CN-119066955-B - Wet clutch residual life prediction method and system

CN119066955BCN 119066955 BCN119066955 BCN 119066955BCN-119066955-B

Abstract

The invention discloses a method and a system for predicting the residual life of a wet clutch, wherein the method comprises the steps of obtaining clutch data to be predicted, inputting the clutch data to be predicted into an improved IG process model, obtaining a predicted value of the residual life of the clutch, and training the improved IG process model through a training set, wherein the training set is friction coefficient degradation data. The invention is beneficial to prolonging the reliable service life of the clutch, reducing the maintenance cost and providing effective support for the health management and predictive maintenance of the vehicle.

Inventors

YU LIANG
FENG YUQING
ZHENG CHANGSONG
XIONG CENBO
LI HEYAN
OUYANG XIANGJUN
JI DELIN

Assignees

北京理工大学

Dates

Publication Date: 20260508
Application Date: 20240806

Claims (5)

1. A wet clutch remaining life prediction method, comprising: Acquiring clutch data to be predicted; inputting the clutch data to be predicted into an improved IG process model, and obtaining a predicted value of the residual life of the clutch, wherein the improved IG process model is formed by training a training set, and the training set is friction coefficient degradation data; introducing random drift parameters and random vibration parameters to modify a parameter structure of the IG process model; the improved IG process model further includes: The method comprises the steps of defining prior distribution and posterior distribution for parameters by using a Bayesian method, performing posterior sampling on an IG process model modified by structural parameters by using an MCMC algorithm, and obtaining parameter statistics to be estimated of the improved IG process model; The probability density function of the random drift parameter μ is: The probability density function of the random vibration parameter lambda is as follows: wherein ω, κ, η and γ are the super parameters to be estimated by the model, B is the normalization factor, Is a gamma function, mu is a random drift parameter, lambda is a random vibration parameter, phi is a probability density symbol of standard normal distribution; the probability density function of the IG process model with the modified structural parameters is as follows: wherein Λ (t) is a time dependent shape function, and Y t = Y (t) is a degradation observation at time t.
2. The method of claim 1, wherein obtaining the coefficient of friction degradation data comprises: Acquiring initial friction coefficient degradation data of the clutch by using a sensor; and carrying out outlier rejection processing on the initial friction coefficient degradation data to obtain the friction coefficient degradation data.
3. The method for predicting remaining life of a wet clutch as claimed in claim 1, wherein the inputting of the clutch data to be predicted to the improved IG process model further comprises: Acquiring latest degradation observation data; And setting degradation data of a preset time window length based on the latest degradation observation data.
4. The method of claim 1, wherein obtaining the predicted remaining life of the clutch comprises: Predicting a degradation state based on the improved IG process model; Judging whether the degradation state exceeds a preset fault threshold value, and acquiring the predicted value of the residual life of the clutch according to a judging result.
5. The wet clutch residual life prediction system is characterized by comprising a bench test module, a data processing module, a numerical modeling module and a performance prediction module; The bench test module is used for designing a cyclic engagement test of the clutch system under the actual vehicle condition, and collecting all friction coefficient degradation data of the clutch system from the time of being put into use to the time of ending the test by using a sensor; the data processing module is used for carrying out outlier rejection processing on the friction coefficient degradation data; The numerical modeling module is used for constructing an improved IG process model of a modified parameter structure containing random drift and random vibration; The performance prediction module is used for acquiring a predicted value of the residual life of the clutch based on an improved IG process model; the improved IG process model further includes: The method comprises the steps of defining prior distribution and posterior distribution for parameters by using a Bayesian method, performing posterior sampling on an IG process model modified by structural parameters by using an MCMC algorithm, and obtaining parameter statistics to be estimated of the improved IG process model; The probability density function of the random drift parameter μ is: the probability density function of the random vibration parameter λ is: wherein ω, κ, η and γ are the super parameters to be estimated by the model, B is the normalization factor, Is a gamma function, mu is a random drift parameter, lambda is a random vibration parameter, phi is a probability density symbol of standard normal distribution; the probability density function of the IG process model with the modified structural parameters is as follows: wherein Λ (t) is a time dependent shape function, and Y t = Y (t) is a degradation observation at time t.

Description

Wet clutch residual life prediction method and system Technical Field The invention belongs to the field of disciplinary intersection combining equipment reliability engineering with probability and statistical analysis, and particularly relates to a wet clutch residual life prediction method and system. Background The wet multi-plate clutch is an important power transmission component of a vehicle comprehensive transmission device and is widely applied to various special vehicles. The wet type multi-plate clutch mainly relies on friction pairs formed by friction plates and separating plates to realize power transmission and interruption, so that high power transmission and control are easy to realize. However, the cumulative damage caused by the frictional wear effect gradually thins and smoothes the frictional surface, losing the intended design function over the life. Thus, effective monitoring of wet clutch degradation data, i.e., performance characteristics (Performance Characteristic, PC), may provide a reliable historical information basis for effective predictive maintenance measures and predicting remaining useful life (REMAINING USEFUL LIFE, RUL). By means of the accurate prediction method, the service life and reliability of the wet clutch can be prolonged, and the working cost in the using stage is reduced. Conventional reliability analysis based on lifetime data incurs higher costs in terms of acquisition and implementation of the available data. The PC reflects the evolution of clutch performance more accurately than life data, and also contains more life information. Degradation analysis, including degradation modeling and parameter estimation, has proven to be an important tool kit, especially for devices with limited test time and sample size. Furthermore, we can use failure mechanisms to guide modeling of the product degradation process, conduct reliability analysis and predict RUL, which has instructive implications for state-based maintenance. An appropriate degradation model is critical to clutch degradation characteristics and reliability representation. Physical-based methods typically require a clear mathematical model to quantitatively characterize the degradation behavior of the system, which is almost impossible for complex clutch systems. Degradation modeling based on random processes is a more promising option because of the ability to capture random dynamics and cell heterogeneity (typically including random drift and random vibration) in the degradation process. In particular, the Inverse Gaussian process (IG) has been particularly prominent in recent studies and has proven to be an attractive model of the degradation process. Although many students have achieved a certain research effort in exploring the decline of the operating characteristics and the prediction of the remaining life of the clutch by using methods of finite element simulation and physical models, etc. However, there is still a large gap in the degradation analysis and the RUL prediction of complex wet clutch systems based on stochastic processes, in which many studies do not fully consider the unit heterogeneity between the same and different clutch devices, including long-term trends and short-term effects, and in which, in addition, the existing studies mostly obtain higher model performance by model initialization with a large amount of offline data, neglecting the importance of on-line real-time diagnosis of the devices, which is an insufficient for expensive clutch systems. Therefore, there is a need for a life prediction model that effectively guides the degradation prediction and real-time prediction of RUL of a clutch system based on degradation data, guides the performance of an on-demand maintenance strategy and health management, and ensures efficient and reliable operation of the clutch during the service phase. Disclosure of Invention In order to solve the technical problems, the invention provides a wet clutch residual life prediction method and a wet clutch residual life prediction system, which can solve the limitations of the prior art in terms of degradation analysis and residual life prediction of a clutch. In order to achieve the above object, the present invention provides a wet clutch remaining life prediction method, comprising: Acquiring clutch data to be predicted; inputting the clutch data to be predicted into an improved IG process model, and obtaining a predicted value of the residual life of the clutch, wherein the improved IG process model is formed by training a training set, and the training set is friction coefficient degradation data; the improved IG process model is characterized in that random drift parameters and random vibration parameters are introduced, and the parameter structure of the IG process model is modified. Optionally, acquiring the coefficient of friction degradation data includes: Acquiring initial friction coefficient degradation data of the clutch by usin