A Combined Interpolation Method for Cross Correlation Based Particle Velocity Measurement

Cross correlation techniques have been proved to be a valuable tool for online continuous velocity measurement of particulate solids in pneumatic pipelines. In order to reduce computational complexity the sampling frequency is usually kept as low as possible, and the peak in the correlation function is found by interpolating the correlation function. Parabola functions are commonly used as parametric models of the cross correlation function in time delay estimation. However, the parabolic-fit interpolation method introduces a bias at low sampling rate to the center frequency ratio of input signal. In this paper, a combined interpolation method is proposed to improve the estimation accuracy. Experiments are carried out to evaluate the performance of the proposed interpolation method for low sampling rate. The experimental results have been promising and have shown the potential of the proposed method for particle flow velocity measurements.