CN-122021313-A - Intelligent track smoothness adjusting method and system for dynamic detection and big data analysis

Abstract

The invention provides an intelligent regulation method and system for track smoothness of dynamic detection and big data analysis, and belongs to the technical field of railway engineering maintenance. The method comprises the steps of collecting track geometric data, vehicle dynamics response data and environment data in real time through a multi-dimensional sensor array arranged on an operation train, uploading the data to a cloud big data platform for fusion and deep mining, constructing a track state degradation prediction model by using a machine learning algorithm, accurately predicting the development trend of track irregularity, automatically generating a targeted track fine adjustment scheme based on a prediction result, and issuing the track fine adjustment scheme to an intelligent maintenance robot deployed on a line or guiding a manual maintenance team. The system comprises a data acquisition layer, a cloud analysis layer and an intelligent execution layer. The invention realizes the transition from the periodic maintenance to the predictive maintenance of the track state, obviously improves the maintenance efficiency, the precision and the smoothness of the line, and reduces the maintenance cost of the whole life cycle.

Inventors

• ZHU ZHEN
• DU BAOFENG
• LI HAIYING
• CHEN MIN
• ZHU YI
• LIU YOUJIN
• SUN DAWEI
• LI BIN
• LIU DONG
• SANG DI
• XU TAOPING

Assignees

• 中国铁路上海局集团有限公司淮安高铁基础设施段

Dates

Publication Date

20260512

Application Date

20260130

Claims (8)

1. 1. The intelligent track smoothness adjusting method for dynamic detection and big data analysis is characterized by comprising the following steps of: S1, dynamically acquiring track state data in real time through a detection system deployed on an operation train; S2, uploading the track state data to a central processing platform, and constructing a track smoothness state evaluation and trend prediction model by utilizing a big data analysis and machine learning algorithm; s3, automatically generating a personalized track adjustment scheme based on an output result of the prediction model; And S4, issuing an adjustment scheme instruction to automatic maintenance equipment on the line or providing the adjustment scheme instruction for maintenance personnel to execute.
2. 2. The intelligent regulation method for track smoothness of dynamic detection and big data analysis according to claim 1, wherein in step S1, the detection system comprises a multi-dimensional sensor array comprising an inertial measurement unit, a laser displacement sensor, an optical line camera, an acoustic sensor and a global navigation satellite system receiver.
3. 3. The intelligent regulation method for track smoothness of dynamic detection and big data analysis according to claim 1, wherein in step S2, the track smoothness state evaluation and trend prediction model is constructed by using big data analysis and machine learning algorithm, and specifically comprises the following steps: S21, carrying out data cleaning, alignment and fusion on multi-source heterogeneous track state data to form a track state data cube with consistent space-time; S22, extracting time domain, frequency domain and time domain characteristics related to track smoothness degradation from a data cube; S23, training a prediction model based on a gradient lifting decision tree or a long-short-term memory network by utilizing historical data, and predicting the evolution value of the geometric parameters of the track in a specific time period in the future by taking the characteristics as input.
4. 4. The intelligent track smoothness adjustment method based on dynamic detection and big data analysis of claim 1, wherein in step S3, a personalized track adjustment scheme is automatically generated, specifically comprising calculating an optimal adjustment amount through an optimization algorithm based on predicted track geometry overrun positions and degrees and combining line design parameters, track structure types and material characteristics, wherein the adjustment scheme comprises a torque adjustment instruction aiming at a ballastless track fastener system or a track lifting amount and tamping strength instruction aiming at a ballastless track.
5. 5. The intelligent track smoothness adjusting method based on dynamic detection and big data analysis according to claim 1, wherein in step S4, the automatic maintenance device is an intelligent maintenance robot moving independently, the intelligent maintenance robot is integrated with a high-precision positioning module, a machine vision recognition module and various executing mechanisms, automatically positions to a target adjusting point according to an adjusting scheme instruction, and completes torque adjustment of a fastener bolt or accurate adjustment of a sleeper position.
6. 6. An intelligent track smoothness adjustment system for dynamic detection and big data analysis, for implementing the method according to any one of claims 1 to 5, comprising: The data acquisition layer is composed of detection units deployed at the bottoms of the multiple trains of operation trains and is used for acquiring track state data in real time; The cloud analysis layer comprises a big data storage module, a machine learning algorithm module and a decision generation module and is used for processing data, training a prediction model and generating an adjustment instruction; And the intelligent execution layer is formed by one or more intelligent maintenance robots deployed on the line and used for receiving and executing the adjustment instruction from the cloud analysis layer.
7. 7. The intelligent track smoothness adjusting system for dynamic detection and big data analysis according to claim 6, wherein data transmission is performed between the data acquisition layer and the cloud analysis layer through a 5G or satellite communication network, and instruction issuing and state feedback are performed between the cloud analysis layer and the intelligent execution layer through a wireless local area network or a mobile communication network.
8. 8. The intelligent regulation system for track smoothness in dynamic detection and big data analysis according to claim 6, wherein the intelligent maintenance robot further comprises a state feedback module for retesting the regulated track state after executing the regulation task and feeding retest data back to the cloud analysis layer.

Description

Intelligent track smoothness adjusting method and system for dynamic detection and big data analysis Technical Field The invention relates to the technical field of railway engineering maintenance, in particular to an intelligent regulation method and system for track smoothness of dynamic detection and big data analysis. Background Track smoothness is a core index for ensuring safe, stable and high-speed running of trains. Along with the rapid development of high-speed railways, the requirements on track smoothness are increasingly severe. Under the action of factors such as train cyclic load, environmental change and the like, the track can generate uneven settlement and geometric shape and position change, so that smoothness is reduced. The traditional track maintenance mode mainly relies on a large detection vehicle to detect a line periodically (e.g. once every 10-20 days), and after finding out an overrun place, a manual or large road maintenance machine (e.g. a tamping vehicle) is arranged for maintenance. The mode has obvious defects of 1) poor timeliness, namely that time is delayed from disease discovery to maintenance implementation, smoothness is continuously deteriorated during the period to influence riding comfort and safety margin, 2) high cost, high operation cost of large-scale detection vehicles and road maintenance machines, high influence on transportation efficiency due to the fact that 'skylight points' operation is required to be occupied, and 3) poor precision and pertinence, namely that most of traditional maintenance is interval type adjustment, and the lack of accurate adjustment strategy for each fastener or sleeper is easy to cause 'over maintenance' or 'under maintenance'. In recent years, research attempts are made to utilize vehicle-mounted monitoring equipment, but the research is limited to state monitoring and alarming, and the vehicle-mounted monitoring equipment cannot be deeply integrated with a predictive maintenance model and automatic execution equipment, so that a complete intelligent closed-loop solution is formed. Disclosure of Invention Aiming at the technical defects, the invention aims to provide an intelligent track smoothness adjusting method and system for dynamic detection and big data analysis, which realize predictive maintenance and fine adjustment of track smoothness. In order to achieve the above purpose, the invention adopts the following technical scheme: An intelligent track smoothness adjusting method for dynamic detection and big data analysis comprises the following steps: S1, dynamically acquiring track state data in real time through a detection system deployed on an operation train; S2, uploading the track state data to a central processing platform, and constructing a track smoothness state evaluation and trend prediction model by utilizing a big data analysis and machine learning algorithm; s3, automatically generating a personalized track adjustment scheme based on an output result of the prediction model; And S4, issuing an adjustment scheme instruction to automatic maintenance equipment on the line or providing the adjustment scheme instruction for maintenance personnel to execute. Preferably, in step S1, the detection system comprises a multi-dimensional sensor array, wherein the multi-dimensional sensor array comprises an inertial measurement unit, a laser displacement sensor, an optical line camera, an acoustic sensor and a global navigation satellite system receiver. Preferably, in step S2, the construction of the track smoothness state estimation and trend prediction model by using the big data analysis and machine learning algorithm specifically includes the following steps: S21, carrying out data cleaning, alignment and fusion on multi-source heterogeneous track state data to form a track state data cube with consistent space-time; S22, extracting time domain, frequency domain and time domain characteristics related to track smoothness degradation from a data cube; S23, training a prediction model based on a gradient lifting decision tree or a long-short-term memory network by utilizing historical data, and predicting the evolution value of the geometric parameters of the track in a specific time period in the future by taking the characteristics as input. Preferably, in the step S3, an individualized track adjustment scheme is automatically generated, specifically comprising calculating an optimal adjustment amount through an optimization algorithm based on the predicted geometrical overrun position and degree of the track and combining the line design parameters, the track structure type and the material characteristics, wherein the adjustment scheme comprises a torque adjustment instruction aiming at a ballastless track fastener system or a track lifting amount and tamping strength instruction aiming at a ballastless track. Preferably, in step S4, the automatic maintenance equipment is an intelligent maintenance robot moving au