摘要
在新的机型投入生产之前,需要安排每个零件的工艺路线,如果该工作能实现自动化,就可以事先快速地给出一个基本正确的方案,为后续的并行协调提供方便。文章将介绍使用机器学习技术建立工艺路线分工模型的思路和方法,首先是任务性质的分析、数据清洗、特征工程、one-hot编码等准备工作,然后使用随机森林算法训练工艺分工预测模型,对模型的性能进行审查,以及模型学到的内容进行解释,最后对更先进的机器学习技术在工艺路线分工中的应用进行展望。
Before the new machine is put into production, it is necessary to arrange the process route of each part. If the work can be automated, a basic and correct scheme can be given quickly in advance, so as to provide convenience for the subsequent parallel coordination. This paper will introduce the ideas and methods of using machine learning technology to establish the model of process route division of labor. First of all, it is the preparatory work of task nature analysis, data cleaning, feature engineering, one-hot coding and so on. Then, the stochastic forest algorithm is used to train the process division prediction model, the performance of the model is reviewed, and the contents of the model are explained. Finally, the application of more advanced machine learning technology in the process route division of labor is prospected.
出处
《科技创新与应用》
2019年第21期91-92,共2页
Technology Innovation and Application