Parameter Estimation of Structures from Static Response. II: Numerical Simulation Studies

Abstract

The behavior is studied of a force‐error estimator and a displacement‐error estimator using Monte Carlo simulation and a bowstring truss as the model problem. The most important factors governing the performance of any parameter‐estimation algorithm are the quality and quantity of measured data. To assess the performance of the force‐error estimator and the displacement‐error estimator, we examine their behavior with respect to the amount of noise in the measurements, the spatial distribution of measurements, the number of spatial measurements, the number of load cases, and the character of load cases (specifically, loading patterns). In addition to the extensive study of the effects of the physical character of the data and the experiment, we examine the effects of different starting values of the unknown parameters used to initiate the iteration.