CN-122021275-A - Train key component RAMS evaluation method based on life spectrum

CN122021275ACN 122021275 ACN122021275 ACN 122021275ACN-122021275-A

Abstract

The invention relates to the technical field of rail transit, in particular to a train key component RAMS evaluation method based on a life spectrum, which is characterized in that strain gauges, acceleration sensors and temperature sensors are distributed on a bogie bearing piece and a traction motor of a train, stress, vibration and temperature time sequence data are synchronously acquired at a sampling frequency not lower than 1kHz, and the unified time scale alignment of multi-source data is realized by combining speed, traction/braking force and transverse acceleration signals acquired by a train control management system; the data processing is divided into real-time acquisition of a train end and high-precision analysis of a ground end, and the service life spectrum is optimized through online learning and data reconstruction, so that the high-precision evaluation of the running state and the residual service life of key parts of the train is realized, and a scientific decision basis is provided for train running maintenance; the method can realize the health management of the whole life cycle of the train part and improve the operation safety, reliability and maintenance efficiency.

Inventors

LIU XUEMEI
LV JINLING
ZHENG RUIFANG
CUI SHIMING

Assignees

中车长春轨道客车股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260120

Claims (10)

1. The train key component RAMS evaluation method based on the life spectrum is characterized by comprising the following steps of: The method comprises the steps of firstly, arranging strain gauges, acceleration sensors and temperature sensors on a bogie bearing piece and a traction motor of a train, synchronously collecting stress, vibration and temperature time sequence data at a sampling frequency not lower than 1kHz, and carrying out unified time scale alignment with a train speed and traction/braking force signal obtained by a Train Control Management System (TCMS) in real time; The second step, carrying out working condition-environment self-adaptive preprocessing on the collected original data, wherein the preprocessing comprises the steps of decomposing and reducing noise of a vibration signal by adopting a wavelet packet to reserve a characteristic frequency band related to the natural frequency of a component, eliminating thermal stress interference of a strain signal by utilizing a temperature compensation algorithm, and eliminating and interpolating an abnormal point by adopting a median absolute deviation method to replace the abnormal point; Step three, working condition segmentation is carried out on the running process according to the speed change rate, the traction/braking force instruction and the transverse acceleration, typical working condition sections such as acceleration, cruising, braking, curves and the like are identified, a three-parameter rain flow counting method is adopted in each section to extract stress circulation, and meanwhile, the average temperature and the vibration intensity under the circulation are recorded to form a high-fidelity life spectrum taking the load amplitude, the average value, the circulation times, the accompanying temperature and the accompanying vibration intensity as dimensions; step four, a fatigue and abrasion coupled composite damage model is established based on a life spectrum, the total damage amount is calculated, and the residual life is predicted; Combining the residual life and a component failure rate model, dynamically generating reliability, availability, maintainability and safety (RAMS) indexes of a future operation interval, and triggering hierarchical early warning when the failure probability exceeds a set threshold; Step six, generating a refined maintenance work order with a maintenance time window and an operation list according to RAMS indexes and early warning grades and combining an operation diagram and spare part inventory, and issuing and executing; Step seven, completing data acquisition, preprocessing and life spectrum construction at the train end, uploading life spectrum abstract data to the ground end, and carrying out composite damage calculation, RAMS index generation and maintenance work order optimization at the ground end to realize real-time and high-precision division work cooperation of the train end; and step eight, when each working condition section is finished, data reconstruction and updating are carried out, service life spectrum data of each working condition section is dynamically optimized by using a feedback mechanism, and the optimization process is self-adaptively adjusted by an online learning algorithm, so that the evaluation accuracy and reliability of the system are improved.
2. The method for evaluating a train key component RAMS based on a life spectrum according to claim 1, wherein the temperature compensation algorithm is as follows: Wherein, the In order to measure the strain in the test, In order to be able to achieve a thermal expansion coefficient of the material, For the purpose of actually measuring the temperature, Is the reference temperature.
3. The method for evaluating the RAMS of the key train components based on the life spectrum according to claim 1, wherein the calculation of the fatigue life is based on an S-N-T three-dimensional database, the database is established by a material high-temperature fatigue test, in the life spectrum calculation, for each stage of load cycle, the corrected fatigue life value is retrieved from the S-N-T database according to the corresponding average temperature, and the fatigue life is dynamically adjusted according to factors such as the speed of a vehicle, the temperature, the stress cycle and the like, and the method comprises the following steps: Step one, determining the load amplitude and the load times of each working condition section; step two, according to the temperature change, dynamically searching the fatigue life after corresponding temperature correction from an S-N-T database; and thirdly, considering load circulation, temperature fluctuation and stress amplitude when the fatigue life is calculated, and further optimizing a fatigue damage assessment model.
4. The life spectrum-based train key component RAMS evaluation method according to claim 1, wherein the total damage amount of the composite damage model is as follows: Wherein the first item is fatigue damage, For the ratio of cycle times to fatigue life after temperature correction, the second term is wear damage, Is the first The intensity of the vibration in the time period, In order to keep the time of the action, And Is the material wear coefficient.
5. The life spectrum-based train key component RAMS evaluation method of claim 1, wherein the failure rate model adopts double-parameter Weibull distribution, and the failure probability is: Wherein, the As a function of the shape parameter(s), And (3) for the service life of the features, the parameters are updated in a rolling way by combining the full life cycle data of the components and the operation data of the similar components through a maximum likelihood estimation method.
6. The method for evaluating the RAMS of the key train components based on the life spectrum of claim 1, wherein the early warning stage comprises a monitoring stage, an early warning stage and an emergency stage, the emergency stage is corresponding to the prior maintenance when the operation is planned to be stopped last time, the early warning stage is judged according to the state of the system, and on the premise of meeting the operation safety requirement, the early warning stage triggers maintenance early warning and automatically generates an emergency maintenance scheme.
7. The method for evaluating the train key component RAMS based on the life spectrum of claim 1, wherein the data communication between the train end and the ground end adopts the redundant combination of 4G/5G cellular communication and a satellite link so as to ensure the complete transmission of data in weak coverage areas such as mountainous areas, tunnels and the like, and the uploaded data packet comprises a life spectrum abstract, an RAMS intermediate result and a sensor state identifier.
8. The method for evaluating the RAMS of the key train components based on the life spectrum according to claim 1, wherein the ground maintenance platform is provided with a full life cycle database for storing life spectrum evolution history, maintenance records and RAMS indexes of each component, and providing a visual function of 'one component one file', so that operation and maintenance personnel can trace back life consumption track, past maintenance condition and health trend of any component from the operation to the present, and the ground platform can optimize a prediction model based on real-time data so as to dynamically adjust maintenance strategies.
9. The life spectrum-based train key component RAMS evaluation method according to claim 1 is characterized in that a data preprocessing and feature extraction unit at a train end is realized by adopting an FPGA or a high-performance embedded processor so as to ensure real-time parallel processing of multichannel high-frequency signals above 1kHz, and the FPGA comprises a rain flow counting hardware logic unit and a wavelet packet filtering unit.
10. The method for evaluating the RAMS of the key train component based on the life spectrum of the train according to claim 1, wherein the method comprises training historical data through data acquired in real time by utilizing a deep learning algorithm, automatically optimizing life spectrum data reconstruction of each working condition segment, and continuously adjusting maintenance decisions based on an optimization model so as to improve prediction accuracy and system performance of future data.

Description

Train key component RAMS evaluation method based on life spectrum Technical Field The invention relates to the technical field of rail transit, in particular to a train key component RAMS evaluation method based on a life spectrum. Background With the development of high-speed railways and urban rail transit, the reliability, safety and life management of key train components such as bogie carriers and traction motors are increasingly important. The existing RAMS (reliability, availability, maintainability and safety) assessment method mainly relies on statistical life models, empirical data or single sensor information to calculate, and has the problems of insufficient precision, untimely response to actual working conditions, unreasonable task division between a vehicle end and a ground end, lack of pertinence in maintenance decisions and the like. For example, the traditional method is difficult to consider the coupling effect of load circulation, temperature change and vibration influence on the damage of the components, and fine maintenance arrangement according to the real-time health state of the components cannot be realized, so that the maintenance cost is high, the utilization rate of spare parts is low, and even potential safety hazards exist. Disclosure of Invention The invention aims to provide a train key component RAMS evaluation method based on a life spectrum, so as to solve the problems in the background technology. In order to achieve the above purpose, the present invention provides the following technical solutions: a train key component RAMS evaluation method based on life spectrum comprises the following steps: The method comprises the steps of firstly, arranging strain gauges, acceleration sensors and temperature sensors on a bogie bearing piece and a traction motor of a train, synchronously collecting stress, vibration and temperature time sequence data at a sampling frequency not lower than 1kHz, and carrying out unified time scale alignment with a train speed and traction/braking force signal obtained by a Train Control Management System (TCMS) in real time; The second step, carrying out working condition-environment self-adaptive preprocessing on the collected original data, wherein the preprocessing comprises the steps of decomposing and reducing noise of a vibration signal by adopting a wavelet packet to reserve a characteristic frequency band related to the natural frequency of a component, eliminating thermal stress interference of a strain signal by utilizing a temperature compensation algorithm, and eliminating and interpolating an abnormal point by adopting a median absolute deviation method to replace the abnormal point; Step three, working condition segmentation is carried out on the running process according to the speed change rate, the traction/braking force instruction and the transverse acceleration, typical working condition sections such as acceleration, cruising, braking, curves and the like are identified, a three-parameter rain flow counting method is adopted in each section to extract stress circulation, and meanwhile, the average temperature and the vibration intensity under the circulation are recorded to form a high-fidelity life spectrum taking the load amplitude, the average value, the circulation times, the accompanying temperature and the accompanying vibration intensity as dimensions; step four, a fatigue and abrasion coupled composite damage model is established based on a life spectrum, the total damage amount is calculated, and the residual life is predicted; Combining the residual life and a component failure rate model, dynamically generating reliability, availability, maintainability and safety (RAMS) indexes of a future operation interval, and triggering hierarchical early warning when the failure probability exceeds a set threshold; Step six, generating a refined maintenance work order with a maintenance time window and an operation list according to RAMS indexes and early warning grades and combining an operation diagram and spare part inventory, and issuing and executing; Step seven, completing data acquisition, preprocessing and life spectrum construction at the train end, uploading life spectrum abstract data to the ground end, and carrying out composite damage calculation, RAMS index generation and maintenance work order optimization at the ground end to realize real-time and high-precision division work cooperation of the train end; and step eight, when each working condition section is finished, data reconstruction and updating are carried out, service life spectrum data of each working condition section is dynamically optimized by using a feedback mechanism, and the optimization process is self-adaptively adjusted by an online learning algorithm, so that the evaluation accuracy and reliability of the system are improved. Preferably, the temperature compensation algorithm is: Wherein, the In order to measure the strain in the test,In or