Machining process route planning: Deep reinforcement learning guided by graph-based historical process data

doi:10.1016/j.aei.2026.104533

Machining process route planning: Deep reinforcement learning guided by graph-based historical process data

2026-03-04

0

PRE

AI

DOI：10.1016/j.aei.2026.104533

Original

Original request for help

Save

Abstract

En 中文

• A knowledge-guided framework decouples knowledge extraction from DRL optimization. • A GCN extracts historical data to create a trusted and drastically reduced DRL search space. • A unique dual-mode state grounds the DRL policy learning in validated historical processes. • The resulting framework enables retraining-free process planning for unseen parts.

Keywords:

Deep reinforcement learning

Graph neural networks

Process planning

Knowledge-guided framework

Historical data mining

Machining process route planning: Deep reinforcement learning guided by graph-based historical process data

Abstract

Journal

Organization