View in EDS

Integrated multi-task learning and ensemble models for energy consumption prediction and optimization in precision milling spindles.

Saved in:
Bibliographic Details
Title: Integrated multi-task learning and ensemble models for energy consumption prediction and optimization in precision milling spindles.
Authors: Shi, Chenyang1 (AUTHOR), Xie, Yang1 (AUTHOR) xiey_just@just.edu.cn, Wu, Jianzhao2 (AUTHOR), Zhang, Chaoyong3 (AUTHOR)
Source: International Journal of Advanced Manufacturing Technology. Jun2026, Vol. 144 Issue 9/10, p7171-7187. 17p.
Subjects: Energy consumption forecasting, Spindles (Machine tools), Genetic algorithms, Multi-objective optimization, Ensemble learning, Particle swarm optimization, TOPSIS method
Abstract: The complex interplay and inherent trade-off between energy consumption and efficiency in precision milling processes have emerged as critical technical bottlenecks impeding sustainable development. Traditional single-task prediction models are found to struggle with ensuring accurate energy consumption forecasts and satisfying process quality constraints, while lacking support for generalization and efficient optimization. Accordingly, a precision-milling spindle energy consumption prediction model integrating multi-task learning (MTL) with the XGBoost ensemble algorithm is proposed. First, a precision-milling energy consumption prediction model providing a 95% confidence interval was constructed. By incorporating cross-task feature sharing within the MTL framework, high-precision predictions of spindle energy consumption, machine noise, and material removal rate (MRR) were achieved. Second, a multi-objective optimization model combining a genetic algorithm (GA) and particle swarm optimization (PSO) was developed, effectively addressing traditional algorithms' susceptibility to local optima and slow convergence. Furthermore, the entropy-weighted TOPSIS method was applied for the comprehensive evaluation and ranking of the Pareto solution set, enabling the identification of optimal cutting parameter combinations that satisfy process requirements and guide milling operations. As a result, a 14.01% reduction in spindle energy consumption, a 14.37% increase in machining efficiency, and a 15.21% reduction in machining noise were achieved. These research outcomes can foster theoretical innovation and facilitate engineering applications in advanced energy efficiency management for digital manufacturing equipment. [ABSTRACT FROM AUTHOR]
Copyright of International Journal of Advanced Manufacturing Technology is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:The complex interplay and inherent trade-off between energy consumption and efficiency in precision milling processes have emerged as critical technical bottlenecks impeding sustainable development. Traditional single-task prediction models are found to struggle with ensuring accurate energy consumption forecasts and satisfying process quality constraints, while lacking support for generalization and efficient optimization. Accordingly, a precision-milling spindle energy consumption prediction model integrating multi-task learning (MTL) with the XGBoost ensemble algorithm is proposed. First, a precision-milling energy consumption prediction model providing a 95% confidence interval was constructed. By incorporating cross-task feature sharing within the MTL framework, high-precision predictions of spindle energy consumption, machine noise, and material removal rate (MRR) were achieved. Second, a multi-objective optimization model combining a genetic algorithm (GA) and particle swarm optimization (PSO) was developed, effectively addressing traditional algorithms' susceptibility to local optima and slow convergence. Furthermore, the entropy-weighted TOPSIS method was applied for the comprehensive evaluation and ranking of the Pareto solution set, enabling the identification of optimal cutting parameter combinations that satisfy process requirements and guide milling operations. As a result, a 14.01% reduction in spindle energy consumption, a 14.37% increase in machining efficiency, and a 15.21% reduction in machining noise were achieved. These research outcomes can foster theoretical innovation and facilitate engineering applications in advanced energy efficiency management for digital manufacturing equipment. [ABSTRACT FROM AUTHOR]
ISSN:02683768
DOI:10.1007/s00170-026-18267-3