A method for predicting the life of a vehicle shock absorber

Abstract

The application discloses a kind of life prediction methods of automobile shock absorber, and the application relates to life evaluation technical field.The steps of the method include: collecting engine speed, torsional vibration amplitude, oil temperature, load torque, and forming engine parameter set after preprocessing;Collect standard parameters of the same type engine, obtain the spectral peak by processing the torsional vibration amplitude in parameter set, calculate the engine resonance frequency in combination with standard parameters;Based on STL time series decomposition method, construct shock absorber operating state threshold model, input oil temperature and load torque to obtain threshold value, compare resonance frequency and threshold value to determine shock absorber damping state;Calculate torsional vibration change rate, compare with standard parameters of the same type, to obtain the working state of shock absorber;Based on cumulative damage theory, construct shock absorber life prediction model, input damping state, working state and engine parameter set, output remaining life.The application makes the prediction of the remaining life of the automobile shock absorber more accurate by considering multi-dimensional real-time operating condition parameters.

Description

Technical Field

[0001] This invention relates to the field of life assessment engineering, specifically to a method for predicting the life of automotive shock absorbers. Background Technology

[0002] As a core component for transmitting engine torsional vibration and mitigating shock, the lifespan of automotive shock absorbers directly affects vehicle driving safety, comfort, and transmission system reliability. With the increasing demands for intelligent vehicles and long-lifecycle maintenance, the traditional "periodic replacement" model can no longer meet the needs of precise maintenance. The industry urgently needs shock absorber lifespan prediction technology based on real-time operating conditions to avoid transmission failures caused by sudden malfunctions, reduce the cost waste from over-maintenance, and ensure the stable and coordinated operation of the engine and transmission system.

[0003] Currently, automotive shock absorber life assessment mostly employs traditional bench test calibration techniques. These tests obtain the standard fatigue life of the shock absorber, and in actual operation and maintenance, only a single real-time operating parameter such as torsional vibration amplitude or engine oil temperature is monitored to calculate the shock absorber's total lifespan damage. This is then combined with the standard fatigue life to calculate the remaining lifespan of the automotive shock absorber. However, existing technologies rely solely on a single real-time operating parameter when assessing the remaining lifespan of automotive shock absorbers, neglecting multi-dimensional real-time operating parameters such as engine speed, torsional vibration amplitude, engine oil temperature, and load torque. This underestimates the impact of total lifespan damage on the shock absorber's lifespan, leading to inaccuracies in the prediction of the remaining lifespan of automotive shock absorbers. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method for predicting the lifespan of automotive shock absorbers, which solves the problem of underestimating the impact of damage throughout the entire life cycle of shock absorbers by not considering multi-dimensional real-time operating parameters.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for predicting the lifespan of an automotive shock absorber, comprising the following steps: Step S1: Collect engine parameters, preprocess the engine parameters to obtain an engine parameter set, which includes engine speed, torsional vibration amplitude, oil temperature and load torque; Step S2: Collect standard parameters of the same model engine, process the torsional vibration amplitude of the engine parameter set using the short-time Fourier transform method to obtain the engine spectrum peak value; calculate the engine resonance frequency based on the standard parameters of the same model engine and the engine spectrum peak value. Step S3: Based on the STL time-series decomposition method, construct a shock absorber operating state threshold model; input the engine oil temperature and load torque into the shock absorber operating state threshold model, and output the shock absorber operating state threshold; by comparing the engine resonance frequency and the shock absorber operating state threshold, obtain the shock absorber damping state; Step S4: Calculate the torsional vibration change rate based on the engine speed and torsional vibration amplitude; compare the torsional vibration change rate with the standard parameters of the same model engine to obtain the working state of the shock absorber; Step S5: Based on the cumulative damage theory, construct a shock absorber life prediction model; input the shock absorber damping state, shock absorber working state and the engine parameter set into the shock absorber life prediction model, and output the remaining life of the shock absorber.

[0006] Preferably, engine parameters are collected and preprocessed to obtain an engine parameter set including: Collect engine parameters during vehicle operation, including key physical quantities such as engine speed, torsional vibration amplitude, oil temperature, and load torque; Suppose the collected raw engine parameters are in the following set. :

[0007] in Let n represent the parameter vector at the i-th sampling time, where n is the number of samples; Because the engine parameter set A may contain sensor noise, missing values, or outliers, using it directly will affect the accuracy of subsequent analysis; therefore, it needs to be preprocessed using a preprocessing function. Purify and standardize A; Construct the preprocessing function and define the preprocessing function. Its core operations include noise filtering, which uses sliding window mean filtering to suppress high-frequency noise; missing value imputation, which uses linear interpolation based on time series to complete the data; and normalization, which eliminates differences in units. The preprocessing procedure is as follows:

[0008] in, This is the preprocessed engine parameter set. This is the standardized parameter vector.

[0009] Preferably, standard parameters of the same engine model are collected, and the torsional amplitude values ​​in the engine parameter set are processed using the short-time Fourier transform method to obtain the engine spectrum peak values, including: The following is a set of standard parameters for the same engine model retrieved from the historical database:

[0010] in, It is the rated speed; It is the calibrated torsional vibration range; It is the resonant frequency threshold; It is the historical noise variance; It is the benchmark value for judging the damping state, which is obtained through statistical analysis of historical bench test data and normal real vehicle operation data; The maximum allowable deviation is obtained through statistical analysis of the fluctuation characteristics of the same set of historical normal data; a benchmark value for judging the damping state is defined. The upper limit of the reasonable fluctuation range; It's tolerance; It is the target torsional amplitude; This refers to the standard service life of the vibration damper. It is damage throughout the entire life cycle; First, torsional vibration amplitude is calibrated, then short-time Fourier transform is performed, and finally wavelet soft thresholding is used for secondary denoising. Torsional amplitude calibration:

[0011] in, It is the calibrated torsional vibration amplitude, which is used to eliminate individual differences in engines and scale the real-time torsional vibration signal to the standard amplitude range. It is the real-time torsional vibration amplitude value, which comes from the engine parameter set B output by S1; It is the upper limit of the torsional vibration range in the standard parameters, derived from ; It is the lower limit of the torsional vibration range in the standard parameters, derived from... ; It is the standard rated torsional vibration amplitude, based on historical bench test data, used to characterize the typical torsional vibration amplitude of a specific engine model under standard rated operating conditions; its function is to eliminate individual differences and ensure that the signal amplitude conforms to the model reference. The short-time Fourier transform process is as follows:

[0012]

[0013] in, It performs a short-time Fourier transform on the calibrated torsional vibration amplitude, which converts the time-domain vibration signal into a time-frequency domain representation to locate the resonant frequency as it changes over time. It is the calibrated torsional amplitude value, derived from the torsional amplitude calibration output; It is a Hamming window function; T is the window length; It is a time shift parameter; is the frequency; j is a mathematical constant representing the imaginary unit, which is the standard symbol for frequency domain transformation operations; The wavelet soft thresholding secondary denoising process is as follows:

[0014]

[0015] in, It performs wavelet soft thresholding denoising on the torsional amplitude value after short-time Fourier transform. Its function is to suppress frequency domain noise, retain the resonant main frequency component, and improve the frequency extraction accuracy. It is the torsional amplitude value after short-time Fourier transform; It is a threshold; It is the historical noise variance, which comes from N is the number of signal sampling points, determined by the window length T and the sampling rate; The engine spectrum peak extraction process is as follows:

[0016] in, It is the peak value of the engine spectrum, and its function is to extract the frequency component with the strongest energy as the candidate resonant frequency; It is the torsional amplitude value after wavelet soft thresholding denoising; It is the starting point of the k-th time window; T is the window length, which is consistent with the STFT window length.

[0017] Preferably, based on the standard parameters of the same model engine and the engine's peak spectral density, the engine resonant frequency is calculated as follows: The calculation process for the engine resonance frequency is as follows:

[0018]

[0019] in, The system's inherent frequency serves to provide a theoretical reference for the inherent frequency, used to verify the rationality of the measured peak frequency. It is the equivalent torsional stiffness, which is obtained by material mechanics test to obtain the stiffness value of the damper rubber at a specific temperature and operating frequency, and then obtained after standardization. It is the equivalent moment of inertia, which is directly measured through a torsional pendulum experiment of a physical component; It is the engine resonance frequency. The purpose of the verification formula is to filter out abnormal peaks and ensure the physical rationality of the resonance frequency. It is the resonant frequency threshold, taken from .

[0020] Preferably, based on the STL time-series decomposition method, a threshold model for the operating state of the vibration damper is constructed, including: The process of constructing a vibration damper operating state threshold model using STL time series decomposition is as follows: Obtain raw load torque timing data from engine parameter set B. ,

[0021] in, It is the raw load torque time series data, which comes from engine parameter set B. Its function is to separate the long-term trend, temperature-related periodicity and random fluctuations of load torque. It is a trend component, dominated by load torque; It is a periodic component, affected by the periodic temperature of the engine oil; These are residual components used to calculate the threshold; Output component sequence { }, { }, { This is directly used as input for calculating the hierarchical threshold; The process of calculating the prediction error for each component sequence is as follows:

[0022] in, It is the prediction error, and its function is to quantify the prediction deviation of each component and provide a basis for threshold calculation. These are component prediction values, extrapolated using LOESS; This is the actual component value; The process of performing exponential smoothing on the error sequence is as follows:

[0023] in, It is a smoothed error, its function is to eliminate random fluctuations and extract a stable error trend, initial value. Derived from the statistical values ​​of the initial sampling segment; It is a smoothing factor; The process of calculating the threshold of each component based on the smoothed error sequence is as follows:

[0024]

[0025] in, It is the component threshold, and its function is to define the boundary of abnormal fluctuations of each component; It is the smoothed error. The mean; It is the smoothed error. Standard deviation; It is the threshold adjustment coefficient, which dynamically adjusts the threshold strictness according to load / temperature conditions. It is the rated load torque, from ; It is the real-time load torque, from dataset B; It is the engine oil temperature, from dataset B; The process of synthesizing the operating state threshold of the vibration damper is as follows:

[0026] in, It is the operating state threshold of the shock absorber, and its function is to integrate multi-component information to form a comprehensive state index.

[0027] Preferably, the damper damping state is obtained by comparing the engine resonance frequency with the damper operating state threshold, including: The rules for determining the damping state of a shock absorber are as follows:

[0028] The function of this determination formula is to determine the damping state of the shock absorber based on the dynamic threshold. It is the resonant frequency; It is the ideal frequency calibrated on the test bench; , , It is the damping compensation coefficient calibrated by bench testing. 0.3, 0.15, 0.3.

[0029] Preferably, the torsional vibration change rate is calculated based on the engine speed and torsional vibration amplitude, including: Torsional vibration rate The calculation process is as follows:

[0030] in, It is the torsional vibration rate of change, and its function is to eliminate the influence of rotational speed and focus on the characteristics of torsional vibration energy change. It is the instantaneous rate of change of torsional amplitude, reflecting the sudden change in vibration energy; It is the engine's real-time speed; It is a speed normalization factor that eliminates the influence of speed on the rate of change. It is the rated torsional vibration amplitude under operating conditions, from .

[0031] Preferably, the operating state of the shock absorber is obtained by comparing the torsional vibration rate with the standard parameters of the same model engine: The process for determining the working status of the shock absorber is as follows: Torsional vibration rate Standard parameters of the same model engine In , The operating status of the shock absorber was obtained by comparison:

[0032]

[0033] in, It is the real-time normalized rate of change, which quantifies the degree of deviation between the current state and the benchmark. It is the baseline rate of change, which is the bench test calibration value; It is the maximum allowable deviation, from Its function is to determine whether the shock absorber is working properly; The rules for classifying abnormal states are as follows:

[0034] Minor anomalies are recorded by the system without alarm; moderate anomalies trigger a yellow warning on the dashboard; and severe anomalies trigger a red alarm and torque limiting control. This formula quantifies the severity of the anomaly and triggers a graded response mechanism.

[0035] Preferably, based on the cumulative damage theory, the vibration damper life prediction model includes: The process of constructing the vibration damper life prediction model is as follows: Input the pre-processed engine parameter set B, damper damping status, damper operating status, and damper service life. Standard parameter set ; The initial damage factor construction process is as follows: Input vibration damper service life and Standard service life of vibration dampers ;

[0036] in, It is the initial damage factor, and its function is to convert the percentage of running time into the initial damage value; The shock absorber has reached the end of its service life. This is the standard service life of the shock absorber, taken from... ; SN curve temperature correction: First, calculate the corrected fatigue constant, and then perform temperature correction on the SN curve; The calculation process for the temperature correction factor is as follows:

[0037] in, It is a temperature correction factor, which is used to quantify the effect of temperature on the fatigue life of materials. It is the material activation energy, that is, the activation energy of the rubber layer material of the core elastic element in the torsional vibration damper. It is a predetermined parameter for the rubber material of the damper, which is determined in advance through dynamic thermomechanical analysis test. It is the gas constant; It is the engine oil temperature, from dataset B; The physical meaning is to convert the measured Celsius temperature of the engine oil into the thermodynamic temperature Kelvin. The calculation process for the corrected fatigue constant is as follows:

[0038] in, It is the corrected fatigue constant, and its function is to generate the temperature-corrected material fatigue constant; It is the original fatigue constant, derived from ; It is a temperature correction factor; The process of correcting the SN curve is as follows:

[0039]

[0040] in, It is the stress amplitude, and its function is to convert torsional vibration into material stress; It is the torsional vibration-stress conversion coefficient, and the specific value was determined through previous bench tests. It is the real-time torsional vibration amplitude; It is the theoretical fatigue life, and its function is to calculate the theoretical fatigue life under the current stress. It is a material index, which refers to the core elastic element material, namely rubber, used to absorb vibration energy in torsional vibration dampers. The specific value is determined through material fatigue tests. This is the corrected fatigue constant; The real-time damage calculation process is as follows:

[0041]

[0042] in, It represents the number of cycles, and its function is to calculate the number of load cycles within the sampling period. It is the engine's real-time speed; It is the sampling interval; i is the excitation order, taken from S2, i=3 or 6; It is the damage increment, and its function is to quantify the damage increment in the current sampling period; It is the theoretical fatigue life; Life cycle injury The synthesis calculation process is as follows:

[0043]

[0044]

[0045] in, It is a full life cycle damage, and its function is to synthesize historical damage and real-time damage; It is the initial damage; It is the increase in damage; It is a damage correction coefficient, which is used to correct the accumulated damage based on the damper damping state output by S3 and the damper working state output by S4.

[0046] Preferably, the damper damping state, damper operating state, and engine parameter set are input into the damper life prediction model, and the output of the remaining damper life includes: Remaining life of the shock absorber The calculation process is as follows:

[0047] in, It is the remaining life of the vibration damper, and its function is to calculate the interpretable remaining life of the engineering. This is the standard service life; It is damage throughout the entire life cycle.

[0048] This invention provides a method for predicting the lifespan of automotive shock absorbers, involving machine learning and deep learning technologies, which has the following beneficial effects: (1) The life prediction method for automotive shock absorbers integrates real-time parameters from multiple dimensions, including engine speed, torsional vibration amplitude, oil temperature, and load torque, and combines them with the cumulative damage theory of Arrhenius temperature correction to dynamically adapt to the impact of different working conditions on the life of the shock absorbers. For example, the fatigue constant is corrected by oil temperature, and the damage increment is corrected by damping state and working state, thereby reducing the life prediction error and achieving a much higher accuracy than traditional single-parameter static prediction schemes.

[0049] (2) The life prediction method of the automobile shock absorber outputs layered maintenance suggestions through dual verification of "damping state judgment + working state classification". When the remaining life is ≥70% of the standard life, it is judged as "healthy" and no intervention is required; when it is 30%~70%, "decay warning" is triggered and maintenance is planned; when it is <30%, "immediate replacement" is prompted, avoiding cost waste or fault omission caused by traditional maintenance and reducing the average annual maintenance cost of a single vehicle.

[0050] (3) The life prediction method of the car shock absorber adapts to different working conditions by using the standard parameters of the same engine model. It does not rely on complex bench test data and directly calculates the operating parameters collected in real time by the car CAN bus without the need for additional sensors. The prediction model is compatible with short-time Fourier transform and wavelet denoising data preprocessing modules, and can be directly connected to the existing car monitoring system, simplifying the engineering implementation process and adapting to most life assessment scenarios of the same engine model shock absorber. Attached Figure Description

[0051] Figure 1 This is a flowchart of a method for predicting the lifespan of an automotive shock absorber proposed in this invention.

[0052] Figure 2 The remaining life hierarchy diagram is calculated for the life prediction method of an automotive shock absorber proposed in this invention. Detailed Implementation

[0053] 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.

[0054] Please see Figure 1-2 This invention provides a technical solution: a method for predicting the lifespan of automotive shock absorbers. Specifically, the following method for predicting the lifespan of automotive shock absorbers is provided; please refer to [link / reference]. Figure 1 The method includes the following steps: Step S1: Collect engine parameters, preprocess the engine parameters to obtain an engine parameter set, which includes engine speed, torsional vibration amplitude, oil temperature and load torque.

[0055] The engine's torsional vibration is transmitted to the damper through the drive shaft system. Its vibration energy (torsional amplitude x) and frequency characteristics (resonance frequency) are also considered. Oil temperature is the main cause of shock absorber damage. The damping characteristics can be indirectly altered by affecting the stiffness of the rubber.

[0056] Engine data acquisition and engine parameter definitions are as follows: Collect engine parameters during vehicle operation, including key physical quantities such as engine speed, torsional vibration amplitude, oil temperature, and load torque.

[0057] Suppose the collected engine parameters are in the following set. :

[0058] in The parameter vector at the i-th sampling time (e.g.) =[Speed, Torque amplitude, Oil temperature, Load torque]).

[0059] Because engine parameter A may contain sensor noise, missing values, or outliers (such as peak interference in torsional amplitude caused by rapid acceleration), directly using it will affect the accuracy of subsequent analysis. Therefore, it needs to be preprocessed using a preprocessing function. A is purified and standardized.

[0060] Construct the preprocessing function and define the preprocessing function. Its core operations include: noise filtering, which uses sliding window mean filtering to suppress high-frequency noise; missing value filling, which fills in the data based on time series linear interpolation; and normalization, which eliminates dimensional differences (such as mapping rotational speed to the [0,1] interval).

[0061] The preprocessing procedure is as follows:

[0062] in, This is the preprocessed engine parameter set. This is the standardized parameter vector.

[0063] The preprocessed engine parameter set B has the following data quality requirements: engine speed reflects the frequency of the vibration excitation source; torsional amplitude retains the effective vibration energy characteristics; and oil temperature / load torque characterizes the operating load level.

[0064] This step involves collecting engine speed, torsional vibration amplitude, oil temperature, and load torque engine parameters, and preprocessing the data (including noise reduction, missing value imputation, and standardization) to generate a high-quality engine parameter set. This provides a clean and uniform data foundation for subsequent steps, ensuring the reliability of the input for resonance frequency extraction (S2) and state analysis (S3-S4).

[0065] Step S2: Collect standard parameters of the same model engine, process the torsional vibration amplitude of the engine parameter set using the short-time Fourier transform method to obtain the engine spectrum peak value; calculate the engine resonance frequency based on the standard parameters of the same model engine and the engine spectrum peak value.

[0066] Traditional denoising algorithms (such as Fourier transform and mean filtering) lack time-frequency local analysis capabilities, making them unsuitable for time-varying signals like torsional vibration amplitude, or prone to over-smoothing and loss of key vibration details. In contrast, short-time Fourier transform can simultaneously analyze the time and frequency characteristics of a signal in the time-frequency domain, accurately locating periods of abnormal vibration. Wavelet soft-threshold denoising can adaptively distinguish between signal and noise, suppressing random electromagnetic interference while fully preserving key features such as damper damping changes and resonance mutations. This significantly improves the extraction accuracy of torsional vibration amplitude and engine resonance frequency, providing more reliable basic data for subsequent damper life prediction and effectively addressing the shortcomings of traditional algorithms in terms of insufficient time-frequency resolution and loss of details.

[0067] The data source is defined as follows: The data sources collected in this step mainly include the preprocessed engine parameter set B output from step S1 and the standard parameters of the same model engine.

[0068] The following is a set of standard parameters for the same engine model retrieved from the historical database:

[0069] in, It is the rated speed; It is the calibrated torsional vibration range; It is the resonant frequency threshold; It is the historical noise variance; It is the benchmark value for judging the damping state, which is obtained through statistical analysis of historical bench test data and normal real vehicle operation data; The maximum allowable deviation is obtained through statistical analysis of the fluctuation characteristics of the same set of historical normal data; a benchmark value for judging the damping state is defined. The upper limit of the reasonable fluctuation range; It's tolerance; It is the target torsional amplitude; This refers to the standard service life of the vibration damper. It is damage throughout the entire life cycle.

[0070] The joint data processing flow is as follows: First, torsional vibration amplitude is calibrated, then short-time Fourier transform is performed, and finally wavelet soft thresholding is used for secondary denoising.

[0071] Torsional amplitude calibration (real-time torsional amplitude value) Compared with standard parameters of (Comparison)

[0072] in, It is the calibrated torsional vibration amplitude, which is used to eliminate individual differences in engines and scale the real-time torsional vibration signal to the standard amplitude range. It is the real-time torsional vibration amplitude value, which comes from the engine parameter set B output by S1; It is the upper limit of the torsional vibration range in the standard parameters, derived from ; It is the lower limit of the torsional vibration range in the standard parameters, derived from... ; It is the standard rated torsional vibration amplitude, based on historical bench test data, used to characterize the typical torsional vibration amplitude of a specific engine model under standard rated operating conditions. Its purpose is to eliminate individual differences and ensure that the signal amplitude conforms to the model reference.

[0073] The short-time Fourier transform process is as follows:

[0074]

[0075] in, It performs a short-time Fourier transform on the calibrated torsional vibration amplitude, which converts the time-domain vibration signal into a time-frequency domain representation to locate the resonant frequency as it changes over time. It is the calibrated torsional amplitude value, derived from the torsional amplitude calibration output; It is a Hamming window function; T is the window length; It is a time shift parameter; is the frequency; j is a mathematical constant representing the imaginary unit, and is the standard symbol for frequency domain transformation operations.

[0076] The wavelet soft thresholding secondary denoising process is as follows (wavelet thresholding is superimposed in the frequency domain):

[0077]

[0078] in, It performs wavelet soft thresholding denoising on the torsional amplitude value after short-time Fourier transform. Its function is to suppress frequency domain noise, retain the resonant main frequency component, and improve the frequency extraction accuracy. It is the torsional amplitude value after short-time Fourier transform; It is a threshold; It is the historical noise variance, which comes from N is the number of signal sampling points, which is determined by the window length T and the sampling rate.

[0079] The engine spectrum peak extraction process is as follows:

[0080] in, It is the peak value of the engine spectrum, and its function is to extract the frequency component with the strongest energy as the candidate resonant frequency; It is the torsional amplitude value after wavelet soft thresholding denoising; It is the starting point of the k-th time window; T is the window length, which is consistent with the STFT window length.

[0081] The calculation process for the engine resonance frequency is as follows (verification) (Whether it is within the valid range):

[0082]

[0083] in, The system's inherent frequency serves to provide a theoretical reference for the inherent frequency, used to verify the rationality of the measured peak frequency. It is the equivalent torsional stiffness, which is obtained by material mechanics test to obtain the stiffness value of the damper rubber at a specific temperature and operating frequency, and then obtained after standardization. It is the equivalent moment of inertia, which is directly measured through a torsional pendulum experiment of a physical component; It is the engine's resonant frequency. The purpose of the verification formula is to filter out abnormal peaks (such as noise interference) and ensure the physical rationality of the resonant frequency. It is the resonant frequency threshold, taken from .

[0084] when At that time, the engine spectrum peak value As the inherent frequency of the system; when Then, extend the window length T of the short-time Fourier transform, and use the new window length T to re-process the original time-domain vibration signal. Perform a short-time Fourier transform to obtain new spectral peaks.

[0085] The engine excitation correlation process is as follows (outputting the final engine resonant frequency and correlation order):

[0086] in, is the engine's excitation frequency, which serves as a theoretical calculation value to match with the actual measured vibration signal, thereby ultimately determining the engine's resonant frequency; the value of i is 3 or 6, which is the principal order of the six-cylinder engine. This formula calculates the 3rd and 6th principal harmonics of the six-cylinder engine.

[0087] Order matching must satisfy the following formula:

[0088] in, It is tolerance, taken from , is a preset value directly called, determined through extensive preliminary experiments and data analysis, taking into account the vibration characteristics of this engine model, the accuracy of the measurement system, and the allowable frequency deviation range; n is the real-time engine speed, in rpm revolutions per minute, which is substituted into the formula. The calculation shows that the physical meaning of this formula is to convert the rotational speed into the corresponding excitation frequency, based on real-time data from engine parameter set B. It is a class; It is the resonant frequency.

[0089] This step integrates the engine parameter set B (including real-time speed, denoised torsional vibration amplitude, oil temperature, and load torque) preprocessed in step S1 with the factory standard parameters of the same engine model (including rated speed, calibrated torsional vibration range, and resonant frequency threshold). Short-time Fourier Transform (STFT) is applied for time-frequency analysis and Hamming window segmentation. Combined with wavelet soft thresholding secondary denoising technology, high-frequency noise interference is eliminated, ultimately accurately calculating the resonant frequencies (e.g., 70Hz, 150Hz) characterizing the matching relationship between engine excitation and the transmission system. This result provides a key engine resonant frequency input for the subsequent damper operating status assessment in step S3, establishing a theoretical benchmark for damper life prediction.

[0090] Step S3: Based on the STL time-series decomposition method, construct the shock absorber operating state threshold model; input the engine oil temperature and load torque into the shock absorber operating state threshold model, and output the shock absorber operating state threshold; by comparing the engine resonance frequency and the shock absorber operating state threshold, obtain the shock absorber damping state.

[0091] Input the preprocessed engine parameter set Based on engine parameter set The damping state of the damper is determined by the damper operating state threshold output from the damper operating state threshold model.

[0092] The process of constructing a vibration damper operating state threshold model using STL time series decomposition is as follows: Obtain raw load torque timing data from engine parameter set B. ,

[0093] in, It is the raw load torque time series data, which comes from engine parameter set B. Its function is to separate the long-term trend, temperature-related periodicity and random fluctuations of load torque. It is a trend component, dominated by load torque; It is a periodic component, affected by the periodic temperature of the engine oil; It is the residual component, used to calculate the threshold.

[0094] Output component sequence { }, { }, { It can be directly used as input for calculating the stratification threshold.

[0095] The process of calculating the prediction error for each component sequence is as follows:

[0096] in, It is the prediction error, and its function is to quantify the prediction deviation of each component and provide a basis for threshold calculation. These are component prediction values, extrapolated using LOESS; This is the actual component value.

[0097] The process of performing exponential smoothing on the error sequence is as follows:

[0098] in, It is a smoothed error, its function is to eliminate random fluctuations and extract a stable error trend, initial value. Derived from the statistical values ​​of the initial sampling segment; It is a smoothing factor. =0.2.

[0099] The process of calculating the threshold of each component based on the smoothed error sequence is as follows:

[0100]

[0101] in, It is the component threshold, and its function is to define the boundary of abnormal fluctuations of each component; It is the smoothed error. The mean; It is the smoothed error. Standard deviation; It is the threshold adjustment coefficient, which dynamically adjusts the threshold strictness according to load / temperature conditions. It is the rated load torque, from ; It is the real-time load torque, from dataset B; It is the engine oil temperature, from dataset B.

[0102] when When the real-time load torque is greater than the rated load torque and the temperature is less than 40 degrees Celsius, the threshold adjustment coefficient is... 3.5; when , When the real-time load torque is greater than the rated load torque and the temperature is not less than 40 degrees Celsius, the threshold adjustment coefficient is... 3; when , That is, when the temperature is less than 40 degrees Celsius and the real-time load torque is not greater than the rated load torque, the threshold adjustment coefficient is... 2; Threshold adjustment coefficient for remaining cases 2.5.

[0103] The process of synthesizing the operating state threshold of the vibration damper is as follows:

[0104] in, It is the operating state threshold of the shock absorber, and its function is to integrate multi-component information to form a comprehensive state index.

[0105] Traditional methods often rely on single parameters (such as torsional vibration amplitude alone) or static models to determine the state of shock absorbers, ignoring multi-parameter coupling and dynamic changes in operating conditions, which can easily lead to errors. This new operating state threshold model, based on historical data and simultaneously linking engine resonance frequency with multi-dimensional parameters of the shock absorber's operating state, can accurately capture the coupling and interaction relationships between parameters. Combined with dynamic threshold comparison, it can adapt to the parameter linkage characteristics under different operating conditions (such as load fluctuations and temperature changes), significantly improving the accuracy of shock absorber operating state determination and providing more reliable support for calculating torsional vibration rate of change. This effectively solves the problems of insufficient accuracy and poor adaptability caused by traditional methods due to single parameters and static determination.

[0106] The process for determining the damping state of a shock absorber is as follows: By the engine resonant frequency and vibration damper operating status threshold Compare them.

[0107] The formula for calculating the adaptive factor is as follows:

[0108] in, It is an adaptive factor, whose function is to convert the shock absorber operating state threshold into a relative quantity for state determination; It is a preset baseline threshold.

[0109] The rules for determining the damping state of a shock absorber are as follows:

[0110] The function of this determination formula is to determine the damping state of the shock absorber based on the dynamic threshold. It is the resonant frequency; It is the ideal frequency calibrated on the test bench; , , It is the damping compensation coefficient calibrated by bench testing. 0.3, 0.15, 0.3.

[0111] when When the damping force of the shock absorber is too large, it is determined to be an overdamped state, which causes the vibration of the system to decay too quickly and the resonance frequency to deviate from the "ideal value" and be higher than the ideal resonance frequency. when At that time, it was determined to be In this state, the damper damping force is within the optimal design range, the system vibration attenuation characteristics best meet expectations, and the deviation of the resonant frequency from the "ideal value" is within a small allowable range. when At that time, it was determined to be If the damping force of the shock absorber is too small, the vibration of the system will decay too slowly, and the resonance frequency will deviate from the "ideal value" and be lower than the ideal resonance frequency.

[0112] This step uses STL time-series decomposition to decompose the preprocessed engine parameter set B (including oil temperature and load torque) into trend components, periodic components, and residual components. Based on the smoothed sequence of prediction errors of each component, hierarchical dynamic thresholds, including trend thresholds, periodic thresholds, and residual thresholds, are calculated. Finally, the shock absorber operating state thresholds are weighted and synthesized to provide a dynamic threshold benchmark for subsequent shock absorber damping state determination, thereby achieving anomaly detection capability that adapts to operating conditions.

[0113] Step S4: Calculate the torsional vibration change rate based on the engine speed and torsional vibration amplitude; compare the torsional vibration change rate with the standard parameters of the same model engine to obtain the working state of the shock absorber.

[0114] Input the preprocessed engine parameter set B (including the engine's real-time speed n and real-time torsional vibration amplitude x(t)), and output the damper damping state and standard parameters of the same engine model from step S3. ={ , }, It is the benchmark value for judging the damping state. This is the maximum allowable deviation.

[0115] First rate of change The calculation process is as follows:

[0116] in, It refers to the time point from idle speed to operating speed, corresponding to the torque at speeds below 700 rpm. ; It refers to the torque stabilization point, determined by the torque when the speed fluctuation is less than 5% for 10 seconds. ; It is the torsional amplitude value at idle speed; It is the torsional vibration amplitude at the stable operating point; the purpose of this formula is to detect structural anomalies in large-span torque changes and prevent mechanical failure.

[0117] The decision logic is when the first rate of change... If the value exceeds the structural threshold (the structural threshold is a critical value determined in advance through experiments or simulations, representing the upper limit of the rate of change of vibration that the torsional vibration damper can withstand on the mechanical structure), a structural failure alarm will be issued; otherwise, steady-state detection will be performed.

[0118] The process for confirming the steady-state operating point is as follows (calculating the steady-state volatility). ):

[0119] in, and It is the real-time torsional amplitude value of adjacent sampling points; It is the sampling time interval, determined by the system sampling rate; the formula is used to confirm whether the system has entered a stable working state, providing a basis for subsequent analysis.

[0120] The criterion is that when the steady-state volatility... When the value is less than the steady-state threshold, which is obtained from a large amount of data analysis, the output and target torsional amplitude values ​​are... .

[0121] Torsional vibration rate The calculation process is as follows:

[0122] in, It is the torsional vibration rate of change, and its function is to eliminate the influence of rotational speed and focus on the characteristics of torsional vibration energy change. It is the instantaneous rate of change of torsional amplitude, reflecting the sudden change in vibration energy; It is the engine's real-time speed; It is a speed normalization factor that eliminates the influence of speed on the rate of change. The rated operating condition torsional vibration amplitude refers to the representative standard torsional vibration amplitude measured through a large number of previous bench tests when the engine of this model is running stably under standard rated operating conditions.

[0123] The adaptive correction process for the damper's damping state is as follows (reference value) (Dynamically adjusted)

[0124] in, It is a correction reference value, and its function is to dynamically adjust and determine the reference based on the damping state of the shock absorber. and It is the damping compensation coefficient calibrated by bench testing; , and It is the damped state, which comes from the output of step S3.

[0125] The process for determining the working status of the shock absorber is as follows: Torsional vibration rate Standard parameters of the same model engine In , The working status of the vibration damper is obtained by comparison.

[0126]

[0127]

[0128] in, It is the real-time normalized rate of change, which quantifies the degree of deviation between the current state and the benchmark. It is the baseline rate of change, which is the bench test calibration value; It is the maximum allowable deviation, from Its function is to determine whether the shock absorber is working properly.

[0129] The abnormal state classification rules are as follows (when the shock absorber is in an abnormal working state, the following state classification is applied):

[0130] Minor anomalies are recorded by the system without alarm; moderate anomalies trigger a yellow warning on the dashboard; and severe anomalies trigger a red alarm and torque limiting control. This formula quantifies the severity of the anomaly and triggers a graded response mechanism.

[0131] The functional failure linkage detection process is as follows:

[0132] in, It is the real-time torsional vibration amplitude; This is the target torsional amplitude value, which is the bench-calibrated torsional amplitude value of the same model engine at rated speed, taken from... This formula is used to detect the deviation between the actual torsional vibration and the target, and to prevent functional failure.

[0133] when If the value is greater than 0.2 (torsional vibration threshold, which is determined based on statistical analysis of P3 parameters in historical bench test data and actual vehicle operation data of this engine model. Specifically, under rated operating conditions, P3 data under normal operating conditions are collected and a probability distribution is formed, and finally a high quantile of the distribution is selected as the threshold), a functional failure alarm is triggered; otherwise, life prediction is performed.

[0134] This step is based on the preprocessed engine parameter set B (including engine speed and torsional vibration amplitude). It calculates the torsional vibration change rate (reflecting the instantaneous change of vibration energy) and compares and analyzes this change rate with the standard parameters of the same engine model. Finally, it outputs the working status of the shock absorber (such as normal or abnormal), providing a real-time working condition judgment basis for the shock absorber life prediction in the subsequent step S5.

[0135] Step S5: Based on the cumulative damage theory, construct a shock absorber life prediction model; input the shock absorber damping state, shock absorber working state and the engine parameter set into the shock absorber life prediction model, and output the remaining life of the shock absorber.

[0136] The shock absorber's service life was read from the ECU via the CAN bus. .

[0137] The process of constructing the vibration damper life prediction model is as follows: Input the preprocessed engine parameter set B (including oil temperature) Real-time torsional amplitude x(t)), damper damping status (S3 output), damper operating status (S4 output), damper service life. Standard parameter set .

[0138] The initial damage factor construction process is as follows: Input vibration damper service life and Standard service life of vibration dampers .

[0139]

[0140] in, It is the initial damage factor, and its function is to convert the percentage of running time into the initial damage value; The shock absorber has reached the end of its service life. This is the standard service life of the shock absorber, taken from... .

[0141] Temperature correction of SN curve (Arrhenius model embedding): First, the corrected fatigue constant is calculated, and then the SN curve is corrected for temperature.

[0142] The calculation process for the temperature correction factor is as follows:

[0143] in, It is a temperature correction factor, which is used to quantify the effect of temperature on the fatigue life of materials. It is the material activation energy, that is, the activation energy of the rubber layer material of the core elastic element in the torsional vibration damper. It is a predetermined parameter for the rubber material of the damper, which is determined in advance through dynamic thermomechanical analysis test. It is the gas constant, a fundamental physical constant in the Arrhenius equation; It is the engine oil temperature, from dataset B; The physical meaning is to convert the measured Celsius temperature of the engine oil into the thermodynamic temperature Kelvin temperature.

[0144] The calculation process for the corrected fatigue constant is as follows:

[0145] in, It is the corrected fatigue constant, and its function is to generate the temperature-corrected material fatigue constant; It is the original fatigue constant; It is the temperature correction factor.

[0146] The process of correcting the SN curve is as follows:

[0147]

[0148] in, It is the stress amplitude, and its function is to convert torsional vibration into material stress; It is the torsional vibration-stress conversion coefficient, and the specific value was determined through previous bench tests. It is the real-time torsional vibration amplitude; It is the theoretical fatigue life, and its function is to calculate the theoretical fatigue life under the current stress. It is a material index, which refers to the core elastic element material, namely rubber, used to absorb vibration energy in torsional vibration dampers. The specific value is determined through material fatigue tests. It is the corrected fatigue constant.

[0149] The real-time damage calculation process is as follows:

[0150]

[0151] in, It represents the number of cycles, and its function is to calculate the number of load cycles within the sampling period. It is the engine's real-time speed; It is the sampling interval; i is the excitation order, taken from S2, i=3 or 6; It is the damage increment, and its function is to quantify the damage increment in the current sampling period; It is the theoretical fatigue life.

[0152] Life cycle injury The synthesis calculation process is as follows:

[0153]

[0154]

[0155] in, It is a full life cycle damage, and its function is to synthesize historical damage and real-time damage; This refers to the initial damage. For a brand new shock absorber, the initial damage is zero, i.e. =0. For in-service vibration dampers. This value represents the total damage accumulated during the previous operating cycle of the vibration damper. This value will be finalized when the system is shut down or under maintenance. Write into the standard parameter set of the same engine model Save it in the middle; when restarted, it will be saved from the middle. The value is used as the initial damage for this prediction. ; It is the increase in damage; It is a damage correction coefficient, which is used to correct the accumulated damage based on the damper damping state output by S3 and the damper working state output by S4.

[0156] Remaining life of the shock absorber The calculation process is as follows:

[0157] in, It is the remaining life of the vibration damper, and its function is to calculate the interpretable remaining life of the engineering. This is the standard service life; It is damage throughout the entire life cycle.

[0158] Remaining life of output vibration damper .

[0159] Traditional vibration damper life prediction often ignores the impact of temperature on material fatigue and relies on static SN curves without considering actual operating conditions, leading to large deviations. This new prediction model offers significant advantages: it introduces the Arrhenius model to dynamically correct the impact of temperature on material fatigue characteristics, adapting to different temperature conditions; based on SN curves and Miner's cumulative damage theory, it accurately quantifies damage accumulation under cyclic loading; and by combining vibration damper operating time, preset time, and engine operating status, it dynamically calculates remaining life, greatly improving prediction accuracy and adaptability to operating conditions. This provides a reliable basis for precise vibration damper maintenance and overcomes the shortcomings of traditional models, such as insufficient accuracy and poor adaptability to operating conditions.

[0160] The rules for determining the health status of shock absorbers are as follows (providing actionable maintenance decisions):

[0161] when If the shock absorber's performance degradation is minimal, it remains within its optimal operating range and provides sufficient assurance for vehicle stability and comfort. No additional testing or maintenance is required; normal use can continue. when When this occurs, it indicates that the shock absorber's performance has shown detectable degradation (such as decreased damping stability and weakened resonance suppression effect), but has not yet reached the level of "endangering driving safety". Early warning and preventive maintenance are required. Try to avoid driving on potholed roads and heavy off-road conditions for a long time to reduce additional impact on the shock absorber and slow down the rate of performance degradation. when When the damper performance has severely deteriorated (damping failure, abnormal vehicle vibration / noise, and "bouncing out of control" when going over potholes may occur), continued use will significantly reduce driving stability and even endanger driving safety. Choose dampers that match the original specifications and performance (prioritize original parts or verified brand replacements) to ensure that the damping characteristics and load-bearing capacity are consistent with the original vehicle design.

[0162] This step, based on the damper's operational life and standard service life, quantifies historical damage using the Arrhenius temperature correction model, calculates the temperature acceleration factor using real-time oil temperature, dynamically corrects the remaining life, and finally outputs an engineering-interpretable remaining life value and maintenance judgment conclusion (health / deterioration warning / immediate replacement), achieving accurate life prediction independent of load spectrum dependence.

[0163] This invention achieves accurate lifespan assessment of automotive shock absorbers through multi-dimensional data fusion and dynamic model correction. The method first collects engine speed, torsional vibration amplitude, and oil temperature parameters in real time, capturing time-frequency features using short-time Fourier transform and combining this with wavelet soft thresholding to preserve key vibration details. Then, it uses a binary regression model to uncover parameter coupling relationships and constructs dynamic operating thresholds based on STL time-series decomposition to accurately determine damping and operating status. The core functionality involves adapting to operating temperature differences using an Arrhenius temperature correction model and quantifying cyclic load damage using Miner's cumulative damage theory to dynamically calculate remaining lifespan. Finally, it outputs graded maintenance recommendations for health status, degradation warnings, and immediate replacement. Compared to traditional static threshold methods, this method reduces prediction errors, lowers the average annual maintenance cost per vehicle, requires no additional sensors, and is suitable for precise maintenance needs under complex operating conditions.

[0164] 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 a process, method, article, or apparatus. Without further limitations, the phrase "comprising an element defined as..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0165] 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 likenesses.

Claims

1. A method of predicting the life of an automobile shock absorber, characterized by, Includes the following steps: Step S1: Collect engine parameters, preprocess the engine parameters to obtain an engine parameter set, which includes engine speed, torsional vibration amplitude, oil temperature and load torque; Step S2: Collect standard parameters of the same model engine, process the torsional vibration amplitude of the engine parameter set using the short-time Fourier transform method to obtain the engine spectrum peak value; calculate the engine resonance frequency based on the standard parameters of the same model engine and the engine spectrum peak value. Step S3: Based on the STL time-series decomposition method, construct a shock absorber operating state threshold model; input the engine oil temperature and load torque into the shock absorber operating state threshold model, and output the shock absorber operating state threshold; by comparing the engine resonance frequency and the shock absorber operating state threshold, obtain the shock absorber damping state; Step S4: Calculate the torsional vibration change rate based on the engine speed and torsional vibration amplitude; compare the torsional vibration change rate with the standard parameters of the same model engine to obtain the working state of the shock absorber; Step S5: Based on the cumulative damage theory, construct a shock absorber life prediction model; input the shock absorber damping state, shock absorber working state and the engine parameter set into the shock absorber life prediction model, and output the remaining life of the shock absorber.

2. The method for predicting the lifespan of an automotive shock absorber according to claim 1, characterized in that, Engine parameters are collected and preprocessed to obtain an engine parameter set, including: Collect engine parameters during vehicle operation, including engine speed, torsional vibration amplitude, oil temperature, and load torque; Suppose the collected raw engine parameters are in the following set. : ； in This represents the parameter vector at the nth sampling time, where n is the number of samples; Because the engine parameter set A may contain sensor noise, missing values, or outliers, using it directly will affect the accuracy of subsequent analysis; therefore, it needs to be preprocessed using a preprocessing function. Purify and standardize A; Construct the preprocessing function and define the preprocessing function. Its core operations include noise filtering, which uses sliding window mean filtering to suppress high-frequency noise; missing value imputation, which uses linear interpolation based on time series to complete the data; and normalization, which eliminates differences in units. The preprocessing procedure is as follows: ； in, This is the preprocessed engine parameter set. This is the standardized parameter vector.

3. The method for predicting the lifespan of an automotive shock absorber according to claim 2, characterized in that, Standard parameters of engines of the same model are collected, and the torsional vibration amplitude values ​​in the parameter set are processed using the short-time Fourier transform method to obtain the engine spectrum peak values, including: The following is a set of standard parameters for the same engine model retrieved from the historical database: ； in, It is the rated speed; It is the calibrated torsional vibration range; It is the resonant frequency threshold; It is the historical noise variance; It is the benchmark value for judging the damping state, which is obtained through statistical analysis of historical bench test data and normal real vehicle operation data; The maximum allowable deviation is obtained through statistical analysis of the fluctuation characteristics of the same set of historical normal data; a benchmark value for judging the damping state is defined. The upper limit of the reasonable fluctuation range; It's tolerance; It is the target torsional amplitude; This refers to the standard service life of the shock absorber. It is damage throughout the entire life cycle; It is the rated load torque; It is the rated torsional vibration amplitude under operating conditions; It is the original fatigue constant; The joint data processing flow is as follows: First, torsional vibration amplitude is calibrated, then short-time Fourier transform is performed, and finally wavelet soft thresholding is used for secondary denoising. Torsional amplitude calibration: ； in, It is the calibrated torsional vibration amplitude, which is used to eliminate individual differences in engines and scale the real-time torsional vibration signal to the standard amplitude range. It is the real-time torsional vibration amplitude value, which comes from the engine parameter set B output by S1; It is the upper limit of the torsional vibration range in the standard parameters, derived from ; It is the lower limit of the torsional vibration range in the standard parameters, derived from... ; It is the standard rated torsional vibration amplitude, based on historical bench test data, used to characterize the typical torsional vibration amplitude of a specific engine model under standard rated operating conditions; its function is to eliminate individual differences and ensure that the signal amplitude conforms to the model reference. The short-time Fourier transform process is as follows: ； ； in, It is the torsional amplitude value after short-time Fourier transform, and its function is to convert the time-domain vibration signal into a time-frequency domain representation to locate the resonant frequency as it changes over time. It is the calibrated torsional amplitude value, derived from the torsional amplitude calibration output; It is a Hamming window function; T is the window length; It is a time shift parameter; is the frequency; j is a mathematical constant representing the imaginary unit, which is the standard symbol for frequency domain transformation operations; The wavelet soft thresholding secondary denoising process is as follows: ； ； in, It is the torsional amplitude value after wavelet soft thresholding denoising. Its function is to suppress frequency domain noise, retain the resonant main frequency component, and improve the frequency extraction accuracy. It is the torsional amplitude value after short-time Fourier transform; It is a threshold; It is the historical noise variance, which comes from N is the number of signal sampling points, determined by the window length T and the sampling rate; The engine spectrum peak extraction process is as follows: ； in, It is the peak value of the engine spectrum, and its function is to extract the frequency component with the strongest energy as the candidate resonant frequency; It is the torsional amplitude value after wavelet soft thresholding denoising; It is the starting point of the k-th time window; T is the window length, which is consistent with the STFT window length.

4. The method for predicting the lifespan of an automotive shock absorber according to claim 3, characterized in that, Based on the standard parameters and peak values ​​of the same engine model, the engine resonant frequency is calculated, including: The calculation process for the engine resonance frequency is as follows: ； ； in, The system's inherent frequency serves to provide a theoretical reference for the inherent frequency, used to verify the rationality of the measured peak frequency. It is the equivalent torsional stiffness, which is obtained by material mechanics testing to determine the stiffness value of the damper rubber at a specific temperature and operating frequency, and then standardized. It is the equivalent moment of inertia, which is directly measured through a torsional pendulum experiment of a physical component; It is the engine resonance frequency. The purpose of the verification formula is to filter out abnormal peaks and ensure the physical rationality of the resonance frequency. It is the resonant frequency threshold, taken from .

5. The method for predicting the lifespan of an automotive shock absorber according to claim 4, characterized in that, Based on the STL time-series decomposition method, a threshold model for the operating state of the vibration damper is constructed, including: The process of constructing a vibration damper operating state threshold model using STL time series decomposition is as follows: Obtain raw load torque timing data from engine parameter set B. , ； in, It is the raw load torque time series data, which comes from engine parameter set B. Its function is to separate the long-term trend, temperature-related periodicity and random fluctuations of load torque. It is a trend component, dominated by load torque; It is a periodic component, affected by the periodic temperature of the engine oil; These are residual components used to calculate the threshold. Output component sequence { }, { }, { This is directly used as input for calculating the hierarchical threshold; The process of calculating the prediction error for each component sequence is as follows: ； in, It is the prediction error, and its function is to quantify the prediction deviation of each component and provide a basis for threshold calculation. These are component prediction values, extrapolated using LOESS; This is the actual component value; The process of performing exponential smoothing on the error sequence is as follows: ； in, It is a smoothed error, its function is to eliminate random fluctuations and extract a stable error trend, initial value. Derived from the statistical values ​​of the initial sampling segment; It is a smoothing factor; The process of calculating the threshold of each component based on the smoothed error sequence is as follows: ； ； in, It is the component threshold, and its function is to define the boundary of abnormal fluctuations of each component; It is the smoothed error. The mean; It is the smoothed error. Standard deviation; It is the threshold adjustment coefficient, which dynamically adjusts the threshold strictness according to the load / temperature conditions. It is the rated load torque, from ; It is the real-time load torque, from dataset B; It is the engine oil temperature, from dataset B; The process of synthesizing the operating state threshold of the vibration damper is as follows: ； in, It is the operating state threshold of the shock absorber, and its function is to integrate multi-component information to form a comprehensive state index.

6. The method for predicting the lifespan of an automotive shock absorber according to claim 5, characterized in that, The damper damping state is obtained by comparing the engine resonance frequency with the damper operating state threshold, including: The rules for determining the damping state of a shock absorber are as follows: ； The function of this determination formula is to determine the damping state of the shock absorber based on the dynamic threshold. It is the engine's resonant frequency; It is the ideal frequency calibrated on the test bench; , , It is the damping compensation coefficient calibrated by bench testing. 0.3, 0.15, 0.3; It is an adaptive factor.

7. The method for predicting the lifespan of an automotive shock absorber according to claim 6, characterized in that, Based on the engine speed and torsional amplitude, the torsional vibration rate of change is calculated, including: Torsional vibration rate The calculation process is as follows: ； in, It is the torsional vibration rate of change, and its function is to eliminate the influence of rotational speed and focus on the characteristics of torsional vibration energy change. It is the instantaneous rate of change of torsional amplitude, reflecting the sudden change in vibration energy; It is the engine's real-time speed; It is a speed normalization factor that eliminates the influence of speed on the rate of change. It is the rated torsional vibration amplitude under operating conditions, from .

8. The method for predicting the lifespan of an automotive shock absorber according to claim 7, characterized in that, By comparing the torsional vibration rate with standard parameters of the same engine model, the operating state of the shock absorber is obtained, including: The process for determining the working status of the shock absorber is as follows: Torsional vibration rate Standard parameters of the same model engine In , The operating status of the shock absorber was obtained by comparison: ； ； in, It is the real-time normalized rate of change, which quantifies the degree of deviation between the current state and the benchmark. It is the baseline rate of change, which is the bench test calibration value; It is the maximum allowable deviation, from Its function is to determine whether the shock absorber is working properly; The rules for classifying abnormal states are as follows: ； Minor anomalies are recorded by the system without alarm; moderate anomalies trigger a yellow warning on the dashboard; and severe anomalies trigger a red alarm and torque limiting control. This formula quantifies the severity of the anomaly and triggers a graded response mechanism.

9. The method for predicting the lifespan of an automotive shock absorber according to claim 8, characterized in that, Based on the cumulative damage theory, a vibration damper life prediction model is constructed, including: The process of constructing the vibration damper life prediction model is as follows: Input the pre-processed engine parameter set B, damper damping status, damper operating status, and damper service life. Standard parameter set ; The initial damage factor construction process is as follows: Input vibration damper service life and Standard service life of vibration dampers ; ； in, It is the initial damage factor, and its function is to convert the percentage of running time into the initial damage value; The shock absorber has reached the end of its service life. This is the standard service life of the shock absorber, taken from... ; SN curve temperature correction: First, calculate the corrected fatigue constant, and then perform temperature correction on the SN curve; The calculation process for the temperature correction factor is as follows: ； in, It is a temperature correction factor, which is used to quantify the effect of temperature on the fatigue life of materials. It is the material activation energy, that is, the activation energy of the rubber layer material of the core elastic element in the torsional vibration damper. It is a predetermined parameter for the rubber material of the damper, which is determined in advance through dynamic thermomechanical analysis test. It is the gas constant; It is the engine oil temperature, from dataset B; The physical meaning is to convert the measured Celsius temperature of the engine oil into the thermodynamic temperature Kelvin. The calculation process for the corrected fatigue constant is as follows: ； in, It is the corrected fatigue constant, and its function is to generate the temperature-corrected material fatigue constant; It is the original fatigue constant, derived from ; It is a temperature correction factor; The process of correcting the SN curve is as follows: ； ； in, It is the stress amplitude, and its function is to convert torsional vibration into material stress; It is the torsional vibration-stress conversion coefficient, and the specific value was determined through previous bench tests. It is the real-time torsional vibration amplitude; It is the theoretical fatigue life, and its function is to calculate the theoretical fatigue life under the current stress. It is a material index, which refers to the core elastic element material, namely rubber, used to absorb vibration energy in torsional vibration dampers. The specific value is determined through material fatigue tests. Stress amplitude power of m; This is the corrected fatigue constant; The real-time damage calculation process is as follows: ； ； in, It represents the number of cycles, and its function is to calculate the number of load cycles within the sampling period. It is the engine's real-time speed; is the sampling interval; i is the excitation order, i=3 or 6; It is the damage increment, and its function is to quantify the damage increment in the current sampling period; It is the theoretical fatigue life; Life cycle injury The synthesis calculation process is as follows: ； ； ； in, It is a full life cycle damage, and its function is to synthesize historical damage and real-time damage; It is the initial damage; It is the increase in damage; It is a damage correction coefficient, which is used to correct the accumulated damage based on the damper damping state output by S3 and the damper working state output by S4.

10. The method for predicting the lifespan of an automotive shock absorber according to claim 9, characterized in that, The damper damping state, damper operating state, and engine parameter set are input into the damper life prediction model, and the remaining life of the damper is output, including: Remaining life of the shock absorber The calculation process is as follows: ； in, It is the remaining life of the vibration damper, and its function is to calculate the interpretable remaining life of the engineering. This is the standard service life; It is damage throughout the entire life cycle.

Images