Research On Information Fusion Method Of Arch Dam Structure Safety Monitoring

The environment of high arch dam structure is complex,which will lead to the reduction of structural safety performance,various damage and cracks,etc.in serious cases,it will lead to structural instability and safety problems.Therefore,to ensure the safe operation of the arch dam structure and improve the operating efficiency of the structure is very important for the study of the safety monitoring of the high arch dam structure.In recent years,it is a hot topic to use the dynamic test signal to monitor the operation status of the structural engineering.The acquisition of the real signal is the premise of the safety monitoring of the structural operation status.The signal is dynamically fused by the information fusion technology to obtain the overall operation characteristic information of the structure,so as to increase the accuracy of the monitoring results and ensure the safe and stable operation of the structure.In this paper,the Laxiwa arch dam is taken as the research object.Using the on-line monitoring data of the structure,the safety operation state of the high arch dam structure is monitored and studied from the structure feature information extraction and multi-source information fusion technology.(1)Dam monitoring signal processing.The reason of dam structure vibration is complex,which affects the purity of structure vibration signal.The direct analysis of monitoring signal will affect the reliability of results.In this paper,AVMD-SVD method is used to filter the monitoring signal,and it is applied to the simulation experiment with other methods.The results show that this method has better noise reduction ability,can effectively filter out the high-frequency signal and avoid the interference of low-frequency signal.In order to further ensure the practicability of the method,AVMD-SVD method is used to process the monitoring data of Laxiwa arch dam,and the obtained signals can better reflect the real vibration state of the structure and further ensure the accuracy of structural operation analysis.(2)Research on information fusion method of vibration signal.In order to make full use of the monitoring data and complete the dynamic fusion,the correlation function algorithm and variance contribution rate method are introduced,and the data of the multi measuring points of the Laxiwa arch dam structure are processed.The structure shows that the fusion results of variance contribution rate method can better reflect the overall characteristics of the structure,and the data after fusion contains more feature information.In order to get the operation status of the structure at different times accurately,the relationship between the monitoring data and the operation status of the arch dam structure is used to establish the HMM model,which takes the monitoring information of different types of arch dam as the input data,and then obtains the dynamic decision results of the structure at different times.The results show that the structure needs to be overhauled to prevent the damage from aggravating at time 17.In order to make full use of the monitoring data under multiple working conditions,mwmpe model is established in this paper.In this method,the entropy change curve after fusion is used to comprehensively reflect the operation state of the structure.When the operation state of the arch dam structure changes,the entropy changes.When the arch dam structure runs stably,the entropy value is basically unchanged.Therefore,mwmpe method can realize the real effect of the multi-point data on the dam structure Time on-line monitoring to ensure the safe operation of the structure.(3)Research on prediction of structural vibration.It is the key and difficult problem to predict the vibration trend of the structure engineering by using the dynamic monitoring signal.By accurately predicting the vibration trend of the structure,and then characterizing the operation of the structure,to ensure that the operation of the structure is at a safe level.The results show that AVMD-KELM model can save training time and has strong generalization ability in the process of prediction,and also has good effect in the prediction of arch dam structure at multiple measuring points.Considering the coupling of multiple measuring points,the AVMD-KELM prediction model is established for the data of 13 measuring points on the top of arch dam after fusion,which achieves good results and further reflects the validity of the prediction model.At the same time,the data after fusion can make full use of massive information,simplify the amount of data,and the obtained signals can better represent the overall vibration trend of the structure,providing a strong reference for the realization of online monitoring of the operation of the structure.