Research On Thinned Array And Near-field Imaging Algorithm For Phased Array Three-Dimensional Sonar

Phased array three-dimensinal(3D)sonar is a new type of three-dimensional acoustic imaging equipment,which was developed in recent years to meet the increasing demand of 3D detection of underwater targets.With excellent imaging resolution and real-time imaging performance,it can be applied for target detection in turbid water in the harbor and coastal regions.There is a wide range of applications in port security,underwater engineering,marine science research,aquaculture and defense.Two critical issues in the development of phased array 3D sonar systems are the high hardware complexity of system and the angle coverage limitation of near-field imaging.In order to address those two critical issues,this dissertation researches thinned array and near-field imaging algorithm for phased array 3D sonar.The research work will contribute to reducing the hardware complexity of system and enlarging the angle coverage of near-field imaging,and advance the state of phased array 3D sonar and related technology in both theoretic and practical aspects.The main research contents and achievements are listed as follows.Considering the transducer array performance requirements of phased array 3D sonar,the variables,constraints and objectives of thinned array optimization problems are analyzed according to the function relation between the excitation of transducers and array pattern.The analysis focuses on the constraints and related handling methods of side lobe level and array gain in the thinned array optimization.A theoretical model of both far-field and near-field thinned array optimization problem is proposed based on the above analysis.Moreover,the models of echo signal and beamforming algorithms are established,and the difference of signal processing in far-field and near-field are analyzed.The study reveals the theoretical applicability of the near-field beamforming algorithm that based on traditional Fresnel approximation delay.For the high hardware complexity of phased array 3D sonar system caused by the large scale of receiving transducer array,this dissertation studies the thinned array optimization method in both heuristic and deterministic aspects.From the heuristic aspect,a modified teaching-learning based optimization(MTLBO)algorithm is proposed.The MTLBO algorithm has fast convergence and requires no algorithm specified parameter.Numerical results show that the performance of MTLBO outperforms other methods through comparison.From the deterministic aspect,an improved sequential convex optimizations(ISCVX)method is presented.The ISCVX requires fewer iterations,and the optimization results of ISCVX method are deterministic yet very close to those of the heuristic optimization algorithm,which is more suitable for practical engineering applications.The optimization example of transducer array for typical phased array 3D sonar system shows that the proposed methods can effectively reduce the hardware complexity of sonar system,with over 86% thinned rate of transducer array.For the angle coverage limitation of near-field imaging in the phased array 3D sonar system,a near-field sub-regional time delay(NSTD)method,based on the Fresnel approximation,is proposed in this dissertation.The NSTD method can broaden the angle coverage of near-field imaging,and maintain the delay accuracy of the original Fresnel delay area.Based on the NSTD,near-field beamforming algorithms based on FFT and CZT are proposed respectively.Compared to the traditional direct beamforming algorithm,the proposed method can greatly reduce the parameter storage requirement and computational load during the imaging process.The theoretical analysis and simulation results show that the proposed near-field imaging algorithm can extend the effective detection area of phased array 3D sonar system by over 40%.In consideration of the reqirements for high resolution and wide viewing angle performance of phased array 3D sonar system,design thinned array in both far-field and near-field.A systematic imaging algorithm is proposed and simulated for the thinned array based 3D sonar system considering the difference between far-field and near field imaging.An experimental platform was built to verify the 3D imaging algorithm,which shows that the proposed algorithm can effectively detect the targets in both far-field and near field.Moreover,the imaging experiment is also carried out in a high resolution and wide-angle 2D imaging sonar system,which demonstrates the effectiveness of the proposed algorithm.The results of this study on thinned array and near-field imaging algorithm can effectively reduce the amount of transducers and widen the near-field viewing angle,which shows the feasibility of this research work in practical application.