The Research On Comparation And Optimization Setting Of Sensor In Bridge Health Monitoring

With the rapid development of science and technology, Civil Engineering Structures, such as water-power engineering, harbor engineering, ocean engineering of hydraulic structure and bridge, tunnel, industrial and civil construction of civil engineering structure, its'scale and difficulty of construction are increasing. Problems like material deterioration and damage accumulation will be formed in structures during their service, which are caused by design load and varieties of environmental factors, so the security and stability of the structures will be affected, and direct or indirect economic losses caused by failure or sudden destruction will be serious. Theoretical and practical significance can be achieved by damage identification and real time monitor for vital structures.Sensor system is an important component of the bridge structural health monitoring system. Theoretically, the more sensors are used, the more comprehensive structural response information is obtained. But because of running state of structures and economic reasons, for a bridge it is impossible to place sensors on every degree of freedom. The fundamental principle of placement optimization of sensors is using as few as sensors to obtain reliable and comprehensive information of bridge health state.This paper summarized the comparison and selection, optimal placement and the present research and the newest progress of sensor comprehensively. By connecting the finite element model of Mashuihe-Bridge, it developed the modal frequency and vibration mode. Besides, it discussed the rules of effective independence idea (EFI) and modal assurance criterion of sensor's optimal placement detailed. And it applied the later method on Mashuihe-Bridge. The main content is as following:(1) The principle of comparison and selection of sensors is Proposed based on the applicability, Reliability, Durability and economy of sensors.(2) Based on the modal assurance criterion, the paper considered the results of QR decomposition as initial allocation and optimized the allocation of sensors by using the revising step by step accumulating method which developed in this paper to make the maximum nondiagonal matrix component reaching the preset value sooner and complete the optimal allocation.(3) Based on the Strain Modal Assurance Criterion, this paper optimized the allocation of sensors and worked out the place of some sensors according to the message of matrix of strain mode. And then, it completed the combinatorial optimization of places of the rest sensors by applying the improved genetic algorithm.