Color Injection Molding Process Parameters Neural Network-based Particle Swarm Optimization

As the first production process in the seventies of the 20th century, two-color injection molding technology has been more widely used in the foreign countries. As the high-price forming equipment and the complex injection molding process, Multi-component injection molding process has not been widely introduced and applied before. the research and application of two-color injection molding are in an early stage. The research of two-color injection molding forming process and quality control are still poor. This paper conducts a FEM molding for two-color injection, and uses particle swarm neural networks to find optimal forming process.The main work and achievements are as follows:Through the principle of the two-color injection molding, it establishes a reliable finite element model, uses an advanced Moldflow software to simulation. Through the analysis of the simulation results, the effect is good.The paper selects the warpage of the Injection molding as quality indicators, analyzes to the process parameters and the impact of quality. This paper selects ten factors and three levels orthogonal to analyze all parameters significantly. It selects five parameters which have significant effect to the warpage.It proposes the warping prediction models basing on the Bp neural network, verifies the accuracy of the algorithm though the examples. It uses the first melt temperature, mold temperature, melts injection speeding, packing pressure, packing time as the target, establishes a nonlinear maps from the In molding production parameters to the warping deform.It achieves the optimization of process parameters basing on PSO, makes less warping deformation though using PSO and doing optimization of process parameters using BP neural network prediction model as fitness function.