Scholarlink

<p>This paper presents a neural network-based model for predicting Computer Numerical Control (CNC) machines’ trajectory generation and cycle times directly from NC part programs. In CNC machining, trajectory generation involves interpolating through tool path points to create the actual motion paths of the machine tool. However, these trajectory generation algorithms are hidden within the proprietary algorithms of commercial CNC systems. Traditional methods rely on simplified assumptions, often failing for complex toolpaths with sharp transitions and high-speed motions. The proposed neural network model learns complex trajectory generation and cornering algorithms of commercial CNC using a bidirectional long short-term memory (Bi-LSTM) architecture. The study identifies limitations of existing cycle time prediction methods, particularly in handling complex toolpaths with sharp transitions and high-speed motions. Experimental validation on a CNC machine demonstrates the model’s capability to predict the trajectory and 3-axis machining cycle time. By capturing CNC system behavior, this approach enables better process planning, leading to enhanced machining efficiency.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Neural network-based modeling of CNC machines trajectory generation

  • 期刊论文

  • 发表日期：

    2025年7月25日

查看全文
  • Mobin Abdar Esfahani,
  • Behnam Karimi,
  • Yusuf Altintas

摘要

This paper presents a neural network-based model for predicting Computer Numerical Control (CNC) machines’ trajectory generation and cycle times directly from NC part programs. In CNC machining, trajectory generation involves interpolating through tool path points to create the actual motion paths of the machine tool. However, these trajectory generation algorithms are hidden within the proprietary algorithms of commercial CNC systems. Traditional methods rely on simplified assumptions, often failing for complex toolpaths with sharp transitions and high-speed motions. The proposed neural network model learns complex trajectory generation and cornering algorithms of commercial CNC using a bidirectional long short-term memory (Bi-LSTM) architecture. The study identifies limitations of existing cycle time prediction methods, particularly in handling complex toolpaths with sharp transitions and high-speed motions. Experimental validation on a CNC machine demonstrates the model’s capability to predict the trajectory and 3-axis machining cycle time. By capturing CNC system behavior, this approach enables better process planning, leading to enhanced machining efficiency.