Methodological Study On Displacement And Main Contant Of Submarine Based On GA-RBF Neural Network Arithmetic

In initial stage of submarine conceptual design,it is first necessary to do analysis and appraisal for the displacement and main contant of designing submarine.So as to get optimal displacement and main contant which benefit the general performance and tactics &technical indexes.However,at the beginning of the demonstration phase,often only know that the mission or task of designing submarine,many performance indexes and tactics&technical indexes are not clear,so it is of great difficulties to accurately get reasional displacement and main contant.The disquisition of"Methodological Study on Displacement and Main Contant of Submarine based on GA-RBF Neural Network Arithmetic"is to resolve this problem.The research is of great sigificance and value.This research includes two parts:the first part starts from functional modules classification of submarine,derives relations between weight attributes of systems& equipment and normal displacement,establishes standard calculation method of displacement and main contant on condition that weight of various parts of the submarine equals to displacement.The second part is to establish methodological study on displacement and main contant of submarine based on GA-RBF neural network arithmetic.As BP neural network is based on gradient decline,it is difficult to determine the best network structure and easy to get local best smallness which will cause false convergence. Therefore firstly in this paper,study on arithmetic of optimizing RBF neural network applying GA(Genetic Arithmetic) after deducind associated equation about displacement,data samples and GA-RBF neural network are applied to function approach.Numerical tables and calculation method of displacement and main contant based on GA-RBF neural network are established.This study provides a method to calculate displacement and main contant of conventionlly-powerec submarine for initial stage of conceptual design,calculations show that the two arithmetics have enough precission for practical project.