DOA Estimation Of MEMS Vector Hydrophone Based On Improved Intelligent Optimization Algorithm

The development of vector hydrophones has brought convenience to human production and life,and has also added new power for people to further understand various complex and changeable underwater environments.Especially the MEMS vector hydrophone,which is currently in the research hotspot,not only inherits all the features of the previous vector hydrophone,but also has the advantages of small size and good consistency.Equipped with the product of this technology,it can provide a larger platform for its application in practical engineering,thereby promoting the research and improvement of MEMS vector hydroph--ones.This paper starts with the traditional DOA estimation under the intelligent optimization algorithm.Four merhods of DE,GWO,QGA,and SSA are chose and improved.And MUSIC algorithm and ML algorithm are used as the basis for joint optimization.The simulation results and actual effects in array signal processing are studied.The main research contents are as follows:(1)The four improved optimization algorithms are combined with the MUSIC algorithm,and the MUSIC spectrum function is used as the objective function of the optimization algorithm to find the extreme value.Then,the approximate interval is located by the peak value,and then search each angle step by step to finally reach the goal of DOA estimation.To a certain extent,the improvement of the algorithm in most SNR intervals is better than the MUSIC algorithm,with higher estimation accuracy and smaller root mean square error.Finally,the actual measurement experiments results of the second Fenhe Reservoir are all within a very small error.(2)Similarly,the four improved optimization algorithms are combined with the ML algorithm,and the ML rank absolute value function is used as the objective function of the optimization algorithm to directly seek extreme values to achieve the purpose ofsimultaneous multiple DOA estimation.Compared with the original ML algorithm,the complexity is greatly reduced,and the algorithm is improved with fewer iterations and higher efficiency.The same actual measurement experiments also performed well and achieved the expected results.The two direction DOA estimation models proposed in this paper have achieved a certain estimation effect in the actual measurement experiments.This is inseparable from the pre-processing of the data.It provides new ideas for the study of DOA estimation,and also broadens the practical application field of DOA estimation.