A durability test method for window regulators in new energy vehicles

Abstract

This invention relates to the field of durability testing technology, specifically a durability testing method for a new energy vehicle window regulator. The method includes the following steps: obtaining the ratio of current change rate to sliding speed and determining the trend using a differential sliding window; generating trend instructions; calculating the current peak offset and generating delayed pulses through fuzzy classification; collecting symmetric offsets and filtering abnormal sequences; outputting warning levels through a two-parameter clustering model; and correcting abnormal label levels using the Drools engine. In this invention, by dynamically monitoring the ratio of motor load current change rate to window sliding speed, abnormal fluctuations are captured through sliding window analysis. A fuzzy logic classifier is used to classify the current peak offset and generate delayed pulses. Symmetric offsets are calculated and trends are evaluated. Segment positions are compared and level labels are output. Multiple algorithms are integrated to form a closed-loop feedback, improving warning accuracy and testing efficiency.

Description

Technical Field

[0001] This invention relates to the field of durability testing technology, and in particular to a durability testing method for a window regulator in a new energy vehicle. Background Technology

[0002] The field of durability testing technology for new energy vehicles includes methods for evaluating and verifying the performance changes of multiple key components of new energy vehicles during long-term use. By setting simulated usage environments and operating conditions, vehicle components are subjected to long-term, repeated operation tests to determine faults and performance degradation that may occur in real-world applications. Durability testing focuses on multiple aspects, including electric drive systems, power battery systems, vehicle structural components, and internal control and transmission systems. Through scientific and systematic testing methods, data on the performance stability and structural integrity of components under long-term operation are obtained. In the course of its development, this field has gradually formed a complete testing system supported by environmental simulation test benches, load cycling systems, and automated testing platforms, characterized by long testing cycles, complex operating parameters, and detailed data analysis.

[0003] One of the patented methods for durability testing of new energy vehicle window regulators is a test scheme used to evaluate the structural and functional changes of new energy vehicle window regulators under long-term operating conditions. Key aspects of the patent include: continuously and repeatedly opening and closing the window regulator by setting operating cycles and lifting strokes; simulating the working state under varying vehicle voltage conditions through an external control system; using a mechanical loading device to simulate the impact of glass mass and external resistance on the regulator's operation; synchronously collecting response parameters of the regulator at multiple operating stages using a time recording device, current and voltage monitors, and position sensors to form a systematic data record; and periodically sampling and analyzing motor temperature rise and gear wear during the test. This method uses continuous operation control, voltage disturbance simulation, mechanical resistance loading, and multi-dimensional synchronous monitoring as key means to complete the entire testing process.

[0004] Existing technologies employ fixed-cycle operation control and static threshold judgment mechanisms, making it difficult to identify the gradual performance degradation characteristics of the elevator under complex operating conditions such as voltage fluctuations and dynamic changes in mechanical resistance. Environmental simulation test benches rely on preset load cycle modes and lack the ability to analyze the correlation between motor load current and motion parameters in real time, resulting in lag in the detection of abnormal operating conditions. Although mechanical loading devices can simulate constant resistance, they lack a dynamic mapping model between resistance changes and electrical signal responses, failing to capture nonlinear current fluctuations caused by structural damage due to increased gear clearance. Data acquisition systems use discrete monitoring methods and do not perform synchronous correlation analysis of multi-parameter time-series characteristics, causing key fault features to be smoothed and filtered out during data preprocessing. Traditional testing methods rely on empirical threshold settings and manual visual inspection, lacking quantitative analysis methods for weak signals generated by early wear, easily leading to misjudgments or missed detections. Even in the development stage of microcracks on gear tooth surfaces, existing technologies, due to the lack of a correlation model between current peak offset and motion symmetry, struggle to distinguish between normal operating condition fluctuations and early fault signals, causing the warning threshold setting to deviate from the degradation trajectory. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a durability testing method for window regulators in new energy vehicles.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a durability testing method for a new energy vehicle window regulator, comprising the following steps:

[0007] S1: Obtain the change rate of the load current of the lifting motor and the glass sliding speed and calculate the ratio. Calculate the difference of the ratio for continuous periods and substitute it into the sliding window analysis method. Make a judgment based on the fact that the larger the window, the smaller the threshold for the difference trend, and generate a trend judgment instruction.

[0008] S2: Based on the trend judgment instruction, calculate the current peak offset between the current cycle and the previous cycle motor load current peak amplitude, and substitute it into the fuzzy logic classifier and combine it with the environmental disturbance compensation mechanism for dynamic matching and amplitude level division. Generate delay adjustment pulse according to the delay mapping rule corresponding to the amplitude level.

[0009] S3: The time from glass start-up to current surge and from current return to zero to displacement stop during the lifting cycle is collected by the delay adjustment pulse. The symmetry offset of the time is calculated according to the direction segment. Based on the phase lock phenomenon between the symmetry offset and the current peak offset, data with symmetry offset growth exceeding twice the current peak are selected to form a symmetry offset sequence.

[0010] S4: Compare the segment positions of the symmetry offset in the period, calculate the period segment offset ratio, and combine it with the current peak offset to input a two-parameter K-means clustering model for clustering, and output the wear warning level label; S5: Input the wear warning level label data, trend judgment command and symmetry offset sequence into the Drools rule engine and perform Z-score standardization. If the differential trend and symmetry offset trend rise continuously and the peak fluctuation is abnormal, the label level is automatically corrected.

[0011] As a further aspect of the present invention, the trend judgment instruction specifically includes differential trend threshold, incremental trend state, and sliding window analysis result; the delay adjustment pulse includes current peak offset, amplitude level classification result, and delay mapping rule; the symmetry offset sequence specifically refers to time symmetry offset, sliding window proportion evaluation result, and double growth screening condition; the wear warning level label includes segment location comparison parameter, two-parameter K-means clustering model, and wear level classification standard; and the automatically corrected label level specifically includes Z-score standardized data, continuous trend judgment index, and peak fluctuation anomaly.

[0012] As a further aspect of the present invention, the following steps are taken to obtain the ratio of the load current change rate of the lifting motor to the glass sliding speed, calculate the difference of the ratio for consecutive periods, substitute it into the sliding window analysis method, and make a judgment based on the principle that the larger the window, the smaller the threshold for judging the difference trend, and generate a trend judgment instruction:

[0013] S101: Obtain the rate of change of the load current of the lifting motor and the glass sliding speed, calculate the change value of the motor load current and the change value of the glass sliding distance respectively, and calculate the ratio within the same period to generate a ratio sequence;

[0014] S102: Based on the ratio sequence, call the data segments within three consecutive periods, construct a difference data group and a continuous window based on the difference between the ratios of two adjacent periods, arrange and analyze the multiple difference data within the continuous window, and generate a difference trend interval.

[0015] S103: For the differential trend interval, the differential trend judgment threshold is used for discrimination. The direction of data trend change and the increasing property within the multi-window are compared item by item. The differential trend judgment threshold is adjusted according to the window size. For large windows, the differential trend judgment threshold is lowered, and for small windows, the differential trend judgment threshold is raised. It is then determined whether the condition of being greater than the differential trend judgment threshold and the continuous increasing direction are met simultaneously, and a trend judgment instruction is generated.

[0016] As a further aspect of the present invention, based on the trend judgment instruction, the current peak offset between the current cycle and the previous cycle motor load current peak amplitude is calculated, and then substituted into a fuzzy logic classifier combined with an environmental disturbance compensation mechanism for dynamic matching and amplitude level division. The specific steps for generating a delay adjustment pulse according to the delay mapping rule corresponding to the amplitude level are as follows:

[0017] S201: Based on the trend judgment instruction, call the maximum instantaneous current value in the motor load current signal of the current cycle and the previous cycle, calculate the difference between the maximum instantaneous current values ​​in the two cycles, and compare it with the current amplitude change benchmark value corresponding to the current period of the current cycle to obtain the current peak offset.

[0018] S202: Based on the current peak offset, call the amplitude boundary value in the fuzzy logic classifier partitioning rules, classify the position corresponding to the current peak offset into the target amplitude level range, and perform a mapping operation with the hierarchical structure set by the fuzzy logic classifier to generate an amplitude level identifier.

[0019] S203: Call the amplitude level identifier, and according to the pulse adjustment coefficient corresponding to the identifier in the delay mapping rule table preset by the fuzzy logic classifier, load the corresponding pulse control instruction within the period to generate a delay adjustment pulse.

[0020] As a further aspect of the present invention, the specific steps for constructing a symmetrical offset sequence by acquiring the time from glass initiation to current surge and from current return to zero to displacement cessation during the lifting cycle through the delayed adjustment pulse acquisition, calculating the time symmetry offset by directional segment, and selecting data whose symmetrical offset increases by more than twice the current peak value to form a symmetrical offset sequence based on the phase-locking phenomenon between the symmetrical offset and the current peak offset are as follows:

[0021] S301: Obtain the timing data of the lifting cycle in the delay adjustment pulse, extract the start and end time of the glass start-up to the current surge stage and the start and end time of the current return to zero to the displacement stop stage, calculate the duration difference of the two time periods and construct the lifting response time difference according to the action direction.

[0022] S302: Based on the lifting response time difference, and taking the lifting segment and the falling segment time difference as the basis, when determining the offset direction, positive and negative marks are set in the order of lifting segment first and falling segment last. The obtained ratio is then used as the basis for symmetry judgment and classified into the lifting action category structure to obtain the symmetry offset amount.

[0023] S303: Call the symmetric offset, calculate the offset change rate through the sliding window percentage function, compare the difference between the symmetric offset of the current cycle and the previous cycle, if the symmetric offset and the current peak offset increase synchronously, it indicates a mechanical unilateral abnormality, if the symmetric offset and the current peak offset change independently, it indicates electrical interference, determine whether the symmetric offset exceeds twice the current peak offset, and obtain the symmetric offset sequence.

[0024] As a further aspect of the present invention, the specific steps for comparing the segment positions of the symmetry offset in the period, calculating the period segment offset ratio, and combining it with the current peak offset to input a two-parameter K-means clustering model for clustering, and outputting wear warning level labels are as follows:

[0025] S401: Based on the symmetric offset sequence and the current peak offset, an offset time segment is constructed by marking the ratio of the start and end times to the total duration of the cycle. The offset time segment is sorted by the cycle number to form a time axis index table. The cycle segment offset ratio is generated by combining the time ratio distribution of the offset time segment in the cycle.

[0026] S402: Call the periodic segment offset ratio and current peak offset, construct a data set with bivariate attributes from the two values, input it into the two-parameter K-means clustering model, initialize the cluster centers, and generate periodic offset cluster labels according to the spatial distribution of the two variables;

[0027] The two-parameter K-means clustering model includes input variable dimensions, cluster number setting, cluster center initialization, distance metric, iterative convergence criterion, and output structure.

[0028] S403: Based on the periodic offset clustering label, extract the average value of the offset ratio and current peak offset of each sample in the periodic segment in multiple label categories as the label feature benchmark and construct the feature distribution interval. Based on the ratio of the average offset difference between multiple sample categories and the corresponding interval, perform interval positioning determination to obtain the wear warning level label.

[0029] As a further aspect of the present invention, the input to the two-parameter K-means clustering model and the initialization of the cluster centers are performed using the following formula:

[0030] ;

[0031] in, The dynamic weighted distance index representing the p-th sample point. The value represents the measured value of the periodic segment offset ratio. Represents the coordinates of the r-th initial cluster center. This represents the measured value of the peak current offset. Represents the bivariate weighted adjustment factor. , The variance normalization coefficient representing the periodic segment offset ratio. The range normalization coefficient represents the current peak offset.

[0032] As a further aspect of the present invention, the wear warning level label data, trend judgment instructions, and symmetry offset sequence are input into the Drools rule engine and Z-score standardization is performed. If the differential trend and symmetry offset trend rise continuously and the peak fluctuations are abnormal, the specific steps for automatically correcting the label level are as follows:

[0033] S501: Load the wear warning level label data, friction trend judgment command and symmetry offset sequence input items into the Drools rule engine, perform field value reading and pre-matching, execute variable binding and rule activation initialization process on the structure format of the input data, and generate a rule trigger basic variable set;

[0034] The Drools rule engine includes a knowledge base, working memory, rule structure, rule execution flow, and variable binding mechanism;

[0035] S502: Invoke the rule to trigger multiple numerical variables in the basic variable set, calculate the mean and difference according to the field category, and complete the discretization mapping of all variables through Z-score standardization to generate a normalized offset fluctuation sequence;

[0036] S503: Based on the normalized offset fluctuation sequence, filter the fluctuation trajectory of continuous time periods and symmetric offsets related to the differential trend. If the offset increment direction of the two data in multiple continuous segments changes continuously in the positive direction, then perform an update operation on the label level field.

[0037] As a further aspect of the present invention, the discretization mapping of all variables is completed through Z-score standardization, using the formula:

[0038] ;

[0039] in, This represents the weighted standardized feature value of the nth sample in the mth field. Represents the original numerical variable measurement value. This represents the sliding window mean of the m-th field. This represents the dynamic standard deviation of the m-th field. This represents the difference between adjacent data windows of the m-th field. The cross-field association factor representing the nth sample. This represents the global normalization reference constant.

[0040] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0041] This invention dynamically monitors the ratio of the motor load current change rate to the glass sliding speed, and combines this with a sliding window analysis method to judge the differential trend within a continuous period in real time, capturing abnormal fluctuation characteristics during the operation of the elevator. Based on a fuzzy logic classifier, the current peak offset is graded and processed, and a delayed adjustment pulse is generated, significantly improving the response sensitivity to sudden changes in motor load conditions. A multi-dimensional time series correlation model is established through the calculation of symmetric offset and the trend evaluation of the sliding window proportion function, enhancing the ability to identify nonlinear degradation characteristics during the wear process of mechanical components. A two-parameter clustering model is used to compare and analyze the segment positions of the current peak offset and the symmetric offset sequence, breaking through the limitations of traditional single-dimensional threshold judgment and enabling multi-level classification of wear levels. This technical approach deeply integrates time series data analysis, fuzzy logic reasoning, and machine learning algorithms to form a closed-loop feedback mechanism. While reducing reliance on manual intervention, it significantly improves the accuracy and testing efficiency of early wear warning, providing high-resolution data support for the quantitative evaluation of elevator durability performance under operating conditions. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of the main steps of the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0044] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0045] Please see Figure 1 This invention provides a technical solution: a durability testing method for a new energy vehicle window regulator, comprising the following steps:

[0046] S1: Obtain the change rate of the load current of the lifting motor and the glass sliding speed and calculate the ratio. Calculate the difference of the ratio for continuous periods and substitute it into the sliding window analysis method. Make a judgment based on the fact that the larger the window, the smaller the threshold for the difference trend, and generate a trend judgment instruction.

[0047] S2: Based on the trend judgment instruction, calculate the current peak offset between the current cycle and the previous cycle motor load current peak amplitude, and substitute it into the fuzzy logic classifier and combine it with the environmental disturbance compensation mechanism for dynamic matching and amplitude level division. Generate delay adjustment pulse according to the delay mapping rule corresponding to the amplitude level.

[0048] S3: The time from glass start-up to current surge and from current return to zero to displacement stop during the lifting cycle is collected by the delay adjustment pulse. The symmetry offset of the time is calculated according to the direction segment. Based on the phase lock phenomenon between the symmetry offset and the current peak offset, data with symmetry offset growth exceeding twice the current peak are selected to form a symmetry offset sequence.

[0049] S4: Compare the segment positions of the symmetry offset in the period, calculate the period segment offset ratio, and combine it with the current peak offset to input a two-parameter K-means clustering model for clustering, and output the wear warning level label; S5: Input the wear warning level label data, trend judgment command and symmetry offset sequence into the Drools rule engine and perform Z-score standardization. If the differential trend and symmetry offset trend rise continuously and the peak fluctuation is abnormal, the label level is automatically corrected.

[0050] The trend judgment instructions specifically include differential trend threshold, increasing trend status, and sliding window analysis results. The delay adjustment pulse includes current peak offset, amplitude level classification results, and delay mapping rules. The symmetry offset sequence specifically refers to time symmetry offset, sliding window proportion evaluation results, and double growth screening conditions. The wear warning level label includes section location comparison parameters, two-parameter K-means clustering model, and wear level classification standards. The automatically corrected label level specifically includes Z-score standardized data, continuous trend judgment indicators, and peak fluctuation anomalies.

[0051] Please see Figure 1 This invention provides a technical solution: a durability testing method for a new energy vehicle window regulator, comprising the following steps:

[0052] S101: Obtain the rate of change of the load current of the lifting motor and the glass sliding speed, calculate the change value of the motor load current and the change value of the glass sliding distance respectively, and calculate the ratio within the same period to generate a ratio sequence;

[0053] Based on the data acquisition of the lifting motor operation, within the time window t=0s to t=2s, a current value sequence [2.3A, 2.5A, 2.8A, 3.1A, 3.4A] is obtained with a sampling interval of 100ms. Simultaneously, glass displacement sensor data is acquired to obtain a position sequence [0mm, 12mm, 25mm, 39mm, 54mm]. The current difference between adjacent sampling points is calculated to obtain a rate of change sequence [0.2A, 0.3A, 0.3A, 0.3A]. The displacement difference generates a velocity sequence [12mm / s, 13mm / s, 14m]. [m / s, 15mm / s], taking the average current change rate of 0.25A and the average speed of 12.5mm / s within the period from t=0.1s to 0.3s, the ratio calculation 0.25A / 12.5mm / s=0.02A·s / mm is performed. The process is repeated to obtain five cycle ratio sequences [0.0167, 0.0200, 0.0214, 0.0200, 0.0227], as shown in Table 1. Table 1: Motor Operating Parameter Acquisition Table

[0054] time Current Displacement Current change speed ratio 0.1 2.5 12 0.2 12 0.0167 0.2 2.8 25 0.3 13 0.0200 0.3 3.1 39 0.3 14 0.0214 0.4 3.4 54 0.3 15 0.0200

[0055] S102: Based on the ratio sequence, call the data segments within three consecutive periods, construct a difference data group and a continuous window based on the difference between the ratios of two adjacent periods, arrange and analyze the multiple difference data within the continuous window, and generate the difference trend interval.

[0056] Select the data points with indexes 1-3 of the ratio sequence [0.0200, 0.0214, 0.0200]. Calculate the differences between adjacent indices: 0.0214 - 0.0200 = 0.0014 and 0.0200 - 0.0214 = -0.0014. Construct a difference group [0.0014, -0.0014]. Set the window length to 2. Sort the difference values ​​by size to obtain [-0.0014, 0.0014]. Determine the lower limit of the interval: -0.0014. With the upper limit of 0.0014, take the data from indexes 2-4 of the second window [0.0214, 0.0200, 0.0227], calculate the difference between -0.0014 and 0.0027, sort them to get [-0.0014, 0.0027], generate the second interval [-0.0014, 0.0027], and the continuous window data form the trend interval set {[-0.0014, 0.0014], [-0.0014, 0.0027]}.

[0057] S103: For the differential trend range, the differential trend judgment threshold is used for discrimination. The direction of data trend change and the increasing nature of the data within the multi-window are compared item by item. The differential trend judgment threshold is adjusted according to the window size. For large windows, the differential trend judgment threshold is lowered, and for small windows, the differential trend judgment threshold is raised. It is then determined whether the condition of being greater than the differential trend judgment threshold and the continuous increasing direction are met simultaneously, and a trend judgment instruction is generated.

[0058] The trend judgment threshold is set to 0.0020, which is derived from the maximum fluctuation value under normal operating conditions in the data statistics. The upper limit of the first window, 0.0014, is less than the threshold of 0.0020 and is marked as an invalid interval. The upper limit of the second window, 0.0027, is greater than the threshold of 0.0020. The data direction is verified: in the difference sequence [-0.0014, 0.0027], the last term 0.0027 is greater than the previous term -0.0014, but there is no continuous increase. The judgment window only meets the threshold condition but not the direction condition, so the trend judgment instruction is 0 (invalid instruction). When three consecutive windows simultaneously meet the condition that the upper limit is greater than 0.0020 and the difference value is monotonically increasing, the instruction 1 (valid trend) is output.

[0059] Please see Figure 1 This invention provides a technical solution: a durability testing method for a new energy vehicle window regulator, comprising the following steps:

[0060] S201: Based on the trend judgment instruction, call the maximum instantaneous current value in the motor load current signal of the current cycle and the previous cycle, calculate the difference between the maximum instantaneous current values ​​in the two cycles, and compare it with the current amplitude change benchmark value corresponding to the current acquisition period to obtain the current peak offset.

[0061] Based on the valid state where the trend judgment instruction value is 1, extract the maximum current value of the nth cycle. A and the (n-1)th period A, Calculate the difference A, Obtain the reference value of the current cycle amplitude change. A (calculated by the moving average of the maximum current value over the previous 10 cycles), perform offset calculation. A, as shown in Table 2, experimental data, when The time marker is marked as an abnormal offset, generating a peak current offset of 0.05A.

[0062] Table 2 Comparison of Peak Current

[0063] cycle Maximum current (A) Baseline value (A) Offset (A) n-1 3.2 0.24 0.08 n 3.5 0.25 0.05

[0064] S202: Based on the current peak offset, call the amplitude boundary value in the fuzzy logic classifier partitioning rules, assign the position corresponding to the current peak offset to the target amplitude level range, and perform a mapping operation with the hierarchical structure set by the fuzzy logic classifier to generate an amplitude level identifier.

[0065] Set the fuzzy logic amplitude boundary value to [0.1, 0.2]A (based on the 2%-4% safety fluctuation range of the motor's rated current of 5A). Compare the offset of 0.05A with the boundary value. If 0.05 < 0.1, it is determined to be a low amplitude level L1. Look up the classification mapping table to get the identifier code L1. When the offset is 0.15A, 0.1 ≤ 0.15 < 0.2 is classified as a medium amplitude level M2, and the identifier code M2 ​​is generated. If the offset is 0.25A, it is ≥ 0.2 and classified as a high amplitude level H3. Match the level identifier with the preset 9-level classification structure [L1, L2, L3, M1, M2, M3, H1, H2, H3] through a table lookup operation, and output the amplitude level identifier M2.

[0066] S203: Call the amplitude level identifier, and according to the pulse adjustment coefficient corresponding to the identifier in the delay mapping rule table preset by the fuzzy logic classifier, load the corresponding pulse control instruction within the period to generate a delay adjustment pulse;

[0067] Call the adjustment coefficient corresponding to the M2 identifier in the delay mapping rule table (Based on regression analysis of 500 sets of data:) M2 (serial number 2), with reference pulse width Using ms as a reference, calculate the adjusted pulse width. ms, generate a PWM waveform with a duty cycle of 45% (frequency 1kHz, amplitude 12V), load it into the motor drive module, and output a delay adjustment pulse sequence [0,1,0,1,0] (level duration 8ms, interval time 8ms).

[0068] Please see Figure 1 This invention provides a technical solution: a durability testing method for a new energy vehicle window regulator, comprising the following steps:

[0069] S301: Obtain the timing data of the lifting and lowering cycle in the delay adjustment pulse, extract the start and end time of the glass start-up to the current surge stage and the start and end time of the current return to zero to the displacement stop stage, calculate the duration difference between the two time periods and construct the lifting and lowering response time difference according to the action direction.

[0070] Based on delayed pulse waveform data (frequency 1kHz, sampling rate 10kHz), during the rise and fall cycle k, current sensors collect data points [0.5A@1250ms, 1.8A@1255ms, 2.8A@1256ms, 3.5A@1280ms] for the rising phase. The starting point of the current surge is defined as the first sampling point at 1256ms where the current exceeds the baseline by twice the standard deviation (0.2A), and the ending point is defined as three consecutive sampling points at 1280ms where the current stabilizes within the ±5% fluctuation range (3.5A±0.175A). A synchronous displacement sensor records the glass movement starting point at 1256ms (position 0mm→2mm) and the stopping point at 1376ms (position 54mm→54mm). The rise time is then calculated. ms, downgrade time ms (1352ms is the turning point where the current drops from 3.5A to 1.0A), construct the response time difference. ms, indicating direction coefficient (According to ISO 13849-1 standard: an ascent starting earlier than a descent is marked as +1). As shown in Table 3, when the ascent time of period k+1 is detected to be 28ms (1290-1262) and the descent time to be 20ms (1380-1360), the difference is calculated. ms, the direction marker is maintained at +1, and a sequence of rise and fall response time difference values ​​[0ms, 8ms] is generated.

[0071] Table 3 Timing Parameter Analysis Table

[0072] Parameter name Ascending phase Descent phase Start time (ms) 1256 (Current > 0.7A) 1352 (Current < 3.2A) End time (ms) 1280 (current stable ±5%) 1376 (Displacement change < 0.1 mm) Baseline duration (ms) 24 (Design specification value) 24 (Design specification value) Permissible deviation range (ms) ±2 (manufacturer's tolerance standard) ±2 (manufacturer's tolerance standard)

[0073] S302: Based on the time difference between the lifting and lowering responses, and taking the time difference between the lifting and lowering segments as the basis, when determining the offset direction, positive and negative marks are set in the order of lifting segment first and lowering segment last. The obtained ratio is then used as the basis for symmetry judgment and classified into the lifting action category structure to obtain the symmetric offset amount.

[0074] Set a symmetry determination threshold ms (according to Article 9.2.3 of GB / T5226.1-2019 Mechanical and Electrical Safety Standard, taken as 20% of the design duration), when the absolute value of the time difference When the time interval is ms, it is classified into class C1 (completely symmetric). When the period k+2 is detected with an ascent time of 26ms and a descent time of 32ms, the difference is calculated. ms, direction marker (The descending segment starts earlier than the ascending segment), symmetrical offset. According to the three-level classification standard: 0-5ms is C1 (normal), 5-10ms is C2 (warning), and >10ms is C3 (fault). +6ms is classified into C2. When constructing the offset sequence, three consecutive periods of data [0ms, +8ms, +6ms] are taken, and the moving average is calculated. ms, standard deviation ms, when ms or An asymmetric alarm is triggered at ms.

[0075] S303: Call the symmetric offset, calculate the offset change rate through the sliding window percentage function, compare the difference between the symmetric offset of the current cycle and the previous cycle, if the symmetric offset and the current peak offset increase synchronously, it indicates a mechanical unilateral abnormality, if the symmetric offset and the current peak offset change independently, it indicates electrical interference, determine whether the symmetric offset exceeds twice the current peak offset, and obtain the symmetric offset sequence.

[0076] Define the sliding window length Period, calculate the rate of change of the current period k+3 offset +10ms relative to the previous two periods [+8ms, +6ms]. The current peak offset sequence [0.05A, 0.07A, 0.18A, 0.22A] was acquired synchronously, and its rate of change was calculated. When satisfied and When the time interval (0.3>0.2 and 3.4>2) is reached, it is determined to be a mechanical unilateral anomaly. If the rate of change of the symmetrical offset sequence [+6ms, -8ms, +5ms] is detected, it is considered a mechanical unilateral anomaly. The current offset remained stable at 0.15A ± 0.03A. If the noise level is 5, it is determined to be electrical interference (PWM controller noise). This is determined by comparing the symmetry offset. ms and current peak offset A multiple relationship (Exceeding the preset alarm threshold by 20 times), triggering the level 2 fault code F202, generating a symmetric offset sequence with status markers [0(+0), 8(C2), 6(C2), 10(F202)].

[0077] Please see Figure 1 This invention provides a technical solution: a durability testing method for a new energy vehicle window regulator, comprising the following steps:

[0078] S401: Based on the symmetric offset sequence and current peak offset, the offset time segment is constructed by marking the ratio of the start and end times to the total duration of the cycle. The offset time segment is sorted by the cycle number to form a time axis index table. The cycle segment offset ratio is generated by combining the time ratio distribution of the offset time segment in the cycle.

[0079] Based on the symmetric offset sequence [0ms, +8ms, +6ms] and the current peak offset [0.05A, 0.18A, 0.22A], the start time of period k is extracted. ms and end time ms, total calculation period duration ms, construct the offset time segment A time axis index table (Table 4) is generated according to the period numbers k=1 to k=3. When the period offset of k=2 + 8ms is detected, the corresponding segment is... k=3 cycle offset + 6ms corresponds to When calculating the proportional distribution, the mean of three periods is taken. Standard deviation Generate a periodic segment offset ratio sequence [0.0667, 0.0678, 0.0492].

[0080] Table 4 Time Axis Index Table

[0081] Periodic number Start time (ms) End time (ms) Offset ratio 1 1256 1376 0.0667 2 1320 1438 0.0678 3 1465 1587 0.0492

[0082] S402: Call the periodic segment offset ratio and current peak offset, construct a data set with bivariate attributes from the two values, input it into the two-parameter K-means clustering model, initialize the cluster centers, and generate periodic offset cluster labels based on the spatial distribution of the two variables;

[0083] The two-parameter K-means clustering model includes the input variable dimension, cluster number setting, cluster center initialization strategy, distance metric, iterative convergence criterion, and output structure;

[0084] Set the number of clusters K=3 (based on the elbow rule, SSE descent inflection point), and initialize the cluster centers. , , Take sample point p1=(0.0667,0.05). ;in, The dynamic weighted distance index representing the p-th sample point. The value represents the measured value of the periodic segment offset ratio. Represents the coordinates of the r-th initial cluster center. This represents the measured value of the peak current offset. Represents the bivariate weighted adjustment factor ( ), The variance normalization coefficient representing the periodic segment offset ratio. The range normalization coefficient, representing the current peak offset, is used to calculate the weighted distance to C1: , , , Similarly, the distance to C2 is calculated to be 2.14 and the distance to C3 to be 5.62. Therefore, p1 is assigned to class C2, and the center point is iteratively updated. , , Generate cluster label sequence [2,2,1].

[0085] S403: Based on the periodic offset clustering label, extract the average value of the offset ratio and current peak offset of each sample in the periodic segment in multiple label categories as the label feature benchmark and construct the feature distribution interval. Based on the ratio of the average offset difference between multiple samples and the corresponding interval, the interval positioning is determined to obtain the wear warning level label.

[0086] Extract the mean of the offset ratios of the three samples in cluster 1. Peak offset mean Set the reference range , The mean offset of the detected sample point (0.0492, 0.05) was compared to... When the new sample (0.068, 0.19) falls into the cluster 2 interval , When calculating the mean difference ratio The warning levels are divided according to a threshold of 0.1: 0-0.05 is L1 (normal), 0.05-0.1 is L2 (observation), and >0.1 is L3 (warning), generating a wear label sequence [L1, L2].

[0087] Please see Figure 1 This invention provides a technical solution: a durability testing method for a new energy vehicle window regulator, comprising the following steps:

[0088] S501: Load the wear warning level label data, friction trend judgment command and symmetry offset sequence input items into the Drools rule engine, read and pre-match field values, perform variable binding and rule activation initialization process on the structure format of the input data, and generate a set of basic variables for rule triggering;

[0089] The Drools rule engine includes a knowledge base, working memory, rule structure, rule execution flow, and variable binding mechanism.

[0090] Based on wear warning tag sequence Trend command , and symmetry offset sequence Create a Fact object in the Drools rule engine Binding variables ,and Activate rules in the rule base ,when and Triggered on time, initializing working memory and loading the sample dataset. Set the size of the sliding window When the number of times the rule is triggered within the window exceeds the threshold At this time, the basic variable set is generated. As shown in Table 5, when the 4th sample was detected When, update the variable set to .

[0091] Table 5 Rule Triggering Variables Table

[0092] Variable name numerical values Calculation rules label-count 2 Count the occurrence of L2 / L3 labels trend-sum 2 The accumulated value of the trend instruction is 1 offset-avg 7 Offset Arithmetic Mean

[0093] S502: The call rule triggers multiple numerical variables in the basic variable set, calculates the mean and difference according to the field category, and completes the discretization mapping of all variables through Z-score standardization to generate a normalized offset fluctuation sequence.

[0094] Selecting the basic variable set Calculated fields The sliding window mean of (label-count) Standard deviation Difference between adjacent windows Correlation factors (Based on the Pearson correlation coefficient between fields), take (Empirical constant), calculate standardized eigenvalues For fields (trend-sum) calculation , Generate normalized sequences .

[0095] S503: Based on the normalized offset fluctuation sequence, filter the fluctuation trajectory of continuous time periods and symmetric offsets related to the differential trend. If the offset increment direction of the two data in multiple continuous segments changes continuously in the positive direction, then perform an update operation on the label level field.

[0096] Analysis of normalized sequences Calculate the increment for period 1-3 based on the changing trend over three consecutive periods. , When the continuous incremental direction count ( (Number of times) reached At this time, the symmetry shift sequence is detected. Increment , Determine whether two data points simultaneously satisfy the condition that there have been [conditions / restrictions] within the last three periods. This is the second positive change. Currently, only the offset sequence meets the condition, and the delayed update label level is... If a normalized increment appears in the period , and offset increment If all conditions are met, the label will be upgraded to [a higher level]. .

[0097] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A durability testing method for a window regulator in a new energy vehicle, characterized in that, Includes the following steps: S1: Obtain the change rate of the load current of the lifting motor and the glass sliding speed and calculate the ratio. Calculate the difference of the ratio for continuous periods and substitute it into the sliding window analysis method. Make a judgment based on the fact that the larger the window, the smaller the threshold for the difference trend, and generate a trend judgment instruction. S2: Based on the trend judgment instruction, calculate the current peak offset between the current cycle and the previous cycle motor load current peak amplitude, and substitute it into the fuzzy logic classifier and combine it with the environmental disturbance compensation mechanism for dynamic matching and amplitude level division. Generate delay adjustment pulse according to the delay mapping rule corresponding to the amplitude level. S3: The time from glass start-up to current surge and from current return to zero to displacement stop during the lifting cycle is collected by the delay adjustment pulse. The symmetry offset of the time is calculated according to the direction segment. Based on the phase lock phenomenon between the symmetry offset and the current peak offset, data with symmetry offset growth exceeding twice the current peak are selected to form a symmetry offset sequence. S4: Compare the segment positions of the symmetry offset in the period, calculate the period segment offset ratio, and combine it with the current peak offset to input a two-parameter K-means clustering model for clustering, and output the wear warning level label.

2. The durability testing method for new energy vehicle window regulators according to claim 1, characterized in that, The trend judgment instruction specifically includes differential trend threshold, increasing trend status, and sliding window analysis results. The delay adjustment pulse includes current peak offset, amplitude level classification results, and delay mapping rules. The symmetry offset sequence specifically refers to time symmetry offset, sliding window proportion evaluation results, and double growth screening conditions. The wear warning level label includes segment location comparison parameters, two-parameter K-means clustering model, and wear level classification standards.

3. The durability testing method for new energy vehicle window regulators according to claim 2, characterized in that, The specific steps for obtaining the ratio of the change rate of the lift motor load current to the glass sliding speed, calculating the difference of the ratio for consecutive periods, substituting it into the sliding window analysis method, and judging the trend by determining that the larger the window, the smaller the threshold, and generating a trend judgment instruction are as follows: S101: Obtain the rate of change of the load current of the lifting motor and the glass sliding speed, calculate the change value of the motor load current and the change value of the glass sliding distance respectively, and calculate the ratio within the same period to generate a ratio sequence; S102: Based on the ratio sequence, call the data segments within three consecutive periods, construct a difference data group and a continuous window based on the difference between the ratios of two adjacent periods, arrange and analyze the multiple difference data within the continuous window, and generate a difference trend interval. S103: For the differential trend interval, the differential trend judgment threshold is used for discrimination. The direction of data trend change and the increasing property within the multi-window are compared item by item. The differential trend judgment threshold is adjusted according to the window size. For large windows, the differential trend judgment threshold is lowered, and for small windows, the differential trend judgment threshold is raised. It is then determined whether the condition of being greater than the differential trend judgment threshold and the continuous increasing direction are met simultaneously, and a trend judgment instruction is generated.

4. The durability testing method for a new energy vehicle window regulator according to claim 3, characterized in that, Based on the trend judgment instruction, the current peak offset between the current cycle and the previous cycle motor load current peak amplitude is calculated, and then substituted into a fuzzy logic classifier combined with an environmental disturbance compensation mechanism for dynamic matching and amplitude level classification. The specific steps for generating delay adjustment pulses according to the delay mapping rules corresponding to the amplitude levels are as follows: S201: Based on the trend judgment instruction, call the maximum instantaneous current value in the motor load current signal of the current cycle and the previous cycle, calculate the difference between the maximum instantaneous current values ​​in the two cycles, and compare it with the current amplitude change benchmark value corresponding to the current period of the current cycle to obtain the current peak offset. S202: Based on the current peak offset, call the amplitude boundary value in the fuzzy logic classifier partitioning rules, classify the position corresponding to the current peak offset into the target amplitude level range, and perform a mapping operation with the hierarchical structure set by the fuzzy logic classifier to generate an amplitude level identifier. S203: Call the amplitude level identifier, and according to the pulse adjustment coefficient corresponding to the identifier in the delay mapping rule table preset by the fuzzy logic classifier, load the corresponding pulse control instruction within the period to generate a delay adjustment pulse.

5. The durability testing method for a new energy vehicle window regulator according to claim 4, characterized in that, The specific steps for constructing a symmetrical offset sequence by acquiring the time from glass initiation to current surge and current return to zero to displacement cessation during the lifting cycle using the delayed adjustment pulse acquisition, based on the phase-locking phenomenon between the symmetrical offset and the current peak offset, are as follows: S301: Obtain the timing data of the lifting cycle in the delay adjustment pulse, extract the start and end time of the glass start-up to the current surge stage and the start and end time of the current return to zero to the displacement stop stage, calculate the duration difference of the two time periods and construct the lifting response time difference according to the action direction. S302: Based on the lifting response time difference, and taking the lifting segment and the falling segment time difference as the basis, when determining the offset direction, positive and negative marks are set in the order of lifting segment first and falling segment last. The obtained ratio is then used as the basis for symmetry judgment and classified into the lifting action category structure to obtain the symmetry offset amount. S303: Call the symmetric offset, calculate the offset change rate through the sliding window percentage function, compare the difference between the symmetric offset of the current cycle and the previous cycle, if the symmetric offset and the current peak offset increase synchronously, it indicates a mechanical unilateral abnormality, if the symmetric offset and the current peak offset change independently, it indicates electrical interference, determine whether the symmetric offset exceeds twice the current peak offset, and obtain the symmetric offset sequence.

6. The durability testing method for a new energy vehicle window regulator according to claim 5, characterized in that, The specific steps for comparing the symmetry offset in the period, calculating the period segment offset ratio, and combining it with the current peak offset to input a two-parameter K-means clustering model for clustering, and outputting the wear warning level label are as follows: S401: Based on the symmetric offset sequence and the current peak offset, an offset time segment is constructed by marking the ratio of the start and end times to the total duration of the cycle. The offset time segment is sorted by the cycle number to form a time axis index table. The cycle segment offset ratio is generated by combining the time ratio distribution of the offset time segment in the cycle. S402: Call the periodic segment offset ratio and current peak offset, construct a data set with bivariate attributes from the two values, input it into the two-parameter K-means clustering model, initialize the cluster centers, and generate periodic offset cluster labels according to the spatial distribution of the two variables; The two-parameter K-means clustering model includes input variable dimensions, cluster number setting, cluster center initialization, distance metric, iterative convergence criterion, and output structure. S403: Based on the periodic offset clustering label, extract the average value of the offset ratio and current peak offset of each sample in the periodic segment in multiple label categories as the label feature benchmark and construct the feature distribution interval. Based on the ratio of the average offset difference between multiple sample categories and the corresponding interval, perform interval positioning determination to obtain the wear warning level label.

7. The durability testing method for a new energy vehicle window regulator according to claim 6, characterized in that, The input to the two-parameter K-means clustering model and the initialization of cluster centers are performed using the following formula: ； in, The dynamic weighted distance index representing the p-th sample point. The value represents the measured value of the periodic segment offset ratio. Represents the coordinates of the r-th initial cluster center. This represents the measured value of the peak current offset. Represents the bivariate weighted adjustment factor. , The variance normalization coefficient representing the periodic segment offset ratio. The range normalization coefficient represents the current peak offset.

8. The durability testing method for a new energy vehicle window regulator according to claim 6, characterized in that, The method further includes: S5: Input the wear warning level label data, trend judgment command and symmetry offset sequence into the Drools rule engine and perform Z-score standardization. If the differential trend and symmetry offset trend rise continuously and the peak fluctuation is abnormal, the label level is automatically corrected. The automatically corrected label level specifically includes Z-score standardized data, continuous trend judgment index, and peak fluctuation anomaly.

9. The durability testing method for a new energy vehicle window regulator according to claim 8, characterized in that, The wear warning level label data, trend judgment instructions, and symmetry offset sequence are input into the Drools rule engine and Z-score normalization is performed. If the differential trend and symmetry offset trend rise continuously and the peak fluctuations are abnormal, the specific steps for automatically correcting the label level are as follows: S501: Load the wear warning level label data, friction trend judgment command and symmetry offset sequence input items into the Drools rule engine, perform field value reading and pre-matching, execute variable binding and rule activation initialization process on the structure format of the input data, and generate a rule trigger basic variable set; The Drools rule engine includes a knowledge base, working memory, rule structure, rule execution flow, and variable binding mechanism; S502: Invoke the rule to trigger multiple numerical variables in the basic variable set, calculate the mean and difference according to the field category, and complete the discretization mapping of all variables through Z-score standardization to generate a normalized offset fluctuation sequence; S503: Based on the normalized offset fluctuation sequence, filter the fluctuation trajectory of continuous time periods and symmetric offsets related to the differential trend. If the offset increment direction of the two data in multiple continuous segments changes continuously in the positive direction, then perform an update operation on the label level field.

10. The durability testing method for a new energy vehicle window regulator according to claim 9, characterized in that, The discretization mapping of all variables is completed using Z-score normalization, employing the following formula: ； in, This represents the weighted standardized feature value of the nth sample in the mth field. Represents the original numerical variable measurement value. This represents the sliding window mean of the m-th field. This represents the dynamic standard deviation of the m-th field. This represents the difference between adjacent data windows of the m-th field. The cross-field association factor representing the nth sample. This represents the global normalization reference constant.

Images