Design And Implementation Of Fast Sonar Range Prediction Software In Small Memory System

As an important monitoring equipment in Marine environment,sonar plays an important role in various fields.In the process of detecting and identifying underwater targets,the speed of sonar action range prediction is an important reference index to evaluate the performance.In the past,many researchers have used parallel means to improve the prediction speed,using Open MP,but this method has no obvious advantages in the case of clusters.MPI is used for parallel algorithm programming,which is suitable for clusters,but the programming structure of the model is complex and the program is unstable.The function modules of the traditional acoustic model Bellhop have been packaged,which is not conducive to parameter adjustment in actual work,and most sonar detection equipment or upper computer does not support Matlab and other software running,while C++ is more convenient for developers to modify its internal algorithm,and simple operation,high compatibility,can achieve better functions on simpler hardware equipment.Therefore,based on the ray model and the normal wave model,and using C++,this paper designs a fast sonar range prediction model software which can quickly predict the sonar range and analyze the sound field signal on the stable hardware platform with small memory.The ray-acoustic theory has the advantages of fast computation speed and wide application range.Considering the practical application range,the ray-acoustic theory is chosen as the main application method of the model in this paper.Firstly,the Gaussian ray-tracing method is used to track the propagation trajectory of acoustic lines at a certain open Angle,and the given sea depth and furthest distance are divided into a depth-distance grid.Gaussian beam method is used to calculate the influence of stepping points on the sound pressure and propagation loss of grid points.In the scattering model,the small slope approximation theory is used to calculate the single-base reverb and bibase reverb,including seabed parameters and interface rough fluctuation.In order to realize the real-time comparison under the same parameter condition,the propagation loss is calculated by using the normal wave theory.By using the acoustic line arrival structure,the propagation loss in frequency domain and the reverberation in time domain are unified,and the operating distance of sonar is calculated.In this paper,the running time of each module is statistically compared,and it is found that the ray tracing module takes the most time.Considering the optimization of the number of acoustic lines,the running time can be reduced by appropriately reducing the number of acoustic lines.In the process of reducing the number of sound lines,the accuracy of prediction results will inevitably be reduced.Therefore,by comparing and tracking the prediction results of different sound lines,the number of sound lines can be reduced while the error range is appropriate.Finally,360 sound lines are selected for track tracking.In order to increase the operation speed as much as possible,this paper also optimizes the program structure.Finally,the fast sonar range prediction model based on C++ language designed in this paper can run stably on the Window system and Vx Works system.The prediction time is less than 1 minute under 50 sound source depth,and the operating memory is less than 100 MB.