Stochastic Analysis of Human Exposure Assessment by Surrogate Model

Abstract

Despite recent advances in the field, the stochastic analysis of EM models is still time-consuming. Thus, the approximation of the EM model by a surrogate model is often the only choice if a sufficient amount of simulations is required. This paper provides a systematic and novel methodology to implement the polynomial chaos expansion surrogate models in stochastic analysis. The necessary steps in the construction of surrogate model based on polynomial chaos expansion are explained. As an example, we estimate the statistics of the specific absorption rate of a smart watch GSM antenna on a human wrist. Results show that the proposed method is accurate and time-saving in the cases when numerous rounds of simulation are required.