Improved Particle Swarm Optimization And Its Application In Sonar Beam Pattern Optimization

With the proposal of the strategy of maritime power,China has accelerated its exploration of the ocean.Sonar is widely used as the main equipment for underwater information acquisition.With the continuous deepening of ocean exploration and the continuous improvement of actual engineering requirements,the performance of sonar is put forward higher.Beamforming as a Key Technology of Sonar,It directly affects the performance of sonar system.Therefore,it is very important to develop more efficient beamforming technology.The application of conventional beamforming technology is generally based on certain specific conditions.With the increasing complexity of actual engineering requirements on beam pattern,conventional beamforming methods can no longer be fully applicable,and it is urgent to develop beamforming technology that meets the purpose of more projects.As one of the intelligent optimization algorithms,particle swarm optimization(PSO)has been widely applied to various optimization problems,and also to the beam pattern optimization of sonar,showing great potential advantages.However,when particle swarm optimization(PSO)is used to solve the beam pattern optimization problem,premature convergence will easily fall into the problem of local optimal solution.In view of this,this paper proposes an improved particle swarm optimization algorithm based on the sonar beam pattern optimization problem.The main research work is as follows:(1)In order to maintain the diversity of the swarm,genetic operators are introduced into the particle swarm optimization algorithm.After analyzing the shortcomings of roulette selection operator,a selection operator model based on sorting and normal distribution is proposed.In order to improve the efficiency of the algorithm,the two-stage strategy is adopted in the process of iterative optimization.At the beginning of the iteration,the dynamic multi-swarm algorithm is adopted,and the information exchange among the sub-swarms is forbidden.At the later stage of the iteration,the sub-swarms are merged into one for iterative optimization.The simulation experiments on six standard test functions show that the proposed algorithm has a great improvement in optimization ability and stability compared with the related work in the existing literature.(2)The proposed algorithm and dynamic multi-swarm algorithm are applied to beam pattern optimization problem,and the results obtained by the two algorithms are compared.The results show that the proposed algorithm has better optimization ability and convergence speed.At the same time,it shows that the proposed algorithm can effectively solve the beam pattern optimization problem.Aiming at the specific problem of sonar beam pattern optimization,a new model of beam pattern optimization algorithm is proposed based on the algorithm in this paper.For different sonar beam pattern optimization problems,the model uses different swarm initialization strategies to improve the convergence speed of the algorithm.Finally,the model is instantiated according to the given beam pattern optimization problem,then two new algorithms are proposed and applied to the beam pattern optimization problem.The results are analyzed to prove the feasibility and efficiency of the model.