Abstract:Warpage of thin plastic products is influenced by injection processing parameters severely such as a sun visor studied in the paper. It can be minimized by optimizing processing parameters. First of all, a certain number of experiment programs were derived from center composite experiment, then they were simulated by Moldflow with writing down the size of warpage, and then multiply regression model was built between processing and objection by response surface methodology, which made use of the theory of the least square method. After verifying the valid of the model by variance analysis, the best processing parameters combination could be gained by utilizing genetic algorithm to the prediction model with penalty terms.