An online evaluation method for aging state of new energy vehicle wire harness based on impedance spectrum analysis

Abstract

The application belongs to the technical field of new energy vehicle electrical detection and health diagnosis, and discloses a new energy vehicle wire harness aging state online evaluation method based on impedance spectrum analysis. When the target wire harness loop is turned on, a diagnosis excitation signal is injected, and the voltage and current responses of the excitation end and the other end measurement port, the switch state, the wire harness temperature, the loop current and the state of charge signal are synchronously collected. According to the switch state, the transient shielding interval is generated, and the effective sampling points are extracted. The input impedance spectrum and the transmission impedance spectrum are calculated, and the segmented mapping is carried out in combination with the wire harness topology parameters and the segmented propagation time delay parameters. Then, the working condition normalization is carried out based on the working condition reference library, and the segmented aging residual characteristics are obtained. The degradation evaluation model is input, and the segmented aging index, the aging grade and the evaluation confidence are output.

Description

Technical Field

[0001] This invention belongs to the field of electrical testing and health diagnosis technology for new energy vehicles, specifically a method for online assessment of the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis. Background Technology

[0002] High-voltage wiring harnesses in new energy vehicles are used to connect power batteries, electric drive systems and high-voltage accessories, and are responsible for the transmission of electrical energy and control signals. During long-term vehicle operation, the wiring harnesses and connectors are continuously subjected to temperature rise cycles, load fluctuations and mechanical vibrations. The conductor resistance, contact resistance and insulation status will change with the service time, and the circuit impedance and transmission characteristics will drift accordingly. Such changes are usually gradual and condition-dependent, and are not easy to identify directly through a single electrical parameter in the early stages.

[0003] There are two main methods for judging the condition of existing wiring harnesses. One is offline detection when the vehicle is stopped, such as continuity, insulation and withstand voltage testing, which can detect obvious faults, but it is difficult to cover the dynamic degradation during the operation of the whole vehicle. The other is on-board online monitoring, which usually alarms based on single-ended voltage or current thresholds. This method is simple to implement, but the measurement results are easily affected by transients under disturbances such as inverter switching, relay activation and deactivation. At the same time, changes in temperature, loop current and state of charge will cause spectral response shifts, and the reflection and transmission responses of multi-branch wiring harnesses will also have path superposition, resulting in a decrease in the comparability of data of the same segment at different times.

[0004] Therefore, the core deficiency of existing technologies lies in the difficulty of stably distinguishing between reversible response changes caused by changes in operating conditions and irreversible response changes caused by aging under online operating conditions, and forming repeatable and traceable segmented evaluation results under a unified data standard. Summary of the Invention

[0005] The purpose of this invention is to provide an online assessment method for the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an online assessment method for the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis, comprising: When the vehicle's high-voltage system is powered on and the target harness circuit is conducting, the method injects a diagnostic excitation signal into the target harness circuit through the excitation end, and simultaneously acquires the excitation end voltage response signal, excitation end current response signal, the voltage response signal of the measurement port at the other end of the target harness circuit, the current response signal of the measurement port at the other end of the target harness circuit, as well as the vehicle switch status signal, harness temperature signal, circuit current signal and charge status signal under a unified sampling time base.

[0007] In the data processing, a transient shielding interval is first constructed based on the vehicle switch status signal, and the sampling points corresponding to the switch transient are removed. Then, the input impedance spectrum of the excitation end and the transmission impedance spectrum of the measurement port from the excitation end to the other end of the target wiring harness circuit are calculated using the effective sampling points. Subsequently, combined with the pre-stored wiring harness topology parameters and segmented propagation delay parameters, the input impedance spectrum and the transmission impedance spectrum are segmented and mapped to obtain the spectral characteristics of each wiring harness segment. Then, the operating condition reference library is called to normalize the spectral characteristics of each segment to form segmented compensation characteristics under a unified reference operating condition. The difference between the segmented compensation characteristics and the corresponding segmented reference spectrum is then calculated to obtain the segmented aging residual characteristics. Finally, the segmented aging residual characteristics are input into the degradation evaluation model to obtain the wiring harness segmented aging evaluation results.

[0008] In the above technical solution, transient shielding is used to reduce the interference of switching action on impedance spectrum calculation. Synchronous acquisition of the excitation end and the measurement port signal at the other end of the target wire harness circuit is used to simultaneously characterize the circuit input characteristics and transmission characteristics. Segment mapping is used to locate the spectrum change to a specific wire harness segment. Operating condition normalization is used to reduce the coupling effect caused by changes in temperature, circuit current and state of charge. Through the above collaborative processing, the segment characteristics obtained at different sampling times can be compared and support the online and continuous evaluation of the wire harness aging status.

[0009] Furthermore, based on the above online evaluation process, the diagnostic excitation signal adopts a multi-sine combination signal, a linear sweep frequency signal, or a pseudo-random binary sequence signal. The upper limit of the excitation amplitude is determined in the calibration stage. The calibration constraints are the service circuit voltage fluctuation threshold and the communication bit error rate threshold. Through this amplitude constraint, a discriminable impedance spectrum response can be obtained without changing the working state of the service circuit.

[0010] Synchronous acquisition is performed within the sampling permission window, which is generated by the vehicle controller based on the switch status signal and the loop current ripple threshold. The loop current ripple threshold is obtained by statistical analysis of historical operating data during the calibration phase. By controlling the permission window, the calculation process can be avoided during periods when the loop current ripple exceeds the threshold, thus ensuring the comparability of spectral data at different sampling times.

[0011] To suppress transient interference during switching, a transient shielding interval is constructed based on the inverter switching edge time, relay activation time, and relay release time. Protection durations are set before and after each time point, and the protection durations are determined by calibration tests. Impedance spectrum calculations are performed after shielding is completed, which can reduce the impact of transient spikes on spectrum estimation.

[0012] The input impedance spectrum and transmission impedance spectrum are calculated based on the same set of excitation frequencies. The input impedance spectrum is obtained by the ratio of the complex components of the excitation terminal voltage and the complex components of the excitation terminal current at each excitation frequency. The transmission impedance spectrum is obtained by the ratio of the complex components of the voltage at the measurement port at the other end of the target harness loop and the complex components of the excitation terminal current at each excitation frequency. The phase dispersion and amplitude dispersion are calculated for repeated sampling results, and frequency points with coherence below the calibration threshold are removed. Through processing with the same set of frequencies and consistency criteria, the calculation basis of the two types of spectra can be unified and frequency point misjudgment caused by noise can be reduced.

[0013] The harness topology parameters include the branch connection matrix, segment length parameters, segment propagation speed parameters, and connection node reflection coefficient parameters. During segment mapping, the branch connection matrix is ​​used to construct a multipath propagation equation, and the segment propagation delay parameters are used as solution constraints to separate overlapping responses, outputting node identifiers and segment identifiers. Through this mapping process, spectral changes can be located to specific harness segments or connection nodes.

[0014] The operating condition benchmark library is established by collecting data from the same vehicle model and wiring harness number when the wiring harness is newly installed. The data collection dimensions include temperature partition, loop current partition, and state of charge partition. After the wiring harness or connector is replaced, the benchmark spectrum version of the corresponding segment is reconstructed. Through version management, the benchmark data can be guaranteed to be consistent with the current hardware status.

[0015] The operating condition normalization adopts a partitioned interpolation method, which calculates the compensation amount by linear interpolation within adjacent operating condition partitions. The normalization input is the wire harness temperature measurement value, loop current measurement value, and state of charge measurement value at the sampling time. The output is the segmented compensation feature under a unified reference operating condition. This processing is used to weaken the coupling effect of operating condition fluctuations on aging judgment.

[0016] The degradation assessment model is trained using monotonic constraints. The training samples are historical measurement sequences of the corresponding segments in the operating condition benchmark library. Monotonic constraints apply to the relationship between the aging index and the cumulative thermal stress and cumulative electrical stress of the same segment. The cumulative thermal stress is obtained by temperature-time integration, and the cumulative electrical stress is obtained by current-time integration. The aging assessment results of the harness segments include the aging index, aging level, and assessment confidence level. The aging level is divided into three levels: normal, warning, and deterioration. The thresholds for the three levels are determined by the statistical distribution of the operating condition benchmark library. Through this assessment mechanism, continuous determination and consistent classification of the aging status of segments can be achieved.

[0017] The beneficial effects of this invention are as follows: 1. This invention sets a sampling permission window during vehicle operation and constructs a transient shielding interval based on the switching state. It synchronously collects the response signals of the excitation end and the other end of the target harness circuit measurement port under the same sampling time base. Then, it calculates the input impedance spectrum and transmission impedance spectrum on the same excitation frequency point set, so that the data entering the evaluation link has both time base consistency and frequency point consistency. It also suppresses the disturbances to spectrum estimation caused by switching transients and high ripple periods in advance, thereby improving the comparability of segmented characteristics at different sampling times and the stability of online continuous evaluation.

[0018] 2. This invention establishes a topology-constrained multipath propagation equation by introducing a branch connection matrix, segment length parameters, segment propagation speed parameters, and connection node reflection coefficient parameters. It also separates overlapping responses by combining segment propagation delay parameters and outputs segment identifiers and node identifiers. Furthermore, it manages the operating condition benchmark library by temperature partitioning, loop current partitioning, and state of charge partitioning. During the operating condition normalization stage, it calculates compensation by linear interpolation of adjacent partitions, enabling spectral changes to be applied to specific harness segments and compared under a unified benchmark operating condition. This enhances positioning accuracy, version traceability, and cross-operating condition consistency.

[0019] 3. This invention models the aging residual characteristics after segmented compensation together with the corresponding segmented historical measurement sequences. It adopts a monotonically constrained training mechanism for the same segment to establish a monotonically undiminished correlation between the aging index and the cumulative thermal stress and cumulative electrical stress. Then, it combines the statistical distribution of the operating condition benchmark library to determine the three-level thresholds of normal, early warning, and deterioration. This allows the evaluation results to simultaneously output the aging index, aging level, and evaluation confidence. This transforms the aging judgment from a single-moment threshold judgment to a trend-based and graded judgment, thereby improving the stability of early warning, the executability of maintenance decisions, and the consistency of life-cycle management. Attached Figure Description

[0020] Figure 1 This is a flowchart of the online assessment method for the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis, as described in this invention. Figure 2 This is a flowchart of the data acquisition and preprocessing sub-process of the present invention; Figure 3 This is a flowchart of the normalization of operating conditions and aging assessment sub-process of the present invention. Detailed Implementation

[0021] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] like Figures 1 to 3 As shown, this embodiment of the invention provides an online assessment method for the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis. In this embodiment, the assessment object is the target wiring harness circuit in the high-voltage system of the vehicle. The target wiring harness circuit is equipped with an excitation end and a measurement port at the other end of the target wiring harness circuit. The wiring harness topology parameters and segmented propagation delay parameters are determined by the wiring harness design data and calibration data of this vehicle model. The operating condition reference library is established by collecting data from the wiring harness of the same vehicle model and the same wiring harness number under the condition of new installation, and is stored in temperature partition, circuit current partition, and charge state partition.

[0023] When the vehicle's high-voltage system is powered on and the target harness circuit is turned on, a diagnostic excitation signal is injected into the target harness circuit through the excitation terminal. The frequency set and amplitude upper limit of the diagnostic excitation signal are determined during the calibration stage. Subsequently, the excitation terminal voltage response signal, excitation terminal current response signal, voltage response signal of the measurement port at the other end of the target harness circuit, current response signal of the measurement port at the other end of the target harness circuit, vehicle switch status signal, harness temperature signal, circuit current signal, and state of charge signal are synchronously acquired under the same sampling time base.

[0024] After the data collection is completed, a transient shielding interval is generated based on the vehicle switch status signal. The transient shielding interval consists of the protection duration before and after the switch event. The protection duration is determined by the calibration test. Sampling points within the transient shielding interval are removed, and sampling points in non-transient shielding intervals are retained as valid sampling points.

[0025] Based on the effective sampling points, the input impedance spectrum of the excitation end and the transmission impedance spectrum from the excitation end to the measurement port at the other end of the target harness circuit are calculated according to the same set of excitation frequency points. The input impedance spectrum of the excitation end is obtained by the ratio of the complex components of the excitation end voltage to the complex components of the excitation end current, and the transmission impedance spectrum is obtained by the ratio of the complex components of the voltage at the measurement port at the other end of the target harness circuit to the complex components of the current at the excitation end.

[0026] After obtaining the two types of impedance spectra, the harness topology parameters and segmented propagation delay parameters are combined to perform segmented mapping, outputting segmented spectral features. Then, the operating condition reference library is called to perform operating condition normalization on the segmented spectral features. The operating condition normalization adopts linear interpolation of adjacent operating conditions. The input is the harness temperature, loop current and state of charge at the sampling time, and the output is the segmented compensation features under the unified reference operating condition.

[0027] The segmented compensation features are differentially divided with the reference spectrum of the corresponding segment to obtain the segmented aging residual features. Finally, the segmented aging residual features are input into the degradation assessment model to output the segmented aging assessment results of the wiring harness. The training samples of the degradation assessment model come from the historical measurement sequences of the corresponding segments in the working condition reference library. The aging level threshold is determined by statistical distribution. The segmented aging assessment results of the wiring harness include the segmented aging index, the segmented aging level, and the segmented assessment confidence level.

[0028] In this embodiment, the diagnostic excitation signal is generated by the diagnostic excitation module. The diagnostic excitation module is electrically connected to the excitation end of the target wiring harness circuit and shares the same sampling time base with the response acquisition module. The diagnostic excitation module supports three forms: multi-sine combination signal, linear sweep frequency signal, and pseudo-random binary sequence signal. All three signals are constrained by the upper limit of the excitation amplitude. The upper limit of the excitation amplitude is stored and retrieved according to vehicle model, wiring harness number, signal form, and operating condition partition.

[0029] Multi-sine combination signals are formed by superimposing discrete frequency points. Linear sweep frequency signals are determined by the start frequency, end frequency, and sweep duration. Pseudo-random binary sequence signals are determined by the symbol duration and sequence length. The selection of frequency points or equivalent frequency bands follows the frequency band avoidance rules: avoid the protection band where the service power supply switch frequency and its harmonic frequencies are located, and avoid the protection band where the vehicle communication link's working frequency band is located. The protection band width is determined during the calibration stage and written into the parameter table.

[0030] The upper limit of the excitation amplitude is determined during the calibration phase. The calibration constraints include the service loop voltage fluctuation threshold and the communication bit error rate threshold. The service loop voltage fluctuation is calculated as follows: within a calibration window, the maximum absolute deviation of the service loop sampled voltage relative to the window baseline voltage is calculated, and then divided by the baseline voltage to obtain the voltage fluctuation amount. The communication bit error rate is calculated as the ratio of the number of erroneous frames to the total number of frames within the same calibration window. The current excitation amplitude is deemed to pass only when the voltage fluctuation amount is not greater than the service loop voltage fluctuation threshold and the communication bit error rate is not greater than the communication bit error rate threshold.

[0031] During calibration, the process is performed sequentially within the set of operating condition zones consisting of temperature zone, loop current zone, and state of charge zone. Under each operating condition zone, the excitation amplitude is increased in a step-by-step manner for each diagnostic excitation signal form. For each level of amplitude, a fixed number of continuous calibration windows are collected and the above two constraint indicators are calculated. If either indicator exceeds the threshold, the increment is stopped, and the amplitude that passed the verification at the previous level is recorded as the upper limit of the excitation amplitude under that operating condition zone and that signal form. After all operating condition zones are completed, an excitation amplitude upper limit table is generated.

[0032] During online evaluation, the vehicle controller determines the current operating condition zone based on real-time harness temperature, loop current, and state of charge. It then reads the corresponding upper limit of excitation amplitude from the upper limit table as the amplitude control boundary for this diagnostic excitation. If there is no valid calibration record for the current operating condition zone, it reads the calibrated upper limit from adjacent operating condition zones of the same vehicle model, harness number, and signal type, and selects the smaller value as the alternative boundary. This calling rule ensures that the online injection process continuously meets the calibration constraints and maintains data consistency with subsequent transient shielding, impedance spectrum calculation, segmented mapping, and operating condition normalization processing.

[0033] In this embodiment, synchronous acquisition is performed only within the sampling permission window. The sampling permission window is generated in real time by the vehicle controller based on the switch status signal and the loop current ripple threshold. The loop current ripple threshold is determined by statistical analysis of the data collected during the calibration phase and stored separately according to temperature zone, loop current zone, and state of charge zone.

[0034] During the calibration phase, the target wiring harness circuit is in a newly installed and healthy state. The vehicle operates under various working conditions and collects switch status signals and circuit current signals. For each working condition zone, the data during the switch status change period is first removed, and then the circuit current ripple index is calculated within a fixed calculation window. The circuit current ripple index is the difference between the maximum and minimum values ​​of the circuit current within the calculation window. The ripple index of the same working condition zone is statistically distributed, and the circuit current ripple threshold corresponding to the working condition zone is determined according to the preset quantile rule. The fixed calculation window length, sampling period, preset quantile, protection duration, stability judgment duration, and minimum window duration are uniformly determined during the calibration phase and written into the parameter file.

[0035] During the online evaluation phase, the vehicle controller performs sampling permission window determination according to the control cycle.

[0036] First, read the current switch status signal. If a switch status change event is detected, close the sampling permission window and keep it closed for the protection duration.

[0037] Second, when there are no switching status change events and the protection period is not in effect, calculate the current circuit current ripple index.

[0038] Third, determine the operating condition zone based on the current harness temperature, loop current and state of charge, and read the corresponding loop current ripple threshold. If there is no threshold record for the current operating condition zone, determine the threshold by interpolation of the adjacent operating condition zones; if the interpolation conditions are not met, call the conservative threshold.

[0039] Fourth, when the current loop current ripple index is not greater than the corresponding threshold and the condition is continuously met for the stability judgment period, the sampling permission window is opened.

[0040] Fifth, after the sampling permission window is opened, synchronous acquisition is performed to collect the excitation end voltage response signal, the excitation end current response signal, the voltage response signal of the measurement port at the other end of the target harness circuit, and the current response signal of the measurement port at the other end of the target harness circuit, and write them into the timestamp.

[0041] Sixth, within the sampling permission window, if a switch state change event or loop current ripple index exceeds the threshold, the sampling permission window is immediately closed and subsequent sampling in the current window is terminated.

[0042] Data segments with a window duration shorter than the minimum window duration are not included in subsequent spectrum calculations. Data segments with a window duration equal to or greater than the minimum window duration are considered valid sampled data segments and are included in the input impedance spectrum and transmission impedance spectrum calculation steps. This ensures that the data calibrators in the calibration stage and the online evaluation stage are consistent, and that the synchronously sampled data meets the requirements of subsequent segmentation mapping and operating condition normalization processing.

[0043] In this embodiment, the transient shielding interval is used to eliminate transient disturbance sampling points caused by switching action. The transient shielding interval is determined by the inverter switching edge time, the relay activation time, and the relay release time. A protection duration is set before and after each event time. The event time is recorded by the vehicle controller under a unified time base, and the timestamp step size is consistent with the synchronous sampling period.

[0044] In practice, the vehicle controller first generates an event sequence, which includes inverter switching edge events, relay activation events, and relay release events. For each type of event, a forward protection duration and a backward protection duration are set. A shielding sub-interval is constructed using the forward protection duration and the backward protection duration. Then, the overlapping shielding sub-intervals are merged to obtain the transient shielding interval for the current cycle.

[0045] The protection duration is determined by calibration tests. Calibration tests are conducted in the new installation state of the same vehicle model and wiring harness number. Test conditions are organized according to temperature zone, loop current zone, and state of charge zone. Three types of events are triggered in each zone and two-end response data are collected. The average impedance value of the steady-state window before the event occurs is used as the baseline, and the impedance deviation relative to the baseline at each sampling time before and after the event is calculated.

[0046] The backward protection duration is determined according to the following rules: starting from the event time, search backward to the earliest moment when the impedance deviation is not greater than the allowable deviation threshold and remains stable for a continuous period of time. The time difference between the two is defined as the backward protection duration.

[0047] The forward protection duration is determined according to the following rule: the time difference between the event time and the moment when the impedance deviation first exceeds the allowable deviation threshold is defined as the forward protection duration.

[0048] After repeated tests under the same operating condition zone, the forward protection duration and backward protection duration are statistically analyzed separately, and the upper quantile value is written into the protection duration parameter table. The allowable deviation threshold and stability judgment duration are uniformly determined and fixed during the calibration stage.

[0049] During online operation, the vehicle controller determines the current operating condition zone based on real-time wiring harness temperature, loop current, and state of charge, and reads the corresponding forward protection duration and backward protection duration. If there are no direct parameter records for the current operating condition zone, the protection duration is determined by linear interpolation of adjacent operating condition zones. If the interpolation conditions are not met, the larger protection duration in the adjacent operating condition zones is taken as the backoff parameter. Subsequently, a transient shielded interval is generated based on the event sequence, and sampling points within the interval are eliminated. Only unshielded sampling points are retained as valid sampling points and passed to the input impedance spectrum and transmission impedance spectrum calculation steps. When the length of the unshielded sampling segment is less than the minimum effective window duration, the sampling segment does not participate in the spectrum calculation.

[0050] In this embodiment, the calculation of the input impedance spectrum and the transmission impedance spectrum is performed after the sampling permission window is opened and the transient shielding interval is eliminated. After the vehicle controller obtains the valid sampling points, it divides the valid sampling points into measurement windows of the same length. The measurement window length, window overlap rate and number of repeated windows are determined by the calibration stage and written into the parameter file. Each measurement window includes the excitation end voltage response signal, the excitation end current response signal, the voltage response signal of the measurement port at the other end of the target harness circuit and the current response signal of the measurement port at the other end of the target harness circuit. Each signal is recorded under a unified sampling time base and aligned with the reference phase of the diagnostic excitation signal.

[0051] To ensure consistency in the calculation methods for the two types of spectra, both the input impedance spectrum and the transmission impedance spectrum are calculated based on the same set of excitation frequencies. The set of excitation frequencies is determined during the calibration phase and stored in association with vehicle model, wiring harness number, and diagnostic excitation signal format. For multi-sine combination signals, the set of excitation frequencies consists of the frequencies of each sinusoidal component. For linear sweep signals, frequency points are extracted in fixed frequency steps within the sweep interval to form an equivalent set of frequencies. For pseudo-random binary sequence signals, the equivalent set of frequencies is determined based on the symbol duration. The fixed frequency step is determined during the calibration phase and written into the parameter file.

[0052] Within each measurement window, quadrature demodulation is performed on each frequency point in the excitation frequency set to obtain the complex voltage component and the complex current component. The input impedance spectrum is obtained by the ratio of the complex voltage component to the complex current component at each excitation frequency point. The transmission impedance spectrum is obtained by the ratio of the complex voltage component to the complex current component at the other end of the target harness loop at each excitation frequency point. The current response signal at the other end of the target harness loop is used for sampling quality verification. When its deviation from the current response signal at the excitation end exceeds the current consistency threshold, the corresponding measurement window is judged as an invalid window. The current consistency threshold is determined during the calibration stage and written into the parameter file.

[0053] To suppress noise and residual transient effects, a consistency determination is performed for repeated sampling at each frequency point. Within the same sampling permission window, multiple consecutive valid measurement windows are selected, and the complex values ​​of input impedance and transmission impedance are calculated respectively to form amplitude and phase sequences. The phase sequence is expanded before statistics. The amplitude dispersion is calculated as the ratio of amplitude standard deviation to amplitude mean, and the phase dispersion is calculated as the phase standard deviation. The coherence index is calculated based on the amplitude dispersion and phase dispersion. The weight of the coherence index is determined during the calibration stage and written into the parameter file.

[0054] The coherence threshold is determined during the calibration phase. The calibration data comes from the controlled operation data of the same vehicle model and wiring harness number under the condition of new installation. The data segment meets the sampling permission window conditions and completes the removal of transient shielding intervals. The coherence indicators of each working condition zone and each excitation frequency point are statistically analyzed, and the corresponding coherence threshold is determined according to the preset quantile rule. The preset quantile is determined during the calibration phase and written into the parameter file.

[0055] During online operation, the vehicle controller reads the corresponding coherence thresholds according to the current temperature zone, loop current zone, and state of charge zone. If the current operating condition zone is missing parameters, the threshold is determined by interpolation of adjacent operating condition zones. If the parameters of adjacent operating condition zones are missing or the zone span exceeds the allowable range, a stricter threshold backoff is adopted, that is, the higher threshold among the adjacent available thresholds is adopted. For each frequency point, if the coherence index is lower than the corresponding threshold, the frequency point is simultaneously removed from the input impedance spectrum and transmission impedance spectrum. If the number of effective frequency points is lower than the minimum number of frequency points threshold, the data segment of the measurement window will not enter the subsequent segmentation mapping. The minimum number of frequency points threshold is determined by the calibration stage and written into the parameter file.

[0056] After the above processing, the input impedance spectrum and transmission impedance spectrum that share the same excitation frequency set and the same effective frequency determination result are obtained, and are used for subsequent segmented mapping and operating condition normalization processing.

[0057] In this embodiment, segmented mapping is performed after frequency point consistency screening is completed using the input impedance spectrum and transmission impedance spectrum. The input impedance spectrum and transmission impedance spectrum correspond to the frequency domain responses of the excitation end of the target harness loop and the measurement port at the other end of the target harness loop, respectively. The harness topology parameters used for segmented mapping include the branch connection matrix, segment length parameters, segment propagation speed parameters, and connection node reflection coefficient parameters. These parameters are stored and bound according to vehicle model, harness number, and version number.

[0058] First, a segmented directed graph is established based on the wiring harness connection relationships. Connector terminals, branch nodes, and terminal loads are considered nodes, and conductor segments are considered segments. A branch connection matrix is ​​constructed according to the node-segment relationship, with matrix elements taking values ​​of one, negative one, or zero. One indicates that a segment originates from that node, negative one indicates that a segment merges into that node, and zero indicates no connection. The branch connection matrix is ​​used to constrain the set of candidate propagation paths, excluding paths that do not satisfy the topological connection relationships.

[0059] Secondly, the segment propagation delay parameters are calculated based on the segment length and segment propagation speed parameters. The propagation delay parameter for each segment is determined by the ratio of the segment length to the segment propagation speed. For each candidate path, the nominal path delay is obtained by summing the propagation delay parameters of all segments contained in the path. The reflection coefficient parameter of the connecting node is obtained from calibration tests and stored in association with the node identifier. The nominal delay neighborhood width is determined statistically from calibration tests and managed according to temperature zones, loop current zones, and state of charge zones.

[0060] Then, a multipath propagation equation is established at each effective excitation frequency and the input impedance spectrum and transmission impedance spectrum are solved jointly. During the solution, the branch connection matrix is ​​used as the path coefficient constraint and the path nominal delay neighborhood is used as the delay constraint. The path components with overlapping delay neighborhoods are jointly estimated to obtain the amplitude coefficient of each path at the excitation frequency, thereby completing the separation of overlapping responses.

[0061] After completing the overlapping response separation, the segment response contribution is calculated according to the path-to-segment mapping relationship. The segment response contribution is a weighted synthesis of the path amplitude coefficients containing the segment. The weights are linearly combined by the reciprocal of the path traversal count and the absolute value of the reflection coefficient of the corresponding node according to a preset ratio. The current segment response contribution is compared with the corresponding segment response contribution under the baseline state to obtain the segment anomaly. When the segment anomaly exceeds the segment judgment threshold, the segment identifier is output. For connected nodes, the node is jointly judged based on the reflection coefficient offset of the path associated with the node and the mapping residual. When the node exceeds the node judgment threshold, the node identifier is output.

[0062] In this embodiment, the mapping residual is defined as the error energy between the measured and reconstructed values ​​of the input impedance spectrum and the transmission impedance spectrum. If the mapping residual exceeds the residual upper limit, the current window is marked as a low confidence window and will not proceed to the subsequent aging residual calculation. The segment judgment threshold, node judgment threshold and residual upper limit are statistically determined by the calibration stage and stored according to the working condition partition. If there is no corresponding parameter in the current working condition partition during online operation, the parameter is obtained by interpolation according to the adjacent working condition partition. If the interpolation condition is not met, the conservative parameter is called.

[0063] Through the above process, the segmented mapping outputs segment identifiers, node identifiers, and corresponding mapping residuals, and maintains the same time base and data caliber as the preceding sampling permission window, transient shielding interval, and frequency point consistency screening. It can be directly used for subsequent operating condition normalization and segmented aging assessment.

[0064] In this embodiment, the operating condition reference library is used to provide a unified reference for segmented spectrum characteristics. The operating condition reference library is established by collecting data from the same vehicle model and the same wiring harness number under the condition of new wiring harness installation. It is organized and managed according to temperature partition, circuit current partition, and charge state partition. The condition of new wiring harness installation is defined as the state after the target wiring harness circuit and its connector have been assembled and passed the electrical connectivity check, and there are no wiring harness replacement records or connector replacement records in the maintenance records.

[0065] The data used to build the operating condition benchmark library comes only from valid sampled data segments. Valid sampled data segments must simultaneously meet the following conditions: Sampling occurs within the sampling permission window; The transient sampling points of the switch have been removed according to the transient shielding interval; The frequency point consistency screening has been completed for the input impedance spectrum and the transmission impedance spectrum; The number of effective frequency points is not less than the minimum effective frequency point threshold; The number of continuously valid measurement windows is not less than the minimum window number threshold; The minimum effective frequency point threshold and the minimum window size threshold are determined during the calibration phase and written into the parameter file.

[0066] During database construction, the vehicle controller determines the current operating condition zone based on the harness temperature signal, loop current signal, and state of charge signal. For each operating condition zone, valid sampled data segments are collected cumulatively. For each harness segment and each excitation frequency point, the amplitude and phase of the input impedance spectrum and transmission impedance spectrum are statistically analyzed. Outliers are removed using the median absolute deviation rule, and the removal threshold is determined during the calibration phase. After the statistics are completed, the reference spectrum of the harness segment under the operating condition zone is generated. The reference spectrum includes the input impedance reference spectrum and the transmission impedance reference spectrum, and is stored in association with the harness segment identifier, node identifier, operating condition zone identifier, and version number.

[0067] The operating condition baseline library adopts segmented version management. Each version includes vehicle model identifier, wiring harness number, wiring harness segment identifier, operating condition zone identifier, library creation time, validity status, and version number. The validity status includes at least three categories: valid, pending reconstruction, and invalid. After the wiring harness or connector is replaced, the corresponding segment reconstruction process is triggered based on the maintenance record. Before the reconstruction is released, the old version remains valid and continues to participate in online evaluation. The reconstruction release condition is that the new version meets the library creation integrity condition. After the new version is released, the old version becomes invalid.

[0068] When the current operating condition partition lacks a valid version during online evaluation, the reference spectrum of the two adjacent operating condition partitions is used first to generate a temporary reference by linear interpolation of the normalized partition distance. If there is only one adjacent valid partition, the reference of that partition is directly used and marked as a low-confidence reference. The low-confidence reference is only used to maintain the continuity of the evaluation and is written into the evaluation results. It does not replace the formal reconstruction version.

[0069] In this embodiment, the operating condition reference library is input from the dual-end spectrum and established in segments. The output is directly used to generate segmented compensation features for operating condition normalization. This process maintains the same data caliber and time base as the sampling permission window, transient shielding interval, dual-end spectrum calculation and segmented mapping, which can support the engineering implementation of online evaluation, version reconstruction and result traceability.

[0070] In this embodiment, the operating condition normalization is performed after the segmented mapping. The input of the operating condition normalization is the wire harness temperature measurement value, loop current measurement value, state of charge measurement value at the sampling time, and the segmented spectral characteristics within the corresponding window. The segmented spectral characteristics include the input impedance spectral components and transmission impedance spectral components of each wire harness segment at each excitation frequency point. The output of the operating condition normalization is the segmented compensation characteristics under a unified reference operating condition, which is used for subsequent segmented aging residual calculation and degradation assessment.

[0071] The unified reference operating condition is predefined by the operating condition reference library, which includes reference temperature, reference loop current and reference state of charge. The boundaries of temperature zone, loop current zone and state of charge zone are determined by the calibration stage and written into the parameter file. During online operation, the vehicle controller locates the upper and lower adjacent zones of the three dimensions based on the measurement values ​​at the sampling time and calculates the interpolation weights of the three dimensions.

[0072] Subsequently, for each harness segment and each excitation frequency point, the reference versions of the input impedance reference spectrum and transmission impedance reference spectrum in adjacent operating condition partitions are read respectively, and three-dimensional linear interpolation is performed to obtain the new installation state prediction reference spectrum under the corresponding operating condition at that sampling time. During interpolation, the amplitude component and the phase component are processed separately. The phase component is first phase expanded and then interpolated. The reference spectrum of the unified reference operating condition is directly read from the same version system.

[0073] The compensation amount is defined as the difference between the predicted reference spectrum of the new installation under the current operating condition and the reference spectrum of the unified reference operating condition. The segmented compensation feature is defined as the segmented spectrum feature minus the compensation amount. This calculation is performed on the input impedance spectrum component and the transmission impedance spectrum component respectively to ensure that the two types of spectra are normalized under the same caliber.

[0074] When the working condition partition data is missing, the rollback is performed in the following order: if there is only one available partition on a certain dimension, the nearest available partition is taken for that dimension, and interpolation for that dimension is not performed. If there are not enough three-dimensional adjacent partitions, two-dimensional interpolation or one-dimensional interpolation is performed in descending order according to the available dimensions. If an effective prediction benchmark spectrum still cannot be formed, the most recently published effective version of the same harness segment is called and a low confidence flag is written. The low confidence flag is only used for result confidence management and does not change the compensation calculation process.

[0075] Operating condition normalization only processes data windows that meet the following conditions: the data comes from the sampling permission window, the sampling points in the transient shielded interval have been removed, the input impedance spectrum and the transmission impedance spectrum have completed the effective frequency point screening under the same excitation frequency point set, the number of effective frequency points is not less than the minimum frequency point number threshold, and the window length is not less than the minimum effective window duration. The above thresholds are determined by the calibration stage and written into the parameter file.

[0076] The segmented compensation features output in this embodiment are associated with and stored as harness segment identifiers, node identifiers, excitation frequency points, timestamps, and reference version numbers, and are directly passed to the segmented aging residual calculation step. By performing operating condition normalization on the input impedance spectrum and transmission impedance spectrum at the segment level, the reversible offset effects caused by temperature, loop current, and state of charge fluctuations can be reduced, maintaining the comparability and continuity of aging judgment results under different operating conditions.

[0077] In this embodiment, the degradation assessment model is used to map the segmented aging residual characteristics to the segmented aging assessment results of the harness. The degradation assessment model is trained with monotonic constraints. The training samples are from the historical measurement sequences of the corresponding harness segments in the operating condition benchmark library. Each historical measurement sequence is organized in chronological order and includes at least a timestamp, segmented compensation characteristics, segmented aging residual characteristics, harness temperature measurement, loop current measurement, state of charge measurement, number of effective frequency points, window validity marker, and maintenance event marker. The sample is included in the training set only when the window validity marker is valid, the segment identifier is consistent, the number of effective frequency points is not less than the minimum effective frequency point threshold, and the window duration is not less than the minimum effective window duration. The above thresholds are determined by the calibration stage and written into the parameter file.

[0078] To ensure consistency between the training and online evaluation standards, this embodiment models each harness segment independently. The training input includes frequency domain degradation features, stress accumulation features, and data quality features. The frequency domain degradation features are the amplitude offset statistics and phase offset statistics of the segmented aging residual features at each excitation frequency point. The stress accumulation features include cumulative thermal stress and cumulative electrical stress. The data quality features include effective frequency point coverage and mapping residual statistics. The training output is a continuous value aging index, with a larger value indicating a higher degree of degradation.

[0079] The cumulative thermal stress is calculated by recursively integrating over time. The cumulative thermal stress of the kth effective window is equal to the sum of the cumulative thermal stress of the (k-1)th window and the temperature contribution of the current window. The temperature contribution is equal to the non-negative part of the difference between the wire harness temperature measurement value and the reference temperature multiplied by the window duration. The cumulative electrical stress is calculated in the same way. The current contribution of the current window is equal to the preset stress exponent raised to the power of the absolute value of the loop current measurement value multiplied by the window duration. The reference temperature, preset stress exponent, and window duration are all determined during the calibration stage and fixed in the parameter file.

[0080] Monotonic constraints are applied to the relationship between the aging index and the cumulative thermal stress and cumulative electrical stress of the same harness segment. The constraint direction is monotonically non-decreasing. During the training phase, only the model version that has passed the monotonic constraint verification is released. The model version also needs to be verified through playback. The playback verification indicators include playback error not exceeding the playback error threshold and monotonic constraint violation rate not exceeding the violation rate threshold. The threshold is determined during the calibration phase.

[0081] The aging levels are divided into three levels: normal, warning, and deterioration. The level thresholds are determined by the statistical distribution of the operating condition benchmark library. The upper limit threshold of the normal level is taken from the preset high quantile value of the healthy sample distribution, the lower limit threshold of the deterioration level is taken from the preset low quantile value of the degraded sample distribution, and the warning level is between the two. If the degraded sample volume of a certain segment is less than the minimum sample volume threshold, the shared threshold of the same vehicle model, the same wiring harness type, and the same segment topology category will be temporarily used. After the sample volume meets the requirements, the segment-specific threshold will be switched. The threshold version number is managed in association with the operating condition benchmark library version number.

[0082] During online evaluation, the system reads the current segment aging residual characteristics, cumulative thermal stress, and cumulative electrical stress, inputs them into the degradation evaluation model to obtain the aging index, and outputs the aging level according to the corresponding threshold version. The evaluation confidence is obtained by weighting the normalized three indicators: effective frequency coverage, current operating condition partition coverage, and mapping residual consistency. The weights are determined and fixed during the calibration stage. When the baseline version is rebuilt due to wire harness replacement or connector replacement, the affected segment enters the model retraining process. Before the retraining is completed and passed the verification, the old model service is maintained and a low confidence mark is output. After the new model is released, it takes effect simultaneously with the new baseline spectrum version.

[0083] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0084] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for online assessment of the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis, characterized in that, include: When the vehicle's high-voltage system is powered on and the target harness circuit is turned on, a diagnostic excitation signal is injected into the target harness circuit through the excitation terminal. Under the same sampling time base, the excitation end voltage response signal, excitation end current response signal, voltage response signal of the measurement port at the other end of the target harness circuit, current response signal of the measurement port at the other end of the target harness circuit, vehicle switch status signal, harness temperature signal, circuit current signal and state of charge signal are synchronously acquired. A transient shielding interval is generated based on the vehicle switch status signal. Sampling points within the transient shielding interval are then removed to obtain valid sampling points. Calculate the input impedance spectrum at the excitation end and the transmission impedance spectrum from the excitation end to the measurement port at the other end based on the effective sampling points; Based on the pre-stored harness topology parameters and segmented propagation delay parameters, the input impedance spectrum and transmission impedance spectrum are segmented and mapped to obtain segmented spectrum characteristics; Based on the operating condition benchmark library, the segmented spectral features are normalized according to the operating conditions to obtain segmented compensation features. The segmented compensation features are differentially divided with the reference spectrum of the corresponding segments to obtain the segmented aging residual features; The segmented aging residual characteristics are input into the degradation assessment model, and the segmented aging assessment results of the wire harness are output.

2. The online assessment method for the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis according to claim 1, characterized in that: The diagnostic excitation signal is a multi-sine combination signal, a linear sweep frequency signal, or a pseudo-random binary sequence signal. The upper limit of the excitation amplitude is determined by the calibration stage, and the calibration constraints are the service circuit voltage fluctuation threshold and the communication bit error rate threshold.

3. The online assessment method for the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis according to claim 2, characterized in that: Synchronous data acquisition is performed within the sampling permission window, which is generated by the vehicle controller based on the switch status signal and the loop current ripple threshold. The loop current ripple threshold is determined by statistical analysis of historical operating data during the calibration phase.

4. The online assessment method for the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis according to claim 3, characterized in that: The transient shielding interval is determined by the inverter switching edge, the relay activation moment, and the relay release moment, and includes the protection duration before and after each moment. The protection duration is determined by calibration tests.

5. The online assessment method for the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis according to claim 4, characterized in that: The input impedance spectrum and transmission impedance spectrum are calculated based on the same set of excitation frequencies. The input impedance spectrum is obtained by the ratio of the complex components of the excitation terminal voltage to the complex components of the excitation terminal current at each excitation frequency, and the transmission impedance spectrum is obtained by the ratio of the complex components of the voltage at the other end of the measurement port to the complex components of the excitation terminal current at each excitation frequency. Frequency points with coherence below the calibration threshold are removed based on the phase dispersion and amplitude dispersion of repeated sampling.

6. The online assessment method for the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis according to claim 5, characterized in that: The harness topology parameters include branch connection matrix, segment length parameter, segment propagation speed parameter, and connection node reflection coefficient parameter. During segment mapping, the branch connection matrix is ​​used as the equation coefficient constraint to establish the multipath propagation equation, and the segment propagation delay parameter is used as the solution constraint to separate the overlapping response, and the segment identifier and node identifier are output.

7. The online assessment method for the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis according to claim 6, characterized in that: The operating condition benchmark library is established by collecting data from the same vehicle model and wiring harness number under the condition of newly installed wiring harness. The data collection dimensions include temperature partition, loop current partition, and state of charge partition. After the wiring harness or connector is replaced, the benchmark spectrum version of the corresponding segment is reconstructed.

8. The online assessment method for the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis according to claim 7, characterized in that: The operating condition normalization adopts a partitioned interpolation method, which calculates the compensation amount by linear interpolation within adjacent operating condition partitions. The inputs are the wire harness temperature measurement value, loop current measurement value, and state of charge measurement value at the sampling time, and the output is the segmented compensation feature under a unified reference operating condition.

9. The online assessment method for the aging status of wiring harnesses in new energy vehicles based on impedance spectrum analysis according to claim 8, characterized in that: The degradation assessment model is trained using monotonic constraints. The training samples are historical measurement sequences of the corresponding segments in the operating condition benchmark library. The monotonic constraint object is the relationship between the aging index of the same segment and the cumulative thermal stress and cumulative electrical stress. The cumulative thermal stress is obtained by temperature-time integration, and the cumulative electrical stress is obtained by current-time integration. The aging assessment results of the wiring harness segment include aging index, aging level and assessment confidence level. The aging level is divided into three levels: normal, warning and deterioration. The threshold of the three levels is determined by statistical distribution of the working condition benchmark library.

Images