Research On Multi-Frequency Acoustic Measurement Of Suspended Sediment

Suspended sediment is very important for port construction and comprehensive sediment control.Compared with traditional suspended sediment measurement tech-nology,acoustic measurement technology has attracted more and more attention in recent years due to its non-intrusive measurement,high spatial and temporal resolu-tion and profile measurement.The transducer transmits acoustic pulse to suspensions,sediment produces backscattered signal which carries the suspended sediment information,and the in-formation is obtained by applied inversion algorithm on the backscattered signal.The theory of acoustic backscattering was discussed,including the propagation property,attenuation property,backscattering property and the theoretical equation.Then the inversion algorithms were studied.And the error of the inversion algorithms under different simulation condition was compared.In this thesis,a multi-frequency acoustic backscattered system and an experi-mental platform was built.The statistics of the backscattered signal was studied and results showed that the distribution of the backscattered signal was Rayleigh and for the square of backscattered signal was Exponential.The usage of glass spheres de-coupled the calibration of system parameter,and the calibration results showed that the system parameter was a constant.Then a series of measurement experiments was carried out with the calibrated system,the results showed that the system we built can profile the concentration and particle size of the suspended sediment at a low concen-tration.Based on the K-nearest neighbor algorithm,a simple and efficient concentra-tion inversion algorithm was proposed.And its feasibility was preliminarily verified by the experimental results.Finally,an experimental study was carried out on the phenomenon that the con-centration measurement error increased rapidly when the concentration was further increased(greater than lOkgm-3).By increasing the voltage amplitude,the measure-ment error was effectively reduced,and the relationship between SNR and the meas-urement error was qualitatively verified.