Study On Optimal Sensor Placement Criterion And Model Updating Method

The number of sensors arranged in structural health monitoring system is limited due to economic factors,information processing speed,etc. The goal of optimal sensor placement technique is to obtain as much state information as possible by locating fewer sensors in a structure. Model updating technique adjusts and optimizes the initial finite element model to conform to the actual structure based on the information obtained by sensors, thus,the operation state of the structure can be accurately evaluated. Fisher information criterion is widely used as the criterion of optimal sensor placement. Numerical models are used to study the properties of the sensor placement method based on Fisher information criterion. The superiority of the frequency effective independence method with respect to the effective independence method is verified,and frequency effective independence method is applied to locate sensors to collect information for model updating based on frequency response function.Detailed research contents are as follows:(1) For damage parameters identification, when the traditional Fisher information criterion is used for sensor placement, the measuring points are susceptible to gathering in a local sensitivity area, which result in information redundancy. This is not conducive to damage location. To avoid aggregation of the measuring points and improve the ability of damage location, distance coefficient which reflects the degree of information independence is firstly used to correct the Fisher information matrix, and then the measuring points are obtained by maximizing the determinant of the modified information matrix using a sequential algorithm. The method is employed to design the optimal sensor configuration for a simple 16,DOFS chain mass-spring model. The results showed the method can effectively avoid the aggregation of measuring points and solve the problem of information redundancy.(2) The effective independence method locates sensors for modal identification based on Fisher information criterion,and the frequency effective independence method is a generalization of the effective independence method in the frequency domain. A cantilever beam model is used to study the properties of the effective independence method and the frequency effective independence method. The result shows that the effective independence method maximizes the determinant of the Fisher information matrix but does not maximize the linear independence of the target modes. If the target modes chosen in the effective independence method are the same as the excited modes in the target frequency range chosen in the frequency effective independence method, the two methods select the same set of measuring points. The frequency effective independence method is suitable for higher frequency range with high modal density, and eliminates the task of identifying target inodes in the effective independence method.(3) Frequency response function based model updating has more advantages than modal data based one. However, the updated results are affected by several factors, including sensor locations,the frequency range selected and the accuracy of the sensitivity equation. To obtain reliable model updating results, the sensitivity equation of frequency response function is established by perturbation analysis. And the frequency range is reasonably chosen by considering the sensitivity of structural response with respect to parameter variation and the influence of damping on amplitude. Then, the optimal sensor placement method suitable for model updating is further studied to determine the sensor number and locations, which can be used to measure the frequency response data required for model updating. The numerical results show that model updating method based on frequency response function can accurately identify the location and severity of damage.