切削加工の自律型作業設計 : 切削理論アルゴリズム・ニューラルネットワーク統合型の適応予測

概要

論文の詳細を見る
Small sized batch jobs with large product diversification require intelligent machine tools to optimize cutting conditions by adapting operation planning to the machining environments, e.g., the combinations of cutting tool, work material, and machine tool. In this paper, autonomous operation planning, which optimizes machining operations for each machine tool using adaptive prediction of machining processes, is proposed. In adaptive prediction, the machining processes are predicted and optimized more accurately with adapting parameters of governing equations used for analytical prediction, and/or parameters of neural networks through learning of machining results. Adaptive prediction enables us to predict dimensional accuracy with machining point, and to predict surface roughness with machining time. In an operation planning considering tool wear and surface roughness, machining operation is optimized to minimize machining cost by flank wear predicted with analytical approach based on metal cutting theory and by surface roughness predicted with neural network.
1993-10-05