The Structural Health Monitoring System Of Large Span Space Structure Based On Fieldbus

The large and complex space structures have been built more and more, which are complex in the shape, novelty in the structural form, so the safety during the construction and working is concerned. The structural health monitoring can give the working status of the structure using the measurements of the sensors located on the structure, which can give the useful information for the safety of the structure. As the application on the structural health monitoring system on the real structures, the problems such as the choice on the type and arrangement of the sensors, the data collection and management, the structural health monitoring method, the construction of the structural health monitoring system, are studied and researched more and more. The structural health monitoring system based on fieldbus are discussed in the thesis, including the construction of the system and its reliability, the acquisition method on whole structure based on the limited measurements, the optimal placemnts of the sensors, the damage diagnosis on the sensors, the damage extent detection and so on.Firstly, according to the structural health monitoring systems of the real structures and the theory on the fieldbus and distributed control, the structural health monitoring system based on fieldbus is reseached. The structural health monitoring system based on fieldbus is the advanced edition of the structural health monitoring system based on centralized control, which have the higher reliability, stability, anti-interference, system safety, lower cost in maintenance, which has the characteristics such as the expansibilty, the adaptability, the modularization and so on, and which gives the convenience, flexible and unified method for building the structural health monitoring system. According to the advantages of the fieldbus and the structures of the structural health monitoring systems on real structures, the structural health monitoring system based on fieldbus is given, including its components and structure, function modules, hardwares and softwares. In addition, the reliability of the sytem is given based on the failure rate of the sensor subsystems, the basic control units and their connections. The structural health monitoring system of Shenzhen Vanke Center is taken to be an example, several kinds of the structures of the systems are given and it is proofed that the reliability of the structural health monitoring system based on fieldbus is higher than the structural health monitoring system based on centralized and the reliability of the system is higher with the extent of the distributed.Secondly, the structural health monitoring based on the liminted strain measurements is proposed. The aim of the method is to obtain the strain distribution of the whole structure based on the liminted strain measurements; the different numbers of the sensors are located on the substructures according to the simility in the substructures, the strain distribution of the whole structure is obtained by amending the strain distribution of the basic substructure. In the thesis, the principle component analysis method and the system clustering method are used to choose the optimal elements and points for locating the sensors, the fitting method by introducing the continuation is used to obtain the strain distribution of the basic substructure, and the amending method is used to obtain the strain distributions of the other substructure. The method is proofed by the steel structure of the Shenzhen Futian Hinge Center, the identification errors are smaller for the strain distribution of the substructure with higher similarity in the substructures by comparing the theory values and acquisition values.Thirdly, the optimal placement of acceleration senors and the failure dianosis of acceleration sensors based on data fusion are proposed. The method gives the precondition for the effective and reliability of the structural health monitoring seytem and gives the basic theory for realizing the intelligent sensors of the structural health monitoring system based on fieldbus. The coherence data fusion method is used in the thesis, the synthesized support levels and the fusion results can be obtained by calculating the distance matrix and the support matrix. The optimal placement of acceleration sensors is based on the aim to obtaining the optimal identification in modal parameters, the theory measurements are constructed and the optimal placements of acceleration sensors are chosen by the size of the synthesized support level of the theory measurements and real measurements. The failure diagnosis of the acceleration sensors is based on the hardware redundance and time series redundance, the energies of acceleration sensors in frequency domain and the sensor in failure can be known by the small synthesized support level, in addition, the frequency information of the failure sensor can be replaced by the fusion result by the other sensors and used for the diagnosis of the other sensors. The effective and feasibility of the method is proofed on the finite element modle of truss structure and the measurements of acceleration sensors located on the structural health monitoring system of National Aquatics Center.Forthly, the damage extent identification method based on measurements of strain sensors and acceleration sensors is proposed. Two different kinds of measurements are used in the method, which are the strain measurements reflecting the local information of the structure and the acceleration measurements reflecting the whole information of the structure, while the damage extent identification is given from the different points of view, reduce the incompletement of the information, and increase much more known information. Two kinds of damage extent identification methods based on measurements of strain sensors and acceleration sensors are given in the thesis, one is that the synthesized damage extent is obtained by different weight distribution for two damage extent identification results using the D-S evidence theory method; the other one is that the damage extent is obtained by using the combination of the measurements of strain sensors and acceleration sensors to be the input of the neural networks. Through the application on damage extent identification on National Aquatics Center, it is proofed that the damage extent identification based on measurements of the strain sensors and acceleration sensors can synthesize the characteristics of these two kinds of sensors for identifying the structural damage extent.