Strong Scattering Targets Separation Based on Fractional Fourier Transformation in Pulse-to-Pulse Coherent Acoustical Doppler Current Profilers

摘要

The influence of the strong scattering targets (SSTs) on the flow cell velocities, i.e., the fish-bias effect, is a common phenomenon existing widely in real-time measurements of ocean and river current profiles when using pulse-to-pulse coherent acoustical Doppler current profilers (ADCPs), because the SSTs with different velocities and excessively high echo strengths are inevitable in the ocean and river background flow mediums. However, there are no effective means to eliminate the SST-caused biases and recover the current profiles accurately. The common solutions to solve the SST problem either neglect the influence simply, or abandon the affected data directly, which of course causes wrong current profiles or even invalidation of the on-site measurement. Therefore, this paper proposes an SST separation method to eliminate the SST-caused biases in ADCPs. By employing a fractional Fourier transform (FrFT)-based method, the SST echoes are separated from the background ones, so that the accurate estimation of both velocities of the flow cells and the SSTs are obtained through respective signals. First, the proposed SST separation method based on FrFT is theoretically analyzed and validated through a simulation model of flow measurement scenarios, which is constructed with a large number of distributed point targets and several SSTs. Then, an ADCP core algorithm is implemented with traditional correlation theory after the SST separation. Next, the performance of the proposed method is evaluated with different-strength SST signals being superposed to the background echoes. At last, the computational complexity of the FrFT algorithm is evaluated to show that the SST separation algorithm does not affect the real-time measurements of the current velocities significantly. The simulation results show that, even when the averaging power of an SST is 60 dB stronger than that of the averaging background echoes, accurate flow velocities of the affected flow cells can be recovered. Moreover, up to five SSTs around one flow cell with a total power 60 dB higher than that of the background echoes can be separated effectively by the separation algorithm with an affordable computational cost, with which both the current profile and the velocities of the SSTs are accurately obtained.