Model-based parameter estimation for characterizing wave propagation in a homogeneous medium

Abstract A model-based algorithm is presented that uses minimal a priori information and assumptions to adaptively estimate parameters associated with propagating waves in a homogeneous medium. These parameters include transmitter and receiver transfer functions, propagation distances, dispersion, and propagation loss. The algorithm is described in a general framework that accommodates direct arrivals from two or more transmitter–receiver pairs and can be readily adapted to handle application-specific model assumptions. Experimental validation is performed with two sets of ultrasonic guided wave data that conform to two different sets of model assumptions, demonstrating the impact of these assumptions and the ability to successfully estimate model parameters in a dispersive medium.