基于粒子群支持向量回归的原位自生工艺参数优化

Optimization of process of in-situ technology based on SVR and PSO

综合应用激光熔覆和原位反应增强金属基复合材料,是当前金属基复合材料研究领域的一个热点,本文采用该工艺制备铁基表面复合材料,重点考虑该工艺参数的确定问题.根据在不同工艺参数下合成的铁基表面的WC体积分数实测数据集,提出建立不同工艺参数下WC体积分数的支持向量回归预测模型,并与基于人工神经网络模型(ANN)的预测结果进行比较.结果显示:对于相同的训练样本和检验样本,SVR预测模型比ANN预测模型具有更强的泛化能力.最后根据建立的预测模型,应用粒子群算法寻优得到最优工艺参数,该工艺参数在实际实验过程中的应用,验证了该方法的有效性.

The in-situ laser cladding technique is used to prepare reinforced iron matrix surface composites,and the issue of optimizing technique parameters is concerned in this study.According to the experimental dataset on the volume fraction of carbide under different process parameters,the support vector regression（SVR） approach is proposed to establish a model for simulation of the relationship between the volume fraction of carbide and process parameters.The prediction results demonstrate that the estimated errors of the SVR model are all less than those of the ANN model.It is also revealed that the generalization ability of SVR model surpasses that of artificial neural network（ANN） by applying identical training and test samples.The optimal process parameters are obtained by the particle swarm optimization（PSO） optimization of the SVR model,and the results further verified that the method in this paper is effective.