CN-121994406-A - Fault early warning method and system for dynamic torque test platform

CN121994406ACN 121994406 ACN121994406 ACN 121994406ACN-121994406-A

Abstract

The invention discloses a fault early warning method and a system of a dynamic torque testing platform, and relates to the technical field of intelligent fault diagnosis; the method comprises the steps of obtaining a reference health parameter based on a model, extracting dynamic change characteristics from collected information through time sequence decomposition, combining the reference health parameter with a current working condition, obtaining abnormal change characteristics through self-adaptive threshold mechanism comparison, finally tracing the abnormal change characteristics to construct a fault propagation diagram, locating a potential fault propagation path, and generating early warning information comprising fault location and propagation paths. The method solves the technical problems of poor working condition adaptability, inaccurate abnormal identification and uncontrollable fault propagation in the traditional fault early warning, and achieves the technical effects of accurately identifying potential faults, locating fault sources and propagation paths and improving the accuracy and practicality of the fault early warning.

Inventors

TIAN LEI
LI LINGBO
JIANG RENHUA
LI YUTE
Zhang Manqiao
YANG WEN
JIANG TONGMING
YANG CHANGQUN
HUAN JUNJUN
NIU DAODONG
LIN YUANWEN
LI MIAO
QIU SHUI

Assignees

国家石油天然气管网集团有限公司
国家石油天然气管网集团有限公司华南分公司
扬州恒春电子有限公司

Dates

Publication Date: 20260508
Application Date: 20260126

Claims (10)

1. The fault early warning method of the dynamic torque test platform is characterized by comprising the following steps of: Constructing a digital twin model according to the physical structure and the dynamic behavior of the operation equipment, connecting the digital twin model with a dynamic torque test platform, and obtaining dynamic torque information and vibration information of the operation equipment in real time; acquiring a reference health parameter in a normal state based on the constructed digital twin model, and extracting dynamic change characteristics by utilizing the acquired vibration information and dynamic torque information and through time sequence decomposition; According to the extracted dynamic change characteristics, combining the reference health parameters and the current working conditions, and adopting a self-adaptive threshold mechanism to perform working condition state comparison to obtain abnormal change characteristics; and performing feature tracing based on the abnormal change features, constructing a fault propagation diagram, positioning propagation paths of potential faults, and generating early warning information comprising fault positioning and propagation paths.
2. The fault early warning method of a dynamic torque testing platform according to claim 1, wherein the method is characterized in that a digital twin model is constructed according to a physical structure and dynamic behavior of operation equipment, the digital twin model is connected with the dynamic torque testing platform, and dynamic torque information and vibration information of the operation equipment are obtained in real time, and the method comprises the following steps: According to the physical structure, dynamic behavior and mechanical property of the operation equipment, a digital twin simulation model is constructed, the digital twin simulation model can reflect the dynamic response of the equipment under different working conditions, and comprises the geometric structure of the equipment, the physical property of each component and a dynamic behavior model, wherein in the digital twin simulation model, the interaction relationship and the physical rule among the components of the equipment are built by adopting a finite element analysis and fluid dynamics simulation method, and the dynamic behaviors such as load, vibration, torque and the like of the equipment are simulated; Connecting a dynamic torque testing platform of the equipment with a digital twin model through an interface, and acquiring dynamic torque information and vibration information data of the equipment in real time, wherein the dynamic torque testing platform comprises a plurality of sensors and is used for acquiring dynamic torque and vibration signals of the equipment; the sensor is used for collecting torque and vibration dynamic signals of the equipment in the running process, the collected signals are input into the digital twin model, and real-time updating and simulation of the equipment state are carried out.
3. The fault early warning method of the dynamic torque testing platform according to claim 2, wherein obtaining the reference health parameter in the normal state based on the constructed digital twin model comprises: generating a reference health parameter set associated with the working condition parameters when the operation equipment is in a normal state based on the digital twin model; Aligning vibration signals and dynamic torque signals under various working conditions according to the reference health parameter set; Performing time sequence decomposition processing by using the aligned vibration signal and dynamic torque signal to obtain at least two of a trend component, a periodic component and an impact component; And extracting characteristic quantities representing dynamic response from the components respectively, analyzing correlation coupling characteristics between the vibration signals and the dynamic torque signals, and obtaining the reference health parameters, wherein the reference health parameters are used for describing parameter sets of device torque and vibration signal characteristic change rules of running devices under different working conditions in a health state.
4. The fault pre-warning method of a dynamic torque testing platform according to claim 3, wherein extracting dynamic change features through time sequence decomposition by using the collected vibration information and dynamic torque information comprises: The vibration information and the dynamic torque information are subjected to characteristic decomposition according to a time sequence relation through the vibration information and the dynamic torque information obtained by the dynamic torque testing platform in real time, so that a change time sequence chain of the vibration information and the dynamic torque information is obtained; and extracting the change trend, the periodic change and the impact change according to the change time sequence chain, and obtaining the dynamic change characteristics by utilizing the extracted characteristics.
5. The fault early warning method for a dynamic torque testing platform according to claim 4, wherein before the comparing of the working conditions by adopting the adaptive threshold mechanism, the fault early warning method further comprises: based on the reference health parameters, constructing a mapping relation between working conditions and characteristic changes; Carrying out corresponding decomposition of dynamic change characteristics according to the mapping relation to obtain a health parameter change quantity relation under each working condition, and analyzing the working condition change relation of the vibration data and the torque coupling relation to obtain a vibration-torque coupling change quantity relation; establishing a threshold self-adaptive matching model according to the health parameter change quantity relation and the vibration-torque coupling change quantity relation under each working condition; and configuring the self-adaptive threshold mechanism based on the threshold self-adaptive matching model, wherein the self-adaptive threshold mechanism is used for determining characteristic judgment boundaries under different working conditions and health states.
6. The fault pre-warning method of the dynamic torque testing platform according to claim 5, wherein the step of comparing the working condition states by adopting an adaptive threshold mechanism according to the extracted dynamic change characteristics in combination with the reference health parameter and the current working condition to obtain abnormal change characteristics comprises the steps of: utilizing the current working condition to match with the reference health parameter, and identifying a reference health judgment boundary under the current working condition; Analyzing vibration characteristics, torque characteristics and coupling relations thereof according to the dynamic change characteristics; Performing threshold matching with the self-adaptive threshold mechanism by utilizing the reference health judgment boundary, the vibration characteristic, the torque characteristic and the coupling relation of the reference health judgment boundary, the vibration characteristic and the torque characteristic under the current working condition to obtain a self-adaptive threshold; and comparing the dynamic change characteristics through the self-adaptive threshold value, and identifying the dynamic change characteristics which do not meet the threshold value as the abnormal change characteristics.
7. The fault pre-warning method of the dynamic torque testing platform according to claim 6, wherein analyzing the vibration characteristic, the torque characteristic and the coupling relation thereof according to the dynamic change characteristic comprises: according to the dynamic change characteristics, classifying and analyzing the vibration characteristics and the torque characteristics, and respectively identifying the change type, the deviation degree and the corresponding structure association characteristics; Based on the identified change type, deviation degree and corresponding structure association characteristics, carrying out time alignment and change consistency analysis on the vibration characteristics and the torque characteristics, judging the sequence and synchronous change relation of the vibration characteristics and the torque characteristics under the same working condition, and obtaining change association information; based on the change association information, the change association degree between the vibration characteristics and the torque characteristics is analyzed, and the coupling relation and the coupling strength of the vibration characteristics and the torque characteristics are determined and used for representing the association characteristic of abnormal change.
8. The fault early warning method of the dynamic torque testing platform according to claim 1, wherein the feature tracing is performed based on the abnormal change features to construct a fault propagation diagram, comprising: determining causal relation between abnormal change characteristics according to time sequence, characteristic type and coupling relation between vibration characteristics and torque characteristics of the abnormal change characteristics; Based on the causal relationship, a fault propagation diagram for describing the transmission relationship of faults in equipment is constructed by combining the structural relationship, the functional relationship and the component association relationship established in the digital twin model of the operation equipment, wherein nodes of the fault propagation diagram are used for representing equipment components or functional units, and edges of the fault propagation diagram are used for representing the propagation relationship of the faults among the nodes.
9. The fault early-warning method of the dynamic torque testing platform according to claim 8, wherein the performing propagation path localization on the potential fault to generate early-warning information including the fault localization and the propagation path includes: Tracing analysis is carried out on the abnormal change characteristics based on the characteristic types, the occurrence time sequence and the change amplitude of the abnormal change characteristics, and a candidate fault source set is generated; dynamically correcting the weight of each propagation relation in the fault propagation diagram based on the current operation condition parameters and the coupling relation between the vibration characteristics and the torque characteristics; Comprehensively evaluating the occurrence sequence, feature consistency and spatial propagation gradient of the abnormal features of the candidate fault sources, determining the propagation starting point of potential faults, solving a fault propagation path with the highest matching degree with the abnormal change feature set in the fault propagation diagram, and outputting the corresponding fault source position, propagation path and confidence coefficient thereof; and generating early warning information according to the fault source position, the propagation path and the confidence level thereof, wherein the risk level or the propagation path contribution degree of the early warning information is marked based on the contribution degree of each node in the propagation path to abnormal change.
10. A fault early warning system for a dynamic torque testing platform, characterized in that the fault early warning system is used for implementing the fault early warning method for the dynamic torque testing platform according to any one of claims 1-9, and the system comprises: The running equipment information acquisition module is used for constructing a digital twin model according to the physical structure and the dynamic behavior of the running equipment, connecting the digital twin model with the dynamic torque test platform and acquiring the dynamic torque information and the vibration information of the running equipment in real time; The dynamic change feature acquisition module is used for acquiring a reference health parameter in a normal state based on the constructed digital twin model, and extracting dynamic change features through time sequence decomposition by utilizing the acquired vibration information and dynamic torque information; The abnormal change feature acquisition module is used for carrying out condition state comparison by adopting a self-adaptive threshold mechanism according to the extracted dynamic change features and combining the reference health parameters and the current working conditions to obtain abnormal change features; And the early warning information acquisition module is used for carrying out feature tracing based on the abnormal change features, constructing a fault propagation diagram, positioning a propagation path of the potential fault and generating early warning information comprising the fault positioning and the propagation path.

Description

Fault early warning method and system for dynamic torque test platform Technical Field The invention relates to the technical field of intelligent fault diagnosis, in particular to a fault early warning method and system of a dynamic torque test platform. Background The dynamic torque test platform is widely applied in industrial production, and the stable operation of the dynamic torque test platform directly relates to production efficiency and operation safety, and fault prediction and health management are core means for guaranteeing reliable operation of the dynamic torque test platform. In the prior art, the traditional fault early warning is based on fixed threshold judgment and single parameter monitoring, and plays a certain role under a stable working condition. However, as the complexity of industrial working conditions is improved, the traditional technology has obvious limitation that the traditional means cannot accurately capture the coupling characteristics of torque and vibration due to the multi-component linkage characteristics and the dynamic change of working conditions of the dynamic torque test platform, and is difficult to adapt to different working conditions, so that the early warning data is high in one-sided and false alarm rate, and the accurate and comprehensive requirements of equipment fault prediction and health management cannot be met. Disclosure of Invention The application provides a fault early warning method and a fault early warning system for a dynamic torque test platform, which solve the technical problems of poor working condition adaptability, inaccurate abnormal identification and uncontrollable fault propagation in the traditional fault early warning. The application provides a fault early warning method of a dynamic torque test platform, which comprises the steps of constructing a digital twin model according to a physical structure and dynamic behaviors of operation equipment, connecting the digital twin model with the dynamic torque test platform, obtaining dynamic torque information and vibration information of the operation equipment in real time, obtaining standard health parameters in a normal state based on the constructed digital twin model, utilizing the collected vibration information and the collected dynamic torque information to extract dynamic change characteristics through time sequence decomposition, comparing working condition states by adopting a self-adaptive threshold mechanism according to the extracted dynamic change characteristics and combining the standard health parameters and current working conditions to obtain abnormal change characteristics, tracing the characteristics based on the abnormal change characteristics, constructing a fault propagation diagram, positioning a propagation path of potential faults, and generating early warning information comprising fault positioning and propagation paths. The system comprises an operation equipment information acquisition module, a dynamic change feature acquisition module, an abnormal change feature acquisition module and an early warning information acquisition module, wherein the operation equipment information acquisition module is used for constructing a digital twin model according to the physical structure and dynamic behavior of operation equipment, connecting the dynamic torque test platform, acquiring dynamic torque information and vibration information of the operation equipment in real time, acquiring a standard health parameter in a normal state based on the constructed digital twin model, extracting dynamic change features through time sequence decomposition by utilizing the acquired vibration information and the dynamic torque information, the abnormal change feature acquisition module is used for comparing working condition states by adopting an adaptive threshold mechanism according to the extracted dynamic change features and combining the standard health parameter and the current working condition, and the early warning information acquisition module is used for carrying out feature tracing based on the abnormal change features, constructing a fault propagation diagram, positioning a potential fault propagation path and generating early warning information comprising fault positioning and a fault propagation path. One or more technical schemes provided by the application have at least the following technical effects or advantages: According to the application, by constructing a virtual model of the physical structure and dynamic behavior of the laminating equipment, connecting a test platform to collect relevant dynamic data, extracting change characteristics through time sequence decomposition, establishing dynamic judgment standards adapting to different working conditions, comparing reference parameters with the current working conditions, and constructing a fault transfer relation diagram through characteristic tracing, the fault early warning of the dynamic torque