The DOA Estimation Of Vector Hydrophone Based On Two Improved Intelligent Algorithms

The vector hydrophone has taken a very important step for underwater signal processing and improved many performances of the hydrophone.Because of its good performance,it is widely used in underwater information transmission,target detection,target low attitude,tracking and other aspects.In this paper,two kinds of improved intelligent algorithms are used to optimize BP neural network and maximum likelihood estimation respectively to study direction of arrival estimation(DOA)of vector hydrophone1.BP neural network is very sensitive to the selection of initial parameters.If the selection of parameters is improper,BP neural network tends to fall into local minimum and convergence speed is slow.Due to the characteristics of grey wolf optimizer(GWO),the diversity of grey wolf population can not be maintained,and it is easy to fall into local optimum.However,the differential evolution algorithm(DE)has a strong global search ability,simple principle,few controlled parameters and easy implementation.Therefore,the mutation and selection operations of the difference algorithm are introduced into GWO algorithm,which is conducive to maintaining the diversity of the gray wolf population,and the cross selection operation is used to update the individual position of the gray wolf.In this way,the differential gray wolf algorithm,which combines the differential evolution algorithm and the gray wolf algorithm,is obtained.Using the parameters of BP neural network optimized by DEGWO,a DEGWO-BP model is established.The direction of arrival of vector hydrophone array signal is studied,and the error estimates under different SNR are obtained.The simulation results show that the DEGWO-BP model is superior to BP,GWO-BP,PSO-BP and SAPSO-BP models,which not only has better estimation accuracy,better convergence speed and optimization performance.And the DEGWO-BP istested with the data of Fenhe Second Reservoir,and DOA estimation is realized effectively.The results show that the proposed DEGWO-BP model has better general applicability and good application prospects.2.Combining chaos theory with salp swarm algorithm(SSA),an improved CSSA algorithm is proposed,and CSSA is optimized to estimate DOA of vector hydrophone.In this paper,CSSA,SSA and particle swarm optimization(PSO)algorithm respectively optimize the maximum likelihood estimation algorithm DOA estimation in the case of single signal source,double signal source and different low SNR.The simulation results show that the CSSA algorithm proposed in this paper can optimize the parameters of maximum likelihood estimation algorithm to achieve DOA estimation.The mixed algorithm of CSSA algorithm optimization maximum likelihood estimation algorithm is applied to the actual data of Fenhe Second Reservoir for DOA estimation.The results show that the improved algorithm proposed in this paper has better estimation accuracy.Again,it demonstrates that the improved algorithm has a strong guiding significance for practical application.The two methods proposed in this paper,both of which have been processed by computer simulation experiment and actual data,have some advantages over the previous methods,and provide a new idea for the estimation of the direction of arrival of vector hydrophone,hoping to inspire the researchers of the array signal processing and the direction of arrival estimation.