A model-based adaptive algorithm for determination of time-of-flight in ultrasonic measurement

A new method of estimating the time-of-flight of ultrasonic waves is proposed in this paper. The method utilizes an ARMA model to represent the ideal reflected wave. The model is then used to remove each reflected wave successively so that more accurate estimates of the time-of-flight can be obtained for subsequent reflections. The performance of this method has been evaluated using the measurement data ranging from a simple carbon brick to more complex industrial blast furnaces. The estimated time-of-flight can be used to monitor the conditions of the furnace and to detect possible defects.