ISBN: 978-981-18-5852-9 DOI: 10.18178/wcse.2022.04.120
Multi-objective Prediction Method for Hole-Drilling Quality of Multilayer PCB Based on MSVR
Abstract— In the high-speed drilling process of multilayer printed circuit boards, various drilling parameters
and drill wear directly affect the quality of the hole-drillings, which ultimately affects the performance and
yield of multilayer printed circuit boards. The hole quality of multi-layer printed circuit boards mainly
includes multiple indicators such as hole wall roughness, hole position accuracy and nail head thickness. This
study adopts a multi-output support vector regression model to establish a correlation model between drilling
parameters, drill wear and the hole quality of multilayer printed circuit boards, so as to realize the prediction
of the hole-drilling quality under the conditions of variable drilling parameters and different drill wear, and
the differential evolution algorithm is used to optimize the model parameters. The experimental data of
drilling multilayer printed circuit boards with variable drilling parameters show that: the adopted multioutput
support vector regression model has higher prediction accuracy and stronger robustness compared
with other commonly used multi-output regression models, which lays the foundation for the hole-drilling
quality control of multilayer printed circuit boards.
Index Terms—multilayer printed circuit boards drilling, drill wear, various drilling parameters, multi-output
support vector regression, hole quality prediction.
Qifeng Tan
Beijing Key Lab of Precision/Ultra-precision Manufacturing Equipments and Control, Department of Mechanical Engineering, Tsinghua University, China; State Key Laboratory of Tribology in Advanced Equipment, Tsinghua University, China
Guodong Liu
Beijing Key Lab of Precision/Ultra-precision Manufacturing Equipments and Control, Department of Mechanical Engineering, Tsinghua University, China; State Key Laboratory of Tribology in Advanced Equipment, Tsinghua University, China
Yong Li
Beijing Key Lab of Precision/Ultra-precision Manufacturing Equipments and Control, Department of Mechanical Engineering, Tsinghua University, China; State Key Laboratory of Tribology in Advanced Equipment, Tsinghua University, China
Hao Tong
Beijing Key Lab of Precision/Ultra-precision Manufacturing Equipments and Control, Department of Mechanical Engineering, Tsinghua University, China; State Key Laboratory of Tribology in Advanced Equipment, Tsinghua University, China
Cite: Qifeng Tan, Guodong Liu, Yong Li, Hao Tong, " Multi-objective Prediction Method for Hole-Drilling Quality of Multilayer PCB Based on MSVR, " WCSE 2022 Spring Event: 2022 9th International Conference on Industrial Engineering and Applications, pp. 1032-1040, Sanya, China, April 15-18, 2022.