Abstract:Orthogonal design and least squares wavelet support vector machine are integrated to optimize the technological parameters of hydro-mechanical deep drawing process using FEM. The FEM model of hydro-mechanical deep drawing with radial pressure is built by directly replacing hydraulic pressure with node force and element force. Under the orthogonal design method, four processing parameters and correspond three levels of each are studied, which gives importance sequence of four parameters. Learning from orthogonal experiments and several FEM simulations according to single factor alternative, least square wavelet support vector machine (LS-WSVM) can predict the experimental result of given parameters combination and optimized technological parameters. The combination of orthogonal design and LS-WSVM decreases simulation work and provides an effective way for parameters optimization.

Parameters optimization of hydro-mechanical deep drawing with orthogonal design and support vector machine