A knowledge-driven method of adaptively optimizing process parameters for energy efficient turning. (1st January 2019)

• Record Type: Journal Article
• Authors: Xiao, Qinge; Li, Congbo; Tang, Ying; Li, Lingling; Li, Li
• Is Part Of: Energy. Volume 166(2019)
• Page Start: 142
• Page End: 156
• Publication Date: 2019-01-01
• Subjects: Parameter optimization -- Turning process -- Energy efficiency -- Knowledge-driven method; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2018.09.191
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: Selection of optimum process parameters is often regarded as an effective strategy for improving energy efficiency during computer numerical control (CNC) turning. Previous optimization methods are typically developed for specific machining configurations. To generalize the energy-aware parametric optimization for multiple machining configurations, we propose a two-stage knowledge-driven method by integrating data mining (DM) techniques and fuzzy logic theory. In the first stage, a modified association rule mining algorithm is developed to discover empirical knowledge, based on which a fuzzy inference engine is established to achieve preliminary optimization. In the second stage, with the knowledge obtained by investigating the effects of parameters on specific energy consumption covering a variety of configurations, an iterative fine-tuning is carried out to realize Pareto-optimization of turning parameters for minimizing specific energy consumption and processing time. The simulation results show that the method has a high potential for enhancing energy efficiency and time efficiency in turning system. Furthermore, compared with three heuristic optimization techniques, i.e. Genetic Algorithm, Ant Colony Algorithm and Particle Swarm Algorithm, the proposed method demonstrates certain superiority. Highlights: A knowledge-driven optimization method is proposed for energy efficient turning. Power consumption is predicted using a hybrid cluster-regression approach. The parameter optimization method can be performed under multiple machining configurations. Optimization is achieved more sustainably when embodied energy is considered. The extracted knowledge improves the efficiency of the optimization method.