基于改进支持向量机的选区激光熔化参数优化的研究

Research on Selective Laser Melting Parameter Optimization Based on Improved Support Vector Machine

在选区激光熔化成型过程中,通过实验法和试凑法确定合适的参数组合,需要具备一定的实验设备以及人力物力和时间的投入,周期长、成本高。以316L不锈钢选区激光熔化过程的参数选择为例,采用粒子群算法和网格搜索法对支持向量机参数进行优化,建立其选区激光熔化过程的质量参数模型,对加工过程的熔池尺寸进行预测。结果表明,熔池的宽度、高度、深度的预测值与实际值的误差均控制在6%以内,可为选区激光熔化生产的参数选择提供依据。

In the process of SLM, the determination of the appropriate parameter combination by the experimental method and the trial method requires a certain amount of experimental equipment, human,material resources and time input,which has a long cycle and high cost. Taking the parameters selection in selection laser melting process of 316 L stainless as an example, the particle swarm optimization and grid search were used to optimize the parameters of support vector machines.The quality parameters model of selection laser melting process of 316 L stainless steel was established, and the size of molten pool in the processing was predicted. The results show that the error of the model prediction for width, height and depthis is within 6% compared with that of actual values, which can provide references for the parameters selection for laser melting production.