Research On The Theories And Algorithms Of Processing The Bathymetry Outlier In Multibeam Echosounder System

The outliers of Multibeam Echosounder System are major factors influencing the accuracy of hydrography and seabed terrains production. The algorithm of outlier detection, the method of seabed terrains creating and the real-time processing of soundings are introduced in the thesis. The main work and contributions are summarized as follows:1. True values of seabed terrains can't be obtained by Multibeam Echosounder System. The seabed terrains are effectively reconstructed by functions and interpolations, and they are also compared each other. Then, the data of different types of multibeam sounding footprints are synthesized, and the application of the synthetic data without outliers as known data from"sea floor."2. Outliers in multibeam data can not be totally detected by trend surface model of Least-square, and we present the trend surface model of Robust Estimation. In this algorithm, the outliers are detected by equivalent weight. The robust trend surface algorithm is rather robustness. The regulation of boundary values of equivalent weight IGGâ…¢is analyzed, and it is concluded that the lower the data quality is, the smaller the boundary values should be.3. The functional model and stochastic model of the algorithm were created by multibeam data. The covariance function is calculated by combining with equivalent weight factors. Real data set is processed by the Robust Least-squares Collocation algorithm. It shows that outliers are detected correctly and the estimation is more accurate. The algorithm can also be applied to multibeam data security processing for navigation with accurate seabed display.4. Effects of random errors and the relativity of soundings are calculated in Ordinary Kriging algorithm. As variogram function will be disturbed by outliers, we calculate it in combination with equivalent weight factors by Robust Estimation. Outliers will be detected in the iteration and the precision of the algorithm is better. Then, comparisons were made between the algorithms of Robust Least-squares Collocation and Ordinary Kriging. The results indicated that the difference of precisions between them is small. But the former is better to work with detecting outliers.5. Procedure of soundings processing is improved while creating the seabed terrains joined with detecting outliers. By using the real data, the regular square grid model is created based on Robust Least-squares Collocation, and TIN is formed based on the robust trend surface algorithm. It shows that the grids are estimated accurately with detecting outliers at the same time and the latter results also avoid the influence of outliers. The efficiency and precision of creating DTM of seabed are improved.6. CUBE algorithm is discussed in theory. The availability of multiple hypotheses and the rules to determine the best hypotheses are analyzed in detail. The procedure of real-time multibeam data processing and charting are presented, and theories are provided for real-time quality control on soundings. In addition, the Robust Kalman Filter is used for improving the Kalman Filter in the CUBE algorithm, and the availability of improvement is showed by the experiment.