CN-121997425-A - Bridge temperature effect compensation method based on digital twinning

CN121997425ACN 121997425 ACN121997425 ACN 121997425ACN-121997425-A

Abstract

The invention relates to the technical field of bridge construction, in particular to a digital twin-based bridge temperature effect compensation method which comprises the following steps of S1, obtaining temperature data and strain data of a bridge monitoring area, S2, generating a corresponding temperature pulse sequence and strain pulse sequence, S3, outputting an indication signal based on a time sequence analysis result, S4, carrying out temperature effect compensation on the strain data through a pre-built bridge digital twin model, converting a minute-level temperature and strain slow-change signal into a pulse sequence decoupled from physical time through asynchronous event-driven coding, fundamentally solving the contradiction of mismatch between sampling frequency and crack expansion time scale, intelligently distinguishing normal thermal strain and abnormal damage strain through time sequence analysis of a pulse neural network, and driving a digital twin model dynamic switching compensation strategy to accurately capture and separate microscopic fatigue crack transient signals under a strong thermal impact noise background.

Inventors

LI SHENHUI
SUN QIUYING
QU TAO
ZHANG HUAIJUN
LU JINBO

Assignees

青岛畅通市政工程设计有限公司

Dates

Publication Date: 20260508
Application Date: 20260126

Claims (10)

1. The bridge temperature effect compensation method based on digital twinning is characterized by comprising the following steps of: s1, acquiring temperature data and strain data of a bridge monitoring area; S2, based on the temperature data and the strain data, respectively generating a corresponding temperature pulse sequence and a corresponding strain pulse sequence through asynchronous event-driven differential modulation; s3, inputting the temperature pulse sequence and the strain pulse sequence into a pre-trained pulse neural network model for time sequence analysis, and outputting an indication signal based on a time sequence analysis result; And S4, carrying out temperature effect compensation on the strain data through a pre-constructed bridge digital twin model based on the indication signal output by the pulse neural network model, wherein the bridge digital twin model comprises a thermodynamic coupling constitutive relation of the bridge monitoring area.
2. The method for compensating the bridge temperature effect based on digital twin according to claim 1, wherein in S2, a corresponding temperature pulse sequence and a corresponding strain pulse sequence are respectively generated through asynchronous event-driven differential modulation based on the temperature data and the strain data, and the method is characterized in that: Asynchronous event driven differential modulation is respectively carried out on the temperature data and the strain data so as to generate a corresponding temperature pulse sequence and a corresponding strain pulse sequence; Continuously comparing the current value of the monitoring data with a dynamically updated reference value, triggering a pulse event when the difference value between the current value and the dynamically updated reference value is greater than or equal to a preset variation threshold value, generating a pulse representing the variation direction of the pulse event, and adjusting the reference value to be the current value; the monitoring data are temperature data or strain data, and the temperature pulse sequence and the strain pulse sequence formed by pulses according to a triggering sequence are formed by continuously triggering the pulse event.
3. The method for compensating for bridge temperature effects based on digital twinning as recited in claim 2 wherein said predetermined variation threshold comprises a temperature threshold set for temperature data changes and a strain threshold set for strain data changes, each pulse event being characterized by a polarity of said pulse and a sequential relationship between pulse events being characterized by a trigger interval in a pulse sequence.
4. The method for compensating the bridge temperature effect based on digital twin according to claim 1, wherein in S3, the temperature pulse sequence and the strain pulse sequence are input into a pre-trained pulse neural network model for time sequence analysis, and the time sequence analysis is as follows: Establishing a correlation time sequence mode between the temperature pulse sequence and the strain pulse sequence under normal thermodynamic response through the pulse neural network model, wherein the correlation time sequence mode is formed based on a pulse time sequence dependent plasticity mechanism; in real-time monitoring, comparing the strain pulse sequence input in real time with the established associated time sequence mode, and judging whether the time sequence deviation occurs to the strain pulse sequence or not based on the comparison result.
5. The digital twinning-based bridge temperature effect compensation method of claim 4, wherein: the pre-trained impulse neural network model is built by the following steps: acquiring historical temperature data and historical strain data of the bridge monitoring area in a structural health state; converting the historical temperature data and the historical strain data into a historical temperature pulse sequence and a historical strain pulse sequence respectively by an asynchronous event driven differential modulation method; taking the historical temperature pulse sequence and the historical strain pulse sequence as training sample pairs, inputting an initial pulse neural network model, and training by adopting a pulse time sequence dependent plasticity mechanism; wherein the training establishes the impulse neural network model in a correlated time sequence mode between the historical temperature impulse sequence and the historical strain impulse sequence under normal thermodynamic response.
6. The method for compensating for bridge temperature effects based on digital twinning according to claim 5, wherein the output of the indication signal based on the result of the time sequence analysis is as follows: outputting an abnormal indication signal when the strain pulse sequence is recognized to deviate from the related time sequence mode, otherwise outputting a normal indication signal.
7. The method for compensating for bridge temperature effects based on digital twinning as recited in claim 6 wherein said sequence of strain pulses deviates from said associated timing pattern as follows: the absence of a strain pulse within a predetermined time window in which the strain pulse occurs, or the burst of the strain pulse occurs without a time-sequence correlation with the temperature pulse.
8. The method for compensating the bridge temperature effect based on the digital twin according to claim 1, wherein in S4, a bridge digital twin model is constructed: obtaining geometric information and material parameters of the bridge monitoring area; generating a thermodynamic coupling constitutive relation based on the geometric information and the material parameters, and constructing a bridge digital twin model by adopting a finite element analysis method; the bridge digital twin model is used for simulating theoretical strain response under temperature data change and is used for carrying out temperature effect compensation on actually measured strain data.
9. The method for compensating the bridge temperature effect based on digital twin according to claim 8, wherein in S4, based on the indication signal output by the pulse neural network model, the strain data is subjected to temperature effect compensation through a pre-constructed bridge digital twin model, and the method is as follows: When the indication signal output by the pulse neural network model is a normal indication signal, the bridge digital twin model compensates the strain data by adopting a first compensation strategy based on a linear thermal expansion theory; And when the indication signal output by the pulse neural network model is an abnormal indication signal, the bridge digital twin model is switched to a second compensation strategy based on nonlinear fracture mechanics to compensate the strain data.
10. The method for compensating the bridge temperature effect based on digital twin according to claim 9, wherein the first compensation strategy is to calculate a theoretical thermal strain value based on a linear thermal expansion theory according to the change of the temperature data and deduct the theoretical thermal strain value from the actually measured strain data; The second compensation strategy is to freeze temperature compensation calculation based on the first compensation strategy by taking the moment of triggering the abnormality indication signal as a datum point, and recognize residual errors between the actually measured strain data and the frozen theoretical thermal strain value after the datum point as components related to structural damage for output or recording.

Description

Bridge temperature effect compensation method based on digital twinning Technical Field The invention relates to the technical field of bridge construction, in particular to a digital twinning-based bridge temperature effect compensation method. Background The bridge deck is paved with conductive concrete, and is an active snow melting and deicing technology, a conductive network is formed by doping conductive materials (such as steel fibers, carbon fibers, graphite and the like) into the concrete, and after the power is on, electric energy is converted into heat energy, so that the temperature of a road surface is raised to be more than 0 ℃ to realize snow melting and deicing, and the bridge deck paved with the conductive concrete snow melting system can experience severe thermal shock due to disappearance of Joule heat effect at the moment of power failure, namely, the concrete structure is rapidly cooled in a short time. Under the working condition, microscopic fatigue cracks are easy to initiate and expand due to the concentration of thermal stress in the material, and the structure is safe for a long time to form a potential threat. In the prior art, the existing structural health monitoring method based on digital twinning faces fundamental technical dilemma under the scene, the initiation and instability expansion of micro fatigue cracks are transient events in microsecond or even shorter time scale, and the heat conduction and the overall temperature field change of bridge members are slow-change processes in the order of minutes to hours. If high-frequency sampling is adopted to capture crack transient signals, massive redundant data are generated, and great burden is brought to transmission, storage and real-time processing, and if low-frequency sampling adapting to a thermal process is adopted, critical transient signals for damage can be completely revealed by leakage, so that monitoring is invalid. Disclosure of Invention In order to solve the technical problems in the background technology, the invention provides a bridge temperature effect compensation method based on digital twinning, which comprises the following specific scheme: a bridge temperature effect compensation method based on digital twinning comprises the following steps: s1, acquiring temperature data and strain data of a bridge monitoring area; S2, based on the temperature data and the strain data, respectively generating a corresponding temperature pulse sequence and a corresponding strain pulse sequence through asynchronous event-driven differential modulation; s3, inputting the temperature pulse sequence and the strain pulse sequence into a pre-trained pulse neural network model for time sequence analysis, and outputting an indication signal based on a time sequence analysis result; And S4, carrying out temperature effect compensation on the strain data through a pre-constructed bridge digital twin model based on the indication signal output by the pulse neural network model, wherein the bridge digital twin model comprises a thermodynamic coupling constitutive relation of the bridge monitoring area. Further, in S2, based on the temperature data and the strain data, a corresponding temperature pulse sequence and strain pulse sequence are generated by asynchronous event-driven differential modulation, respectively, as follows: Asynchronous event driven differential modulation is respectively carried out on the temperature data and the strain data so as to generate a corresponding temperature pulse sequence and a corresponding strain pulse sequence; Continuously comparing the current value of the monitoring data with a dynamically updated reference value, triggering a pulse event when the difference value between the current value and the dynamically updated reference value is greater than or equal to a preset variation threshold value, generating a pulse representing the variation direction of the pulse event, and adjusting the reference value to be the current value; the monitoring data are temperature data or strain data, and the temperature pulse sequence and the strain pulse sequence formed by pulses according to a triggering sequence are formed by continuously triggering the pulse event. Further, the preset variation threshold includes a temperature threshold set for temperature data changes and a strain threshold set for strain data changes, and in the pulse sequence, each pulse event is characterized by the polarity of the pulse, and the sequential relationship between the pulse events is characterized by a trigger interval. Further, in S3, the temperature pulse sequence and the strain pulse sequence are input into a pre-trained impulse neural network model for time sequence analysis, as follows: Establishing a correlation time sequence mode between the temperature pulse sequence and the strain pulse sequence under normal thermodynamic response through the pulse neural network model, wherein the correlation time sequenc