Optimization And Reconstruction Of Multi-type Monitoring Information Based On Correlation

The sensor is the front-end equipment for structural health monitoring. The basis of structure evaluation is getting the massive monitoring data through the sensors. The validity of sensor data and its reasonable analysis and application directly affect the safety assessment of structural health monitoring. There is a certain correlation between the massive monitoring data, and it is necessary to optimize, evaluate and reconstruct structural monitoring information through the correlation of the structural response.The range determination of structural response method and optimal sensor placement method are proposed. The correlation coefficient matrix and regression coefficient matrix of structural response are established through correlation coefficient and regression coefficient. The structure response associated area is determined by the threshold of the correlation coefficient and the regression coefficient. The optimization method of the sensor placement is determined by the associated area to assess the structural response of the region. A plate space frame model is taken as an example, the optimal displacement monitoring points are selected and the structural displacement is predicted. According to the error analysis of the predicted information, the proposed method is feasible and effective.The method of monitoring information reconstruction based on correlation degree is studied. The correlation degree between the fault points and other measuring points is obtained by calculating the normal data, which is the correlation degree between the reconstructed variable and the response variable s. and the number of response variables will be obtained by the rank of correlation degree. The reconstruction equation of different response is established by partial least square regression method. The factors affecting the reconstruction error which include correlation degree, the quantity of monitoring data, response variable number are discussed. A plate space frame model is taken as an example, the stress data and displacement data is respectively reconstructed by stress data. The three factors affecting the reconstruction error are discussed. According to the error analysis of reconstruction information, the effectiveness of the method and the measure of improving the accuracy of the reconstruction error are verified.The optimization and reconstruction of monitoring information of Shenzhen Bay sports center is researched based on the measured data. The position of the fault sensor and the maximum stress point is d etermined by the analysis of the measured data. The influence factors of the reconstruction error are discussed. Taking two fault points as example, the stress of the fault point and the maximum point are reconstructed. The reliability of the method is proved by comparing with the information reconstruction of different periods, and the stability of the stress association model is proved. By calculating the correlation coefficient and regression coefficient between displacement measur ing points and other nodes, the associated area of the displacement measuring points is determined with the correlation region method.