摘要
为了对等离子喷焊工艺参数进行优化,提高喷焊层的质量,通过径向基函数(radical basis function,RBF)神经网络近似模型和非支配排序遗传算法(non dominated sorting genetic algorithmⅡ,NSGA-Ⅱ)遗传算法相结合的方法,对等离子喷焊试验数据,基于MATLAB平台进行训练,以此来构建显微硬度、磨损量和稀释率的近似模型,利用NSGA-Ⅱ遗传算法对模型进行下一步的多目标优化,最终得到帕累托最优解集,研究了工艺参数间的交互作用。结果表明:利用RBF-NSGA-Ⅱ遗传算法比响应面法能更显著地提高喷焊层质量。可见对等离子喷焊工艺的优化具有一定的参考价值。
To optimize the processing parameters of plasma spray welding and improve the quality of the spray welding layer,by combining radical basis function(RBF)neural network approximation model with non-dominated sorting genetic algorithmⅡ(NSGA-Ⅱ)genetic algorithm,the plasma spray welding test data were trained based on MATLAB platform,on which the approximate model for the microhardness,wear amount,and dilution rate was constructed.The multi-objective optimization of the model in the next step was carried out by using NSGA-Ⅱgenetic algorithm.Finally,the Pareto optimal solution set was obtained,and the interaction among the process parameters was studied.Experimental results show that the RBF-NSGA-Ⅱgenetic algorithm can improve the quality of spray welding layer more significantly than the response surface method.This study has a certain reference value for the optimization of plasma spray welding process.
作者
刘永姜
李俊杰
曹一明
曾艾婧
LIU Yong-jiang;LI Jun-jie;CAO Yi-ming;ZENG Ai-jing(Shanxi Key Laboratory of Advanced Manufacturing Technology,North University of China,Taiyuan 030051,China)
出处
《科学技术与工程》
北大核心
2021年第11期4403-4408,共6页
Science Technology and Engineering
基金
先进制造技术山西省重点实验室开放基金(XJZZ201806)。
关键词
多目标优化
等离子喷焊
神经网络
遗传算法
响应面法
multi-objective optimization
plasma spray welding
neural network
genetic algorithm
response surface method
