Research And Application Of Steel Structure Building Health Monitoring System

Since the reform and opening-up policy,China’s infrastructure construction has made great strides forward,with various emerging technologies and new materials being developed and applied,which has effectively promoted the transformation of the construction industry from traditional brick-and-tile structures to modern architectural styles dominated by metal structures.As steel structures are exposed to natural environmental conditions for a long time,their internal structures are prone to damage,which requires a set of health monitoring system to be designed for in-service steel structures,so as to record the operating conditions of steel structures in real time and avoid accidents.This article takes the steel roof of a museum exhibition hall as the research object,and uses wireless sensor network technology to collect data,locate fault points,predict micro-strains,and evaluate the health status warning level of the structure.The main research contents are as follows:(1)A comprehensive architecture of the health monitoring system for steel structures was designed and established.The factors affecting the structural composition of the monitoring system were analyzed,and sensor models suitable for monitoring were selected.The installation positions were arranged in consideration of the characteristics of the arch structure.Based on the requirements for the system implementation in the later stage,the functions that the sensing subsystem should have were proposed.(2)An improved Sparrow Search Algorithm(ISSA)was proposed to optimize the DV-Hop node location algorithm for solving the problem of low accuracy in fault location in steel structures.Firstly,four communication radii are used to refine the hop count between nodes;Then,the weighted processing average jump distance is used for correction;Finally,the improved sparrow search algorithm is used to locate the fault point.Experiment passed MADIS＿GEN software performs modeling and dimensionality reduction processing,and applies the improved ISSADV-Hop and traditional DV-Hop,IPSODV-Hop,and IGWODV-Hop algorithms to fault location for simulation.The results show that the normalized location error has been reduced by19.64%,14.87%,and 8.96%,respectively.The improved algorithm improves the positioning accuracy,and is more suitable for fault location in steel structures.(3)In order to address the issue of low accuracy in predicting micro-strains in steel structures,this study proposed a micro-strain prediction model based on IWOA-VMD-Bi-LSTM.Firstly,the improved whale optimization algorithm(IWOA)was used for global optimization of the parameters K and penalty factor α of VMD.Then,a prediction model based on Bi-LSTM was used to separately train each modal component.Finally,the predicted values were combined to obtain the final result.Experimental simulation results demonstrated that this prediction model has higher prediction accuracy for deformation in steel structures,with evaluation indicators superior to the other eight models,and the effectiveness of the proposed model was verified.(4)The health monitoring software system for steel structures was implemented.The data visualization of the parameters related to the monitoring points was displayed,and the future operating status of the monitoring points was predicted and analyzed for timely processing.The MADIS＿GEN software was used to model the actual damage state of the steel and calculate the warning limit value.If the measured value exceeds the warning value,the safety warning function will be triggered,and the maintenance personnel will be notified to conduct a safety check on the steel structure.