Abstract.Comparison of the approximation performance of the ultrasonic reflections by simple model functions used in blind deconvolution is presented. New model function was proposed where causality of the signal is easily modelled without the need for piecewise-linear equations. Parameter estimation technique was developed. Performance investigation was done for three model functions using four experimentally collected ultrasonic signals under different signal filtering conditions.

Iterative polynomial model for synchronization of vibrations was proposed. A polynomial was constructed using time delay coordinates. For time delay kernel estimation, a non-uniform embedding was used. Evolutionary optimization algorithms were introduced for non-uniform time delay identification. Performance investigation of proposed method was done for two different vibrations.