The 30-mm cannon's barrel was thin and long. To limit the deformation of the muzzle, the support was designed. The optimization became in need to satisfy the request of the dynamics and mobility.This paper analyzed the 30-mm cannon's barrel fixed with the support from static and dynamics' viewpoint for the information of the stiffness, the frequencies and the response. With finite element analysis program ANSYS, based on the theory of orthogonal design, 16 samples was designed to train the Radial Basis Function Neural Network. With some precision, the well-trained network succeeded in taking place of the ANSYS program to calculate the mechanics character of different structures. Based on the technique of niche, with the optimum theory of Pareto, and by the software of MATLAB, the multi-objective optimization program of Genetic Algorithms was developed to search for the optimum design, which was low in weight and high in frequencies. Verified by the calculation with ANSYS, the structure optimization proved reliable. |