A Diameter Of Cavitation Optimization Design Based On Neural Network

The submarine supercavitation vehicle is a kind of weapon developed by the theory of cavitation. It is focused on by the military nations soon after it is developed. This thesis takes the cavitator diameter of the supercavitation vehicle as research subject. At first,we improves the classic optimization method of engineering structures, a.e. the penalty function method, to make is more easier to compute and simplify. Then we establish the penalty target function with two inequality constraints, and improve the penalty factor to make it unified, which prepares for the merge of the neural network. This thesis merges the improved penalty target function with two inequality constraints into the circuit system analogized by Hopfield neural network, and makes the penalty target function and the energy function of neural network corresponded. By the work we have done, wo convert the problem of classic optimization into the problem of solving the minimum value of the energy function of neural network, which is called Hopfield neural network optimization method.This thesis used neural network method for the cavitator diameter of Supercavit motion body on reasonably optimize the design, so that its reasonable in diameter scope, makes the Supercavit weapons fastest. Firstly established mathematical model of the optimized object, with traditional optimization method. Then used the Hopfield neural network optimization method for modeling, using computer simulation technology to solve, the traditional optimization method and neural network optimization method for comparative analysis, proved the feasibility of neural network optimization method,proved the feasibility of neural network optimization method.