Adaptive optimization of cutting parameters in milling industry considering dynamic tool wear in intelligent manufacturing driven by reinforcement learning

In industrial production, tool wear affects the process of intelligent manufacturing, so real-time monitoring of tool wear is necessary. This article intended to adopt a new approach that combined digital twin modeling with particle filtering, integrating data-driven wear, and simulated wear to achieve accurate evaluation of tool wear. The workpiece material milled in this article is Tc4 (Ti–6Al–4 V). The tool is a hard alloy tool, and tool wear is monitored. Using the digital twin method, combined with data-driven wear values and simulation model-based wear values, the final tool wear monitoring values were obtained through particle wave fusion algorithm fusion. The wear rate before optimization of cutting parameters in this article was 5.6%, and after optimization, it was 4.1%. The adaptive optimization method for cutting parameters in this article can delay tool wear and help improve production efficiency in intelligent manufacturing.