STATISTICAL OPTIMAL SENSOR PLACEMENT TECHNIQUE FOR STRUCTURAL MODEL UPDATING

Structural model updating utilizing measured dynamic responses depends very much on the quantity and quality of the measured data, which in turn depends on the number of sensors and the corresponding locations. Owing to the problems of measurement noise and modeling error, the results of structural model updating are uncertain in nature. A statistical methodology is presented in this paper for the identification of an effective way to install a given number of sensors on a structure to extract as much information as possible, or equivalently to minimize the uncertainties associated with the results of structural model updating. The information entropy is employed as a measure to quantify the uncertainties of model parameters. The problem of optimal sensor placement is then formulated as a discrete optimization problem, in which the information entropy is minimized, with the sensor configurations as the minimization variables. A computationally efficient numerical optimization algorithm is developed following a Genetic Algorithm (GA) in solving this optimization problem. The methodology is illustrated using a simplified bridge model subjected to an impulse excitation. The optimization solutions from the proposed genetic algorithm are compared to those from an exhaustive search. The results show that the proposed genetic algorithm can identify the "true" optimal solution with much less function evaluations than the exhaustive search.