The DOA Estimation Of Vector Hydrophone Based On Neural Network And MUSIC Algorithm

The development of vector hydrophones provides a more convenient tool for people to study the underwater world.Compared with scalar hydrophones,it has more advantages.Its invention is to timely supplement the scalar hydrophone.Vector hydrophones are favored by their small size and strong receiving signal,and have gradually become an important tool for researchers to explore in many fields.In terms of signal processing,the direction of the wave has an important position in its field,and the research on it has never stopped.From the traditional MUSIC algorithm,ESPRIT algorithm and other classical methods to artificial neural network algorithms,the research on the direction of the wave can be varied.In this paper,we estimate the Direction of Arrival(DOA)of vector hydrophones,and study it from two aspects: neural network and MUSIC algorithm.The main contents are as follows:(1)Combining simulated annealing algorithm(SA)with particle swarm optimization(PSO),using the sudden jump action of SA algorithm to avoid PSO algorithm falling into local optimum,and optimizing the weight threshold of BP neural network.The extracted vector hydrophone signal is subjected to the process of covariance,real value,feature decomposition,and the base of the signal subspace is taken as the input of the neural network.The simulation results show that the improved SAPSO-BP model has better estimation performance in the direction of arrival than the traditional PSO-BP and BP models,and the estimation accuracy is higher and the error is smaller.(2)By constructing the genetic particle swarm optimization algorithm(GAPSO)by genetic algorithm(GA)and PSO algorithm,the PSO algorithm is avoided to fall into local optimum by using the crossover and mutation process in GA algorithm.Combined with the MUSIC algorithm to form the GAPSO-MUSIC algorithm,the spectral function of the MUSIC algorithm is used as the fitness function of the GAPSO algorithm to estimate theDOA.Through the simulation experiment and the experiment of the Fenhe Reservoir,the results show that GAPSO-MUSIC has better estimation effect and is more practical.This paper proposes two models,SAPSO-BP and GAPSO-MUSIC,which are used for the direction of arrival estimation of vector hydrophones.Through comparison,it is found that the model proposed in this paper has more advantages and higher precision,which provides a new idea for the research of DOA estimation and has certain reference value.