Algorithms for the measurement of liquid metal coolant flow velocity with correlated thermal signals

The flow velocity of liquid metal coolant can be determined through the measurements of temperature fluctuation recorded by a pair of temperature sensors placed certain distance apart along the flow. Traditionally, this was done using a cross-correlation algorithm to estimate the transit time of the coolant, and thus its velocity. This widely used cross-correlation algorithm, however, suffers from the ambiguity in reading of measurements. To alleviate this problem, the transfer function estimation approach was recently proposed which tends to give more accurate results. In this paper, we introduce a new algorithm that can further improve the accuracy in the transit time estimation using an adaptive inverse system model at a higher cost of computation. When real-time computation is a concern, we propose a second algorithm based on an adaptive filtering approach which makes a sound trade-off between accuracy and computation cost. This algorithm incurs less processing time than the first proposed algorithm with higher accuracy than the two aforementioned conventional algorithms.