Joint Maximum Likelihood Estimation With Angle-frequency Estimation Based On An Improved Intelligent Algorithm

With the development of boda direction(DOA)estimation in the field of array signal processing,a large number of models and methods with superior estimation performance have been applied to it,but most of the traditional estimation processes only focus on the estimation performance of the signal orientation and ignore the effect on the estimation when the orientation is combined with the frequency,so in order to make up for the vector hydrophone signal orientation and frequency combined great likelihood estimation(ML)in hydroacoustic signal processing This paper proposes a method based on the joint azimuth and frequency estimation.In this paper,a method based on the joint azimuth-frequency maximum likelihood is proposed for the estimation of array signals.In addition,to address the shortcomings of simultaneous multidimensional nonlinear search in the traditional joint great likelihood signal orientation-frequency estimation,such as large computational effort and slow speed of traditional grid search,this paper is based on the Tuna Optimisation(TSO)algorithm and the modified Manta Ray Foraging Optimisation(MSMRFO)algorithm,respectively,to perform joint signal orientation-frequency estimation of hydroacoustic signals on the basis of great likelihood estimation to achieve better signal The main research work is as follows:(1)To address the multidimensional non-linearity of the spectral function in the great likelihood estimation,this paper combines the TSO with the joint signal orientation and frequency great likelihood estimation method using its high convergence accuracy in solving the extrema.The simulation results show that the algorithm has accurate angle and frequency estimates for a single source,and has faster convergence and lower root mean square relative error when compared with simulations based on the squirrel search algorithm,the sine and cosine algorithm,the atomic search optimization algorithm and the artificial bee colony algorithm for different numbers of sources。(2)In response to the traditional manta ray optimization algorithm’s low accuracy and its tendency to fall into local optimality,this paper proposes a multi-strategy manta ray optimization algorithm based on five update strategies.The improved algorithm is also verified with the CEC2017 test function package for polymorphic functions.The experimental results show that the improved algorithm not only has a closer function optimum and more stable performance,but also performs well in terms of convergence speed and convergence accuracy.In addition,performance simulations comparing MSMRFO with seven other algorithms for joint azimuth-frequency maximum likelihood estimation show that the multi-strategy manta ray algorithm has better convergence performance,and that the method has faster convergence and lower root mean square relative error at iteration,while maintaining estimation accuracy.The computational complexity of the multidimensional nonlinear problems associated with the use of joint maximum likelihood estimation is greatly reduced.In summary,this paper proposes several improved intelligent algorithms for the joint estimation of azimuthal frequency in hydroacoustic array signal processing.Combined with the performance of simulation experiments,the proposed algorithms have good performance in experimental indicators such as estimation accuracy,convergence speed and relative error,which indicates that the series of studies in this paper have certain significance for hydroacoustic array signal processing.