基于响应曲面法与遗传算法的快速冷热成形工艺优化研究

Efficient Process Parameter Optimization of RHCM Technology Using Response Surface Methodology and Genetic Algorithm

综合Taguchi理论、响应曲面法、遗传算法及数据处理技术,对快速冷热成形工艺进行了优化研究。研究确定了对快速冷热成形制品品质影响显著的工艺因素,采用响应曲面法建立了制品品质预测模型,首度引入均值基的概念,实现了对实验数据的量纲一及规整化处理,在确保评价模型高适配性的同时,保证了对问题描述的客观不变性,进而采用改进遗传算法实现了对预测模型的求解。以LCD电视机前面壳为例,验证了该优化技术的可行性。

By integrating Taguchi method, response surface methodology （RSM）, genetic algorithm （GA） and data processing technology, an efficient optimization of process parameters in RHCM was investigated. The main process parameters which influenced on high-gloss surface part quality efficiently were determined firstly. Then a predictive response surface model for comprehensive quality index of part was created using RSM. In order to eliminate the abnormal matrix during the regressing analysis stage, and ensure the high fitness of estimated model as well, the conception of average matrix base was introduced for the first time during the experimental data processing phase. The re- sponse surface （RS） model was interfaced with an effective developed GA to find the optimum process parameter values. And a front panel of LCD television was considered as a high-gloss part example which was manufactured by RHCM technology. The case results show that it is feasible and effective.