NC code預測加工過程能耗

Abstract

本研究探討了CNC銑床工具機加工過程中的加工時間和能耗預測，並提出了相應的預測模型。研究通過解析NC code，建立了能有效預測加工時間和能耗的數學算法和基於XGBoost的能耗預測模型。通過時間預測模型的結果顯示，在不考慮更換刀把的情況下，考慮加速度與急跳速度對加工時間的預測具有重要意義。在不考慮加速度與急跳速度的情況下，理論計算時間與實際時間的誤差為-1.28%；只考慮加速度時，誤差縮小至-1.08%；而同時考慮加速度與急跳速度時，理論計算時間與實際時間的誤差僅為0.08%。故結果表明考慮機台的加速度和急跳速度可以顯著提高加工時間預測的準確性。能耗預測模型在大多數情況下具有較高的準確性，誤差均在1%以內，但在虎科本校區因輸入特徵較少導致誤差約為3%。總結來說，本研究強調了NC程式碼解析在提升時間和能耗預測準確性方面的價值。未來，本研究可與大型語言模型（LLM）結合，擴大應用範圍，包括檢查NC code編成錯誤、預估加工時間以及閱讀2D/3D圖檔進行自動報價系統評估。這些應用將進一步提升CNC銑床工具機加工過程的效率和準確性，並促進生產管理和能耗管理的改進。

This study investigates the prediction of machining time and energy consumption in CNC machine tool processes and proposes corresponding prediction models. By analyzing NC code, mathematical algorithms and an XGBoost-based energy consumption prediction model were established to effectively predict machining time and energy consumption. The results of the time prediction model show that considering acceleration and jerk is significant for predicting machining time. Without considering acceleration and jerk, the error between theoretical and actual time is -1.28%; when only considering acceleration, the error is reduced to -1.08%; and when both acceleration and jerk are considered, the error between theoretical and actual time is only 0.08%. Therefore, the results indicate that considering the machine tool's acceleration and jerk can significantly improve the accuracy of machining time prediction. The energy consumption prediction model demonstrates high accuracy in most cases, with errors within 1%, but at the NFU Campus, the error is about 3% due to fewer input features. In summary, this study highlights the value of NC code analysis in improving the accuracy of time and energy consumption predictions. In the future, this research can be integrated with Large Language Models (LLM) to expand its application scope, including checking NC code for errors, estimating machining time, and evaluating automatic quotation systems from 2D/3D drawings. These applications will further enhance the efficiency and accuracy of CNC machine tool processes, promoting improvements in production management and energy consumption management.