The Study Of Optimal Sensor Placement For Civil Engineering Structure Based On Modern Intelligent Algorithm

With the rapid development of national economy, civil engineering structures are growing, structural health monitoring(SHM) has becoming one of research hotspot of scholars both at home and abroad. To obtain structure information accurately and timely through reasonable configuration of sensors is the precondition to realize CESHM, and is the important content of civil engineering structure monitoring and assessment. Traditional optimization methods have been applied to the optimal sensor placement(OSP), but all have their own limitations. Modern intelligent algorithms as a random, integrity and heuristic algorithm, haved showed great superiority, and have been widely used in the field of OSP, and have obtained the good effects of monitoring.In this paper, the OSP problem in civil engineering structure was studied based on modern intelligent algorithms by using Modal Assurance Criterion(MAC), and then the monitoring purposes can be completed by using fewer sensors to get more structure health data. the main research content is as follows:(1) The necessary theoretical elaboration about Genetic Algorithm(GA) and Simulated-annealing Algorithm(SA) was done to prepare for the follow-up studies of OSP.(2) The finite element modeling of civil engineering structure was established by ANSYS software, and then the modal analysis was done, and then low order modes was gotten, and the mass matrix and stiffness matrix were extracted.(3) Investigations on the OSP respectively for three different kinds of structure forms high-rise shear wall structure, steel truss bridge, multilayer frame structure were carried out based on GA, SA and Simulated- annealing Genetic Algorithm(GA-SA). The goal of optimization includes the quantity and location choice of measuring point. The study results show that, the above three methods all can achieve OSP for three different kinds of civil engineering structure, and each method provides a set of practical and feasible optimization scheme according to different structure.(4) By analyzing three methods, it can be concluded that GA has obvious superiority about optimization results of the number of measuring points, and SA and GA- SA are lower on running time costs.