Research On Multi-source Heterogen-Eous Information Analysis And Data Fusion For Bridge Monitoring

In the course of bridge operation,various safety hazards will appear,and in fact,many bridge collapse safety accidents have occurred.For this reason,it is very necessary to identify bridge damage in time and deal with it accordingly.The shortcomings of traditional bridge inspection methods such as bridge load test and bridge health monitoring system are also obvious.The main problems of load test are time-consuming and laborious,the need to interrupt traffic,and the amount of test information is small,which leads to inaccurate judgment results.Bridge health monitoring exists.The main problem is that the stability of the sensor is a major test and the establishment of a complete and reliable system requires a huge cost.For some large and medium bridges,the cost of health monitoring is very expensive compared with the cost of the bridge itself,and it is difficult to implement.The bridge short-term monitoring technology solves the shortcomings of traditional detection methods.It does not need to interrupt traffic or high expenses,and the test equipment can be recycled.With the development of science and technology and the improvement of people’s requirements for measurement accuracy,scientific researchers are no longer satisfied with one or one type of monitoring equipment to measure the bridge structure.Various and multiple monitoring equipment have been used to perform multiple monitoring on the bridge.Dimensional monitoring.However,the complexity of the collection of these monitoring equipment is very large and the amount of information is large.If only manual processing and analysis is used,the evaluation error rate will be greatly increased.Therefore,it is necessary to study the information fusion technology in the field of bridge short-term monitoring.To this end,this paper takes a simple beam bridge as the research object,analyzes and fusions multi-source heterogeneous data based on short-term monitoring technology to improve the accuracy of bridge damage identification.This article mainly conducts research from the following aspects:1.Since data is the basis of the bridge short-term monitoring system,the first task is to obtain accurate and reliable monitoring data.Combining the characteristics of short-term monitoring data,processing the short-term monitoring data from the time domain,frequency domain,and time frequency respectively,selecting the processing criteria and methods suitable for this article,and writing a program to realize the processing process,and analyze the analog signal and part of the measured data Processing verified the effectiveness of the method.2.Summarize the multi-source heterogeneous detection content of the bridge short-term monitoring,and then analyze and study the multi-source heterogeneous data processing from two aspects and provide a theoretical basis.One is to use the correlation analysis method to monitor the temperature effect from the bridge The strain and deflection data are separated to obtain monitoring data that is not related to temperature changes;the second is to systematically explain the advantages and disadvantages of several commonly used fusion methods and the calculation process of the fusion algorithm,and the classic DS evidence theory is not good for conflict evidence The shortcomings of the fusion,proposed methods to improve the weighted evidence theory.And the fusion efficiency index is used as a metric to evaluate the effectiveness of the fusion method,which provides a theoretical basis for subsequent research.3.Taking a simple beam bridge as an example,the finite element model is established for simulation analysis,and different damage conditions are set to study the sensitivity of damage indicators to damage.The characteristic value of bridge damage is studied from static test index and dynamic test index respectively.The dynamic test index is also divided into frequency domain damage index and time domain damage index.Numerical examples are used to verify the damage indicators in this paper,and the damage indicators that are sensitive to structural damage and easy to obtain in actual engineering are selected as the evidence of fusion in this paper.4.First,an improved DS evidence theory fusion framework is proposed for the improved weighted fusion algorithm proposed in the second part.Then,taking the third part simulation results as an example,the damage indicators under different simulated damage conditions are fused,and the fusion is performed The results are analyzed to verify the effectiveness of the fusion method.Then a damaged operating bridge was selected for short-term monitoring,and the collected monitoring data was processed and feature extracted,and then the extracted feature values ??were fused to judge the health of the bridge,and compared with the actual situation for further verification The effectiveness of the method.