CN-121977865-A - Train control vehicle-mounted equipment health monitoring method, equipment, medium and product based on D-S evidence theory

CN121977865ACN 121977865 ACN121977865 ACN 121977865ACN-121977865-A

Abstract

The application discloses a train control vehicle-mounted equipment health monitoring method, equipment, medium and product based on a D-S evidence theory, and relates to the technical field of intelligent operation and maintenance of rail transit; the method comprises the steps of generating a second basic probability distribution function through a speed health estimator based on speed core data, generating a third basic probability distribution function through the communication health estimator based on communication auxiliary data, respectively carrying out discount processing by using an adaptive weight method, combining discount processing results twice to obtain a comprehensive basic probability distribution function, and selecting a state with the largest basic confidence distribution value as a health state. According to the application, by introducing (D-S) evidence theory, a health monitoring framework which can quantify uncertainty, effectively fuse multi-source decision information and has high interpretability is constructed, and the safety, reliability and operation and maintenance intelligent level of train operation are remarkably improved.

Inventors

LIU RUI
LIU DAI
CHENG RONG
ZHOU YANLI
WANG CHENG
Yang Luojun
CHEN ZEXIN
PENG QINGMING
YANG HUI

Assignees

华东交通大学

Dates

Publication Date: 20260505
Application Date: 20260403

Claims (10)

1. A health monitoring method of train control vehicle-mounted equipment based on a D-S evidence theory is characterized by comprising the following steps: the train control vehicle-mounted performance data comprises brake core data, speed core data and communication auxiliary data; generating, by a brake health evaluator, a first base probability distribution function based on the brake core data; Generating, by a speed health evaluator, a second base probability distribution function based on the speed core data; The brake health evaluator, the speed health evaluator and the communication health evaluator each generate a basic probability distribution function based on a predefined equipment health status recognition framework Θ, wherein the equipment health status recognition framework Θ= { health H, slight abnormality MA, serious abnormality SA, fault F, uncertainty }; respectively carrying out discount processing on the first basic probability distribution function, the second basic probability distribution function and the third basic probability distribution function by using a self-adaptive weight method to obtain discount processing results; Combining the discount processing results twice by using a Dempster to obtain a comprehensive basic probability distribution function; and selecting the state with the largest basic credibility allocation value in the comprehensive basic probability allocation function as the health state.
2. The method for monitoring health of train control vehicle-mounted equipment based on the D-S evidence theory according to claim 1, wherein the brake core data comprises brake cylinder pressure, wheel cylinder pressure, brake command level, emergency brake trigger signal, brake valve on-off state and brake response delay time; the speed core data comprises real-time speed of the train, issued target speed, acceleration of the train and speed deviation value; the communication auxiliary data comprises a train-ground interaction instruction message, wireless communication link signal strength, a check result of each downlink message, a check result of each uplink message and a rolling calculation past 10-second packet loss rate.
3. The method for health monitoring of a train-controlled vehicle-mounted device based on D-S evidence theory according to claim 1, further comprising, after acquiring train-controlled vehicle-mounted performance data: and marking a time stamp for the train control vehicle-mounted performance data, and performing time synchronization alignment processing on the brake core data, the speed core data and the communication auxiliary data with different sampling frequencies.
4. The method for monitoring the health of a train control vehicle device based on the D-S evidence theory according to claim 1, wherein the generating a first basic probability distribution function by a brake health estimator based on the brake core data specifically comprises: utilizing a brake health evaluator to sequentially match each brake rule in a preset brake health state judgment rule set according to the brake core data to obtain a brake matching result, wherein the brake matching result comprises confidence degrees of the brake core data corresponding to each brake rule respectively; And generating a first basic probability distribution function according to the braking matching result by using a maximum matching principle or a weighted average method.
5. The method for health monitoring of train control vehicle-mounted equipment based on the D-S evidence theory according to claim 1, wherein the formula for obtaining the discount processing result is that the discount processing is performed on the first basic probability distribution function, the second basic probability distribution function and the third basic probability distribution function by using an adaptive weight method, respectively: ; Wherein, the Representing the result after the ith deterministic raw evidence discount; The reliability weight of the ith data source is dynamically adjusted according to the historical performance; Representing a basic probability distribution function, i.e., deterministic raw evidence, generated by the ith data source through its dedicated evaluation rules; A represents a basic probability function comprising the first basic probability distribution function, the second basic probability distribution function and the third basic probability distribution function; Representing deterministic raw evidence generated for the ith data source Performing reliability discount to obtain evidence after discount, wherein the extra uncertainty introduced by insufficient reliability of the data source is explicitly distributed to the whole set Θ of the identification framework; Representing the i-th raw evidence representing uncertainty.
6. The method for monitoring health of a train control vehicle-mounted device based on the D-S evidence theory of claim 1, wherein the discount processing results comprise a first basic probability distribution function discount processing result, a second basic probability distribution function discount processing result and a third basic probability distribution function discount processing result, and the discount processing results are subjected to two-time Dempster combination to obtain a comprehensive basic probability distribution function, and the method specifically comprises the following steps: combining the discount processing result of the first basic probability distribution function and the discount processing result of the second basic probability distribution function to obtain an intermediate fusion result; Using the formula When the first conflict factor is larger than a first preset threshold value, the credibility value distributed to the identification framework theta in the intermediate fusion result is improved, and a conflict log is recorded; combining the intermediate fusion result and the discount processing result of the third basic probability distribution function to obtain a comprehensive basic probability distribution function; And when the second conflict factor is larger than a second preset threshold value, improving the credibility value distributed to the identification framework theta in the comprehensive basic probability distribution function, and recording a conflict log.
7. The method for monitoring the health of a train control vehicle-mounted device based on the D-S evidence theory according to claim 1, wherein the selecting the state with the largest assigned value of the basic belief in the comprehensive basic probability assignment function as the health state specifically includes: Based on the comprehensive basic probability distribution function, according to the formula Calculating the state with the maximum basic credibility allocation value to obtain the health state; And judging the confidence degree and uncertainty measure of the health state based on the comprehensive basic probability distribution function, wherein the uncertainty measure is a confidence value distributed to the identification framework theta.
8. A computer device comprising a memory, a processor and a computer program stored on the memory and capable of running on the processor, characterized in that the processor executes the computer program to implement the method for D-S evidence theory based on-board device health monitoring of any of claims 1-7.
9. A computer-readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the method for health monitoring of a train control on-board device based on D-S evidence theory as claimed in any one of claims 1 to 7.
10. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements the method for health monitoring of a train control on-board device based on D-S evidence theory as claimed in any one of claims 1 to 7.

Description

Train control vehicle-mounted equipment health monitoring method, equipment, medium and product based on D-S evidence theory Technical Field The application relates to the technical field of intelligent operation and maintenance of rail transit, in particular to a train control vehicle-mounted equipment health monitoring method, equipment, medium and product based on a D-S evidence theory. Background In the operation of train control vehicle-mounted equipment, multi-source heterogeneous data such as braking, speed, communication and the like are generated, and the existing health monitoring method mostly adopts threshold judgment or single model fusion, so that inherent noise, ambiguity and uncertainty of the data are difficult to effectively process. When the judgment of different data sources on the equipment state has conflict (such as normal braking but abnormal speed), the traditional method is easy to cause false alarm or missed judgment, and the evaluation result has poor robustness and low confidence. Although the D-S evidence theory can fuse uncertain information, in a train control vehicle-mounted scene, a technical scheme for constructing a structured evidence generation and self-adaptive fusion mechanism aiming at three main core data sources of braking, speed and communication does not exist, and the accuracy and the practicability of health monitoring are restricted. Therefore, a new method capable of systematically utilizing the Dempster-Shafer (D-S) evidence theory to quantitatively model uncertainty in multi-source vehicle-mounted data and realizing high-confidence and high-robustness health state assessment through strict evidence combination rules is needed so as to meet urgent demands of train safety, reliability and intelligent operation and maintenance. Disclosure of Invention The application aims to provide a train control vehicle-mounted equipment health monitoring method, equipment, medium and product based on a D-S evidence theory, which are used for constructing a health monitoring framework which can quantify uncertainty, effectively fuse multi-source decision information and has high interpretability by introducing the Dempster-Shafer (D-S) evidence theory, so that the safety, reliability and operation and maintenance intelligent level of train operation are obviously improved. In order to achieve the above object, the present application provides the following solutions: in a first aspect, the application provides a health monitoring method for train control vehicle-mounted equipment based on a D-S evidence theory, which comprises the following steps: the train control vehicle-mounted performance data comprises brake core data, speed core data and communication auxiliary data; generating, by a brake health evaluator, a first base probability distribution function based on the brake core data; Generating, by a speed health evaluator, a second base probability distribution function based on the speed core data; The brake health evaluator, the speed health evaluator and the communication health evaluator each generate a basic probability distribution function based on a predefined equipment health status recognition framework Θ, wherein the equipment health status recognition framework Θ= { health H, slight abnormality MA, serious abnormality SA, fault F, uncertainty }; respectively carrying out discount processing on the first basic probability distribution function, the second basic probability distribution function and the third basic probability distribution function by using a self-adaptive weight method to obtain discount processing results; Combining the discount processing results twice by using a Dempster to obtain a comprehensive basic probability distribution function; and selecting the state with the largest basic credibility allocation value in the comprehensive basic probability allocation function as the health state. Optionally, the brake core data comprises brake cylinder pressure, wheel cylinder pressure, brake command level, emergency brake trigger signal, brake valve on-off state and brake response delay time; the speed core data comprises real-time speed of the train, issued target speed, acceleration of the train and speed deviation value; the communication auxiliary data comprises a train-ground interaction instruction message, wireless communication link signal strength, a check result of each downlink message, a check result of each uplink message and a rolling calculation past 10-second packet loss rate. Optionally, after acquiring the train control on-board performance data, the method further comprises: and marking a time stamp for the train control vehicle-mounted performance data, and performing time synchronization alignment processing on the brake core data, the speed core data and the communication auxiliary data with different sampling frequencies. Optionally, generating, by the brake health evaluator, a first base probability distribution functio