Cable-stayed bridge safety assessment method based on variable structure dynamic hybrid Bayesian network

Abstract

According to the cable-stayed bridge safety assessment method based on the variable-structure dynamic hybrid Bayesian network, a cable-stayed bridge safety assessment and prediction model is established by using the variable-structure HDBN, BN topologies under all time slices are consistent, and based on a mechanical method and cable stress influence matrix construction, the cable-stayed bridge safety assessment and prediction model is established by using the variable-structure HDBN. The continuous nodes between the adjacent time slices are connected through directed arcs pointing from the previous time slice to the next time slice; a variable-structure HDBN is formed by setting the same transfer network topology and different conditional probabilities among time slices, so as to reflect the time sequence change of the external load level of the cable-stayed bridge; when evaluation is carried out, the maximum stress value, monitored in a certain period of time, of the inhaul cable serves as evidence, a corresponding time slice t is automatically matched, the known evidence is substituted into a corresponding node of the time slice t, all node values of all the time slices of the model are reasoned, and the state probability of system nodes serves as the basis of safety evaluation; according to the method, the safety performance of the cable-stayed bridge can be effectively evaluated only according to a small number of cable node evidences.

Description

technical field [0001] The invention relates to the technical field of bridge engineering, in particular to a safety assessment method for a cable-stayed bridge based on a variable-structure dynamic hybrid Bayesian network. Background technique [0002] Safety accidents of long-span bridges often cause heavy casualties and huge economic losses. Therefore, it is extremely necessary to conduct real-time safety assessments in combination with health monitoring data during bridge service. However, the amount of monitoring data is large and there are many types. How to process, mine and extract the damage characteristics of components (parts) to realize the assessment of the safety of the bridge system has not yet formed a unified standard. In addition, due to the limitation of monitoring cost and site layout conditions, the number of sensors in the actual monitoring system is often limited, making the data incomplete, which is a key problem that needs to be solved urgently in th...

AI Technical Summary

Problems solved by technology

However, compared with pure continuous or discrete BN, the node variable processing of hybrid BN is more complicated

Images