CN-121835437-B - Pier emergency shock absorption monitoring system based on damper failure linkage

CN121835437BCN 121835437 BCN121835437 BCN 121835437BCN-121835437-B

Abstract

The application relates to the technical field of bridge monitoring, in particular to a bridge pier emergency shock absorption monitoring system based on damper failure linkage. The bridge pier emergency shock absorption monitoring system based on damper failure linkage comprises a vibration information acquisition device, a damper working signal acquisition device and a vibration information acquisition device, wherein a plurality of vibration sensors are arranged on a bridge pier to acquire vibration signals of the bridge pier, the damper working signal acquisition device is respectively connected with damper signals arranged on the bridge pier to acquire monitoring signals of the bridge pier, the system is based on bidirectional compensation mechanism modeling and life attenuation process dynamic coupling among dampers, the problem of prediction distortion caused by neglecting failure accumulation effect in traditional monitoring is fundamentally solved, and accurate early warning of bridge shock absorption function cascade failure is realized through vibration time difference positioning compensation topological chain and system level collapse risk simulation.

Inventors

LI FUHAI
LUO SEN
LI HAONIAN
XU TENGFEI

Assignees

西南交通大学

Dates

Publication Date: 20260508
Application Date: 20260311

Claims (7)

1. Emergent shock attenuation monitoring system of pier based on link that attenuator became invalid, its characterized in that includes: the vibration information acquisition device is used for acquiring vibration signals of the bridge pier, wherein a plurality of vibration sensors are arranged on the bridge pier; the damper working signal acquisition device is respectively connected with the damper arranged on the bridge piers in a signal manner to acquire monitoring signals of the bridge piers; the monitoring signals comprise damping force signals and displacement signals at two ends of the damper; The three-dimensional model simulation device simulates a three-dimensional structure of the bridge so as to mark the positions of each damper and each vibration sensor; the damper compensation relation acquisition device is used for respectively acquiring vibration signals of the bridge pier and generating a compensation relation diagram of the damper based on vibration signal time differences of all vibration sensors and monitoring signal changes of all dampers; the device for monitoring the service life state of the damper monitors monitoring signals of the damper in real time, and acquires the service life state of the damper based on the monitoring signals; The bridge risk monitoring device acquires the life state transition process of each damper, and generates systematic risk probability of collapse of the bridge damping system based on the compensation relation diagram and the life state of each damper; Bridge risk monitoring device includes: the compensation relation updating module acquires the compensation relation diagram in real time to generate a compensation relation diagram updating sequence; the service life state updating module of the damper acquires the service life state of each damper in real time so as to generate damper service life state updating data; The damper failure risk calculation module generates global damper systematic failure probability in the next monitoring period according to the compensation relation diagram updating sequence and damper life state updating data; the damper failure risk calculation module includes: The self-adaptive graph convolution layer fuses the compensation relation graph update sequence and the damper life state update data to generate a node characteristic matrix; The compensation dynamic modeling layer captures the rule of the point life state of the node feature matrix along the time dependent compensation relation propagation, and generates hidden features; the systematic failure probability prediction layer is used for aggregating the global risk life state based on the hidden characteristics and outputting the systematic failure probability of the global damper; The systematic failure probability prediction layer comprises: Node level risk assessment network based on hidden life state Generating the probability that all dampers transition to a final life state in the next monitoring period; clustering the division network, and dividing each damper into a plurality of sets by using a spectral clustering algorithm; an aggregate risk computing network that computes risk vectors for each set based on a gated attention mechanism; and the global risk calculation network is used for calculating the systematic failure probability of the damper based on the risk vectors of the sets.
2. The pier emergency vibration reduction monitoring system based on damper failure linkage according to claim 1, wherein the vibration sensor and the damper are matched with each other, and 1 vibration sensor is fixed at a fixed position of each damper; the damper compensation relation acquisition device extracts a time window in which the position of the damper is vibrated from vibration information, and generates an association relation of the damper based on the similarity of monitoring information of each damper in the same time window; the association relationship includes a high association, a medium association, and a low association.
3. The pier emergency vibration reduction monitoring system based on damper failure linkage according to claim 2, wherein the association relation is obtained by the following way: S1, acquiring vibration signals ST i of each vibration sensor S i , carrying out Fourier feature processing on each vibration signal ST i , extracting a feature distortion window SR i of each vibration signal, wherein i represents the index of the vibration sensor; S2, acquiring a monitoring signal GT j of each damper G j , extracting a characteristic distortion window SR j-i of a vibration sensor S i at a position corresponding to each damper G j , and obtaining a characteristic time window GR j of each damper G j , wherein j represents an index of the damper, and j-i represents that the sensor S i and the damper G j are arranged at the same position; S3, extracting a monitoring signal segment uGT j of the damper G j in a characteristic time window GR j , extracting monitoring signal segments uGT v of the rest dampers G v in a characteristic time window GR j , sequentially calculating the similarity between uGT j and uGT v , and if the similarity is higher than a preset threshold, associating the damper G j with the damper G v ; S4, counting the incidence relation probability of the damper G j and the damper G v under all characteristic distortion windows SR i , and setting the unilateral relation between the damper G j and the damper G v as high incidence if the incidence relation probability is not smaller than a first threshold value; If the association probability is lower than the first threshold value and not lower than the second threshold value, setting the unilateral relationship between the damper G j and the damper G v as moderate association; If the association probability is smaller than the second threshold, the one-sided relationship of the damper G j and the damper G v is set to be a low-degree association.
4. The pier emergency vibration reduction monitoring system based on damper failure linkage according to claim 3, wherein the compensation relation graph is a two-dimensional topological graph connected based on association relation; The damper is a node in the two-dimensional topological graph, the association relation is a directed line segment in the two-dimensional topological graph, and the high association, the medium association and the low association respectively correspond to the weights of the line segments in the two-dimensional topological graph.
5. The pier emergency vibration reduction monitoring system based on damper failure linkage according to claim 1, wherein the damper life state monitoring device comprises: the life state optimizer is used for acquiring monitoring signals for converting a plurality of dampers from health to failure so as to define the number of life states of the dampers; An antagonism network trainer, a discriminator based on the life state of the antagonism network training damper; And the life state monitor monitors monitoring signals of all the dampers by adopting the trained discriminators so as to acquire the current life state of each damper.
6. The pier emergency vibration reduction monitoring system based on damper failure linkage according to claim 1, wherein the compensation relation graph updating sequence is as follows ; T represents the number of the compensation relationship graphs; ; Representing a compensation relation diagram of a t-th monitoring period in the compensation relation diagram updating sequence, wherein t represents an index of the monitoring period; A set of nodes is represented and, Represents the directed edge set of the compensation relationship graph, Representing an incidence relation matrix of directed edges in the compensation relation graph; ; ; ; , wherein J represents the total number of dampers, i and J both represent the index of the damper, i corresponds to And Corresponding to the column j of And In the number of rows of the column, The relationship of damper i to damper j is shown, Indicating that damper i is not associated with damper j, Indicating that damper i is associated with damper j; indicating that damper i is not associated with damper j; representing a low degree of association of damper i to damper j; representing a moderate correlation of damper i to damper j; Indicating that damper i is highly correlated to damper j.
7. The pier emergency vibration reduction monitoring system based on damper failure linkage according to claim 6, wherein the damper life status update data is ; ; A life state column of the damper representing a T-th monitoring period in the life state update data of the damper, wherein T represents an index of the monitoring period; ; ; ; ; Wherein, the A set of states of life is represented, Respectively representing a first life state, a second life state and a last life state of the damper, Q represents a total number of life states, Representing the q-th state of life of the damper; representing the state of life of damper j at the t-th monitoring period; Indicating that the life state of the damper j at the t-th monitoring period is 。

Description

Pier emergency shock absorption monitoring system based on damper failure linkage Technical Field The application relates to the technical field of bridge monitoring, in particular to a bridge pier emergency shock absorption monitoring system based on damper failure linkage. Background The statements in this section merely provide background information related to the present disclosure and may not constitute prior art. The bridge damper is a core device for vibration control of a civil engineering structure, and has the function of remarkably reducing the dynamic response of a bridge main body structure by dissipating external excitation energy (such as vehicle load, wind vibration and earthquake action). When the bridge is impacted by dynamic load, the damper generates hysteresis energy consumption through viscous, metal yielding or friction mechanisms, so that the displacement amplitude of the structure is effectively restrained, and the fatigue damage and even the instability damage of the component caused by resonance effect are avoided. The current monitoring method for failure of the bridge damper is mainly based on a prediction model driven by multi-source time-varying parameters. By collecting data such as service life of a bridge, traffic load frequency spectrum, environmental temperature and humidity, seismic event record and the like in real time and combining monitoring of a sensor network on working life states (such as displacement and output) of a damper, a failure probability prediction model based on deep learning (such as LSTM and CNN) or a traditional neural network is constructed. The model sets a threshold triggering mechanism, and automatically triggers an early warning signal when the number of the predicted failure dampers exceeds the preset design redundancy. However, the prior art approach severely ignores the failure accumulation effect of damper clusters. The damping system is used as a whole for cooperative work, and after a single damper fails, the energy consumption task born by the damping system is forcedly transferred to an adjacent damper, so that the latter is in an over-design working condition operation life state for a long time. This compensatory action causes local overload to significantly accelerate peripheral damper fatigue damage, forming a "failure-compensatory-re-failure" vicious circle. In this way, the neural network training process is forced to fit pseudo failure data with compensation interference (namely, the model actually learns the apparent failure rule under the compensation mechanism), so that systematic deviation exists between the prediction result and the actual failure process. Disclosure of Invention The summary of the application is provided to introduce a selection of concepts in a simplified form that are further described below in the detailed description. The summary of the application is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. Some embodiments of the application provide a pier emergency shock absorption monitoring system based on damper failure linkage, which solves the technical problems mentioned in the background art section. As a first aspect of the present application, some embodiments of the present application provide a pier emergency vibration reduction monitoring system based on damper failure linkage, including: the vibration information acquisition device is used for acquiring vibration signals of the bridge pier, wherein a plurality of vibration sensors are arranged on the bridge pier; the damper working signal acquisition device is respectively connected with the damper arranged on the bridge piers in a signal manner to acquire monitoring signals of the bridge piers; the monitoring signals comprise damping force signals and displacement signals at two ends of the damper; The three-dimensional model simulation device simulates a three-dimensional structure of the bridge so as to mark the positions of each damper and each vibration sensor; the damper compensation relation acquisition device is used for respectively acquiring vibration signals of the bridge pier and generating a compensation relation diagram of the damper based on vibration signal time differences of all vibration sensors and monitoring signal changes of all dampers; the device for monitoring the service life state of the damper monitors monitoring signals of the damper in real time, and acquires the service life state of the damper based on the monitoring signals; And the bridge risk monitoring device acquires the life state transition process of each damper, and generates systematic risk probability of collapse of the bridge damping system based on the compensation relation diagram and the life state of each damper. According to the method, the compensation linkage behavior among the damper clusters is accurately analyzed, so that