Analysis Of Ship Shock Response Spectrum Based On Pattern Recognition

With the great variety of modern weapons and the continuous improvement of the performance of weapon systems,warships,as the main offshore combat platform,are facing serious security problems.The research on the impact resistance of warships is the basis of ensuring the vitality of warships and giving full play to their combat effectiveness.The research on the impact environment of warships subjected to Underwater Non-contact Explosion is still one of the most important directions in the field of anti-impact of warships,which has excellent practical application value.There are great differences in the standards of ship impact environment in different countries,but the assessment of ship-borne equipment impact environment and the zoning of impact environment are stipulated in the form of shock response spectrum.Shock response spectrum is one of the most effective methods to describe the ship impact environment.It is influenced by computer technology and artificial intelligence technology.This paper focuses on the artificial intelligence pattern recognition technology under the underwater non-contact explosion based on the impact spectrum description.The application of equipment impact environment-related content is being studied.Because of the high cost of real ship test and the extensive use of pattern recognition theory and technology in the field of anti-shock of naval vessels,the computational load of numerical simulation is greatly reduced.At the same time,the impact environment of underwater explosion test equipment can be evaluated,which has great reference significance..The implementation of the content of this article is mainly in the following aspects:(1)Based on the test data of anti-impact assessment of shipborne equipment ofmedium floating impact platform,the support vector machine optimized by genetic algorithm is used to predict the impact environment at the installation location of equipment under the corresponding working conditions by determining the information of explosive quantity,detonation distance,location of explosive source,location of measuring point and quality of equipment.(2)The traditional method of regularizing the shock spectrum into the design spectrum has great errors in the calculation process.Therefore,a method of converting the shock spectrum and the design spectrum by using BP neural network is proposed.In this paper,the spectral velocity,spectral displacement and spectral acceleration parameters of BP neural network optimized by genetic algorithm are predicted more accurately.(3)There is no uniform standard for the classification of the impact environment of warships in various countries,while the current classification of the impact environment in China mainly refers to the naval powers dominated by the United States.Therefore,taking the impact environment of a certain type of warship in China as an example,this paper proposes a method to divide the impact environment of a warship based on pattern recognition technology.Firstly,the influence of the change of measuring points in longitudinal,transverse and vertical position on the impact environment of warships is evaluated by gray correlation method.Then the numerical statistical characteristics of the change of the impact environment are analyzed.On this basis,the clustering calculation of the impact environment of warships is realized by using SOM neural network,and the boundary of each model is repaired by using S4 VM semi-supervised clustering algorithm.At last,the result of ship impact environment zoning based on pattern recognition technology is obtained.The purpose of this method is to provide new ideas for the research on the impact environment of warships in China.