Data Processing Technology In Multi-Beam Water Flow Measurement

Multi-beam water flow measurement is becoming a hot spot in acoustic Doppler flow measurement because of its wide coverage,high accuracy and high efficiency.Since the multi-beam fluid echo data is more complex than the common fixed-beam ADCP,there is an urgent need for suitable processing methods.Aiming at satisfying this demand,this paper conducts a comprehensive water flow measurement methods research based on multi-beam layered radial velocity and echo intensity data.Main tasks are as follows:Firstly,the basic principles of common fixed-beam ADCP flow measurement and multi-beam flow measurement are briefly introduced,and the characteristics of multibeam data are analyzed,then leading to the overall process of data processing.Secondly,the influencing factors of flow velocity and echo intensity data in the process of multi-beam water flow measurement are analyzed in detail,and processing methods to improve the data quality are proposed separately.For the radial velocity data,decision factor of water layered units measurement outlier is established based on echo signal autocorrelation analysis;an outlier compensation method based on block multiple regression is proposed,and multi-beam layered radial velocity data uses twodimensional filter based on two-dimensional spatial correlation,used to improve the quality of the measured data.For the echo intensity data,a method of compensation for the acoustic wave propagation loss is given according to the sonar equation,and joint denoised using Gaussian filtering and median filtering.Aiming at the difference in data quality caused by the unevenness of underwater scatters,a multi-beam layered data processing method is proposed based on autocorrelation estimation and echo intensity and the effectiveness of the method is verified by simulation analysis.Also aiming at the beam asymmetry problem during the conversion of radial velocity to threedimensional coordinate system velocity in multi-beam velocity measurement,the projection method is used to derive the coordinate transformation matrix on different side and the same side of transmitting fan,which implements the high resolution measurement of the three-dimensional flow velocity,and the simulation verified the correctness of the matrix.Thirdly,Aiming at the application requirements of analysis and extraction of water body turbulence characteristics,the extraction methods of turbulence characteristics based on acoustic multi-beam water body echo in time and space are studied.The nearcircumference method is used to simulate the vortex flow field generated by the propeller disturbance.Feature extraction method based on multi-beam layered echo secondary correlation and Doppler shift correlation are proposed,and vortex simulation data verify the effectiveness of methods.Finally,algorithm software development is completed on the hardware platform based on GPU.According to the characteristics of diversity and independence of multibeam flow data processing,modularity and multi-threaded processing method are designed.At the same time,the measurement data obtained by the pool experiment is analyzed and processed,partially verifying the effectiveness and correctness of the data processing method in this paper.