, but instead outputs a prompt message: "Damage growth is slow; continued monitoring is recommended." At this time, the system continues to execute the S100 to S300 test procedures normally, waiting for more data to accumulate before re-predicting.

[0103] The second verification is the validity of the linear hypothesis: The premise of linear extrapolation is that near the warning starting point, the dynamic damage accumulation index and the number of cycles approximately satisfy a linear relationship. To verify whether this hypothesis holds, we use the intermediate point B at the warning starting point for testing.

[0104] The specific verification method is as follows: First, based on point A and the growth rate... Calculate the linear interpolation result at point B. The basic principle of linear interpolation is: on a straight line, given the value and slope of a previous point, the value of the next point can be calculated. The formula is: (twenty one); The meaning of this formula is: assuming the damage index increases linearly with the number of cycles, then the increment from point A to point B should be exactly equal to the growth rate. (Because adjacent cycles differ by 1). Therefore, the predicted value of point B is equal to the measured value of point A plus the growth rate.

[0105] Then, calculate the actual damage index at point B. Linear interpolation result with point B absolute value of the deviation between : (twenty two); This deviation reflects the degree of discrepancy between the actual data and the linear model. The smaller the deviation, the more valid the linear assumption is; the larger the deviation, the more significant the nonlinear characteristics of the damage evolution.

[0106] In this embodiment, the allowable upper limit for deviation is set at 0.05. This threshold is based on the fact that during the stable expansion phase, the measurement accuracy of the damage index is generally within ±0.02. A tolerance deviation of 0.05 is equivalent to an allowable range of ±2.5 times the standard deviation, which can cover most normal situations. If the deviation exceeds 0.05, it indicates that among the three data points near the warning starting point, the actual value of the middle point deviates significantly from the linear trend.

[0107] when deviation The system still outputs a predicted remaining lifespan, but with an additional message: "Data fluctuates significantly; predictions are for reference only." This provides users with information while also reminding them to use the system with caution.

[0108] when deviation If the linear assumption holds, the prediction result will be output in the normal manner.

[0109] The third verification is the verification of the reasonableness of the remaining lifespan: Predicting remaining lifespan The numerical values ​​themselves also need to be checked for reasonableness. This embodiment sets the lower limit for predicting remaining lifetime to 5 cycles, i.e. .

[0110] The rationale for setting this threshold is that when the predicted remaining lifetime is less than 5 cycles, it means the sample is extremely close to failure, with a very short remaining lifetime. In this case, continuing to predict is not very meaningful; what is more important is to remind the user to pay immediate attention to the sample's condition. At the same time, extremely short predicted values ​​are also easily affected by measurement noise, resulting in low reliability.

[0111] When the predicted remaining lifetime is less than 5 cycles, the output is: "Imminent failure, immediate check recommended," along with the specific predicted cycle count. For example, if the calculated... If the output is: "About to fail, immediate check recommended, remaining lifespan predicted: 3 cycles".

[0112] Comprehensive judgment and output of verification results: After the three verifications are completed, the output format will be determined based on the verification results: Scenario 1: All three verifications pass ( , , The function directly outputs the predicted remaining lifetime in the format: Predicted remaining lifetime: XX cycles.

[0113] Scenario 2: The first verification failed ( Instead of outputting predicted values, it only outputs a prompt message: Damage growth is slow, and continued monitoring is recommended.

[0114] Scenario 3: The first verification passes, but the second verification fails. The system outputs a predicted remaining lifetime, along with the message: "Data fluctuates significantly; the prediction result is for reference only." The output format is: Predicted remaining lifetime: XX cycles (Data fluctuates significantly; for reference only).

[0115] Scenario 4: Items 1 and 2 pass, but item 3 fails. The system outputs a prompt message and a predicted value, in the format: "About to expire, it is recommended to check immediately, predicted remaining lifespan: XX cycles".

[0116] Scenario 5: The first verification passes, but the second and third verifications fail. The system outputs the data in the format of Scenario 4 (with a priority message indicating that the data is about to expire), along with a data fluctuation warning.

[0117] Example 2: Please refer to Figures 1 to 5 As another embodiment of the present invention, this embodiment provides an automotive electrical extreme temperature shock damage testing system, comprising: The programmable rapid temperature change test chamber is used to house automotive electrical components fixed on test fixtures. It automatically completes periodic high and low temperature switching according to preset temperature parameters of -40℃ to +125℃, temperature change rate of 30℃ / min, and extreme value dwell time of 30min, applying cyclic extreme temperature shock thermal fatigue stress to the electrical components under test. It can receive shutdown commands issued by the damage assessor and immediately terminate the temperature cycle when the sample is determined to be at critical failure.

[0118] The high-precision impedance analyzer is connected in parallel to the external port of the automotive electrical system under test via a dedicated test cable. At the end of each high and low temperature dwell phase, 15 to 20 characteristic frequency points are selected logarithmically in the 1kHz to 10MHz frequency band to carry out frequency sweep testing. The impedance magnitude and phase angle of each frequency point are collected to generate the original impedance spectrum sequence in the form of complex impedance, and the measurement data is transmitted to the digital signal processor in real time.

[0119] The digital signal processor is electrically connected to the output of a high-precision impedance analyzer to receive full-cycle impedance spectrum data. Based on the equivalent circuit topology of contact resistance, constant phase angle element, polarization resistance connected in parallel and then series lead inductance, the Levenberg-Marquardt nonlinear least squares algorithm is used to complete the impedance spectrum fitting, extracting four key electrical parameters: contact resistance, constant phase angle element amplitude parameter, polarization resistance, and lead inductance. The normalized rate of change of each parameter is calculated based on the initial parameters at room temperature, and after three-point moving average filtering, they are combined to generate a four-dimensional cumulative damage feature vector, which is then transmitted to the damage assessor.

[0120] The damage assessor's input interface connects to a digital signal processor. On one hand, it calculates the dynamic damage accumulation index in real time by weighting and summing the four-dimensional damage feature vector using pre-calibrated weighting coefficients. On the other hand, it calculates the cumulative rate of change of impedance modulus at each frequency in cycles, and filters out sensitive feature frequencies after three consecutive cycles of stability verification. A dual-threshold failure criterion is configured: either the dynamic damage accumulation index ≥ 0.7 or the impedance modulus at the sensitive feature frequency ≥ 3 times the initial value. If either condition is triggered, the sample is deemed to have reached critical failure, and a shutdown signal is immediately sent to the programmable rapid temperature change test chamber. The current critical failure cycle count is simultaneously stored, and the damage index change trajectory and failure cycle count are synchronously uploaded to the life prediction unit. If the sample experiences a sudden impedance increase failure before reaching the warning threshold of 0.4, the sudden local failure data is separately marked.

[0121] The life prediction unit connects to the data output of the damage assessor and includes a batch statistics module, a remaining life calculation module, and a report generation module. (1) Batch statistics module: Summarize the critical failure cycles of ≥3 samples in the same batch, calculate the arithmetic mean as the average lifetime of the batch, and the minimum value as the minimum lifetime of the batch to form the batch reliability index. (2) Remaining life calculation module: Real-time monitoring of dynamic damage accumulation index. When the index reaches the warning threshold of 0.4 for the first time, the number of warning cycles is locked. Three sets of damage index data are selected: the previous warning, the current warning, and the next warning. The damage growth rate is calculated by the first and last difference method. The remaining cycle life is predicted by linear extrapolation with 0.3 / growth rate. The three checks are completed in sequence: damage growth rate ≥ 0.01, linear interpolation deviation ≤ 0.05, and predicted life ≥ 5 cycles. The life value and corresponding prompt text are output differently according to the check results. (3) Report generation module: After all samples in a single batch have been tested, the module automatically integrates the failure cycle count, average / minimum life of the batch, and remaining life prediction data of each sample, and generates a standardized test report with one click.

[0122] The sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0123] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for testing extreme temperature shock damage to automotive electrical components, characterized in that, Includes the following steps: S100. Place the automotive electrical appliance under test in a programmable temperature change test chamber, and control the temperature of the test chamber to cycle between the lowest temperature and the highest temperature. At the end of each temperature dwell stage, use an impedance analyzer connected in parallel with the port of the electrical appliance under test to scan at multiple characteristic frequency points in a preset frequency band and collect the impedance spectrum sequence that changes with the number of cycles. S200. Perform equivalent circuit fitting on the impedance spectrum for each cycle number, extract at least two key electrical parameters, and calculate the normalized rate of change of each key electrical parameter relative to the initial value, and combine them into a cumulative damage feature vector. S300. Calculate the dynamic damage accumulation index based on the cumulative damage feature vector and monitor the impedance modulus at at least one characteristic frequency point; when the dynamic damage accumulation index reaches the first threshold, or the impedance modulus at the characteristic frequency point exceeds a predetermined percentage of the initial value, determine that the electrical appliance under test has reached the critical failure point, trigger the test termination and record the number of critical failure cycles of the sample. S400. Repeat steps S100 to S300 to obtain the critical failure cycle counts of multiple samples in the same batch. Use the critical failure cycle counts of multiple samples to perform batch lifetime statistics and output the batch reliability index. For a single test sample in the batch, after the dynamic damage accumulation index exceeds the preset early warning threshold, use the measured dynamic damage accumulation index trajectory of the sample to perform linear extrapolation fitting, solve for the predicted remaining lifetime of the sample, and output the prediction results after verification.

2. The method for testing extreme temperature shock damage to automotive electrical systems according to claim 1, characterized in that, The specific steps for acquiring the impedance spectrum sequence that varies with the number of cycles include: The automotive electrical appliance under test is fixed on the test fixture, and a full-band frequency sweep measurement is performed on the electrical port using an impedance analyzer at room temperature. The initial impedance spectrum is recorded as the reference value. Set the temperature cycling parameters: the lowest temperature is -40℃, the highest temperature is 125℃, the temperature change rate is 30℃ / min, and the dwell time at each extreme point is 30 minutes. At the end of each temperature dwell phase, the impedance analyzer is triggered to select 15 to 20 characteristic frequency points in the logarithmic coordinate system, covering the frequency band from 1 kHz to 10 MHz for frequency sweep measurement, and the impedance magnitude and phase angle at each frequency point are recorded to form complex impedance information. The collected data is organized according to the number of temperature shock cycles to form an impedance spectrum sequence that varies with the number of cycles and is then stored.

3. The method for testing extreme temperature shock damage to automotive electrical systems according to claim 1, characterized in that, The specific steps for extracting the cumulative damage feature vector include: Construct an equivalent circuit topology. The equivalent circuit topology is as follows: the contact resistor, the constant phase angle element, and the polarization resistor are connected in parallel, and then connected in series with the lead inductor. The Levenberg-Marquardt nonlinear least squares algorithm is used to fit the measured impedance spectrum to the equivalent circuit topology to obtain the key electrical parameters under the current cycle, including contact resistance, constant phase angle element amplitude parameters, polarization resistance and lead inductance. Using the contact resistance, constant phase angle element amplitude parameters, polarization resistance, and lead inductance at the initial cycle as a reference, calculate the normalized rate of change of each parameter; The four normalized rates of change are combined into a cumulative damage feature vector, and the rate of change sequence of each parameter is filtered using a three-point moving average.

4. The method for testing automotive electrical extreme temperature shock damage according to claim 1, characterized in that, The specific steps for determining the critical failure point include: The dynamic damage accumulation index is calculated by weighted summation of the four normalized rates of change in the cumulative damage feature vector. The sum of the weight coefficients of each rate of change is one and none of them are less than zero. The weight coefficients are determined by pre-experiment calibration. Calculate the cumulative rate of change of impedance modulus at each test frequency point, find the frequency point that makes the cumulative rate of change reach a local maximum as a candidate sensitive characteristic frequency, calculate and update in each loop, and finally confirm the sensitive characteristic frequency after three consecutive loops. A dual-criteria failure determination condition is set. Criterion 1 is that the dynamic damage accumulation index reaches 0.7, and criterion 2 is that the impedance modulus at the sensitive characteristic frequency point reaches or exceeds 3 times the initial value. When either criterion is met for the first time, the electrical appliance under test is determined to have reached the critical failure point. Once the critical failure point is determined, the test is stopped immediately, and the current cycle count is recorded as the critical failure cycle count.

5. The method for testing extreme temperature shock damage to automotive electrical systems according to claim 1, characterized in that, The specific method for batch lifetime statistics is as follows: test at least three samples in the same batch, obtain the critical failure cycle number of each sample, calculate the arithmetic mean of the critical failure cycle number of all samples as the average lifetime, take the minimum value as the minimum lifetime, and output the average lifetime and minimum lifetime as batch reliability indicators.

6. The method for testing extreme temperature shock damage to automotive electrical systems according to claim 1, characterized in that, The specific steps for obtaining the predicted remaining life of the sample by linear extrapolation fitting using the measured dynamic damage accumulation index trajectory of the sample include: Real-time monitoring of the dynamic damage accumulation index; when the dynamic damage accumulation index reaches 0.4 for the first time, record the number of cycles at this time as the warning start cycle number. Based on the number of cycles from the warning start point, dynamic damage accumulation index data points from three consecutive cycles—the cycle before the warning start point, the current cycle from the warning start point, and the cycle after the warning start point—are selected. The damage growth rate is calculated using the first and last point difference method, which is to subtract the damage index from the cycle before the warning start point from the damage index of the cycle after the warning start point, and then divide by 2. Divide 0.3 by the damage growth rate to obtain the remaining number of cycles required from the warning start point to failure, and round it to the nearest integer as the predicted remaining lifetime output.

7. The method for testing extreme temperature shock damage to automotive electrical systems according to claim 1, characterized in that, The specific steps for verifying and outputting the prediction results include: The first verification step: Check whether the damage growth rate reaches or exceeds 0.

01. If it does not reach 0.01, no predicted value will be output, and a message indicating that the damage growth is slow and continued monitoring is recommended will be output. The second verification step is to check the rationality of the linear assumption, that is, to calculate the deviation between the actual damage index of the current cycle at the warning starting point and the linear interpolation result. If the deviation exceeds 0.05, the additional data fluctuation will be large when outputting the predicted remaining lifespan, and the prediction result will be for reference only. The third verification step: Check whether the predicted remaining lifetime has reached or exceeded 5 cycles. If it is less than 5 cycles, output a prompt message indicating that it is about to fail and recommends checking immediately, and output the predicted value at the same time. Based on the combined results of the three verifications, the final output format is determined.

8. The method for testing extreme temperature shock damage to automotive electrical systems according to claim 1, characterized in that: If the test sample has triggered the criterion that the impedance modulus at the sensitive characteristic frequency point changes to more than three times the initial value before the dynamic damage accumulation index reaches 0.4, it is determined to be a sudden local failure, and the individual remaining life prediction step is skipped, and only the number of critical failure cycles is recorded.

9. A testing system for extreme temperature shock damage to automotive electrical components, used to implement the method described in any one of claims 1 to 8, characterized in that, include: The programmable rapid temperature change test chamber is used to house the automotive electrical components under test and apply periodic extreme temperature shocks according to a preset temperature cycling curve. The high-precision impedance analyzer is connected in parallel with the external port of the automotive electrical appliance under test via a test cable. It is used to perform wideband sweep frequency measurement on the electrical port at the end of each temperature dwell phase and output the impedance magnitude and phase angle at each frequency point. A digital signal processor, connected to the output of an impedance analyzer, is used to receive impedance spectrum sequences, perform equivalent circuit fitting, extract key electrical parameters, calculate normalized rate of change, and generate cumulative damage feature vectors. The damage evaluator, connected to the output of the digital signal processor, is used to calculate the dynamic damage accumulation index, identify sensitive characteristic frequencies, perform dual-criteria failure determination, and send a test termination signal to the test chamber when the determination reaches the critical failure point. The life prediction unit is connected to the output of the damage evaluator and is used to receive the critical failure cycle count and dynamic damage accumulation index trajectory to perform batch life statistics and individual remaining life linear extrapolation prediction.

10. The automotive electrical extreme temperature shock damage testing system according to claim 9, characterized in that: The lifetime prediction unit is also equipped with a report generation module, which automatically generates a test report after testing at least three samples, including the critical failure cycle count, batch average lifetime, batch minimum lifetime, and predicted remaining lifetime of each sample.

Images