Volume 48, No 2, 2026, Pages 237-250
Download full text in PDFOptimized Supervised Machining Learning-Based Energy Efficient Optimization of Roller Burnishing Process
Authors:
An-Le Van 
,
Minh-Thai Le ,
Truong-An Nguye ,
Thai-Chung Nguyen ,
Huu-Toan Bui ,
Huu-Phan Nguyen ,
Trung-Thanh Nguyen 
DOI: 10.24874/ti.2028.09.25.12
Received: 22 September 2025
Revised: 28 October 2025
Accepted: 15 December 2025
Published: 15 June 2026
Abstract:
In this work, the burnishing parameters, including the spindle speed (S), feed rate (f), depth of penetration (D), and number of rollers (N) are optimized to maximize the Vickers hardness (VH) and minimize energy consumption (EC) as well as average surface roughness (Ra). Predictive models of burnishing responses are proposed using the optimized Extreme Gradient Boosting (OXGBoost) approach. An efficient algorithm entitled Grasshopper Optimization Algorithm (GOA) is used to create optimal solutions. The entropy method and Pareto-Edgeworth Grierson (PEG) are utilized to calculate weights and select the best data. The findings presented that the optimal S, f, D, and N are 1075 rpm, 0.07 mm/z, 0.06 mm, and, respectively. At the optimal point, the VH is enhanced by 5.1%, while the EC and Ra are reduced by 5.4% and 23.3%, respectively. The EC model was significantly affected by the f, S, N, and D, respectively. The Ra model was significantly affected by the D, f, N, and S, respectively. The VH model was significantly affected by the D, N, f, and S, respectively. The OXGBoost-Entropy-GOA-PEG was a prominent solution to deal with complicated optimization issues, as compared to the conventional one. The outcomes can be applied to enhance energy efficiency and surface properties of the burnishing AISI 5140 process.
Keywords:
Roller burnishing, Energy consumed, Roughness, Hardness, Optimized XGBoost, Grasshopper optimization algorithm
