Multi-objectives design optimization based on multi-objectives Gaussian processes for System-in-Package

Dai, Weijing; Wang, Zhenkun; Xue, Ke

10.1109/ECTC32696.2021.00338

2021

IEEE 71st Electronic Components and Technology Conference (ECTC)

0569-5503

2377-5726

978-1-6654-3120-0

IEEE 71ST ELECTRONIC COMPONENTS AND TECHNOLOGY CONFERENCE (ECTC 2021)

2021-June

2151-2157

JUN 01-JUL 04, 2021

null,null,ELECTR NETWORK

10662 LOS VAQUEROS CIRCLE, PO BOX 3014, LOS ALAMITOS, CA 90720-1264 USA

IEEE COMPUTER SOC

Since electronic devices based on System-in-Package technologies become sophisticated, the number of components and the complexity of internal structure of individual devices increase dramatically. Therefore, how to optimally design such packages turns into one of the most crucial issues. However, this increasing complexity greatly enlarges not only the amount of design variables, but also the design objectives, making it extremely burdensome to examine every design neither by prototyping nor by numerical simulation. For the past, design of experiment, response surface analysis and other common statistical methods are widely used for such purpose, but their effectiveness drops rapidly with the intricacy increasing. To overcome this limitation, we implement multi-objective Gaussian process, Bayesian optimization and multi-objective multifactorial evolution algorithm to form a framework in this study to tackle this design problem with high-dimensional inputs and multiple outputs. Some variations of this framework were tested, and the modelling accuracy and optimization efficiency were used as the indicative merits to evaluate the performance of this framework. This interdisciplinary study shows benefits in introducing advance intelligent algorithms into packaging design process and present a viable approach to achieve both automation and optimization for complicated packaging design.

System-in-Packaging design optimization multi-objectives Bayesian optimization Gaussian processes

第一 ; 通讯

英语

CPCI-S ; EI

Engineering

Engineering, Electrical & Electronic

WOS:000702282700325

20220811668828

Design of experiments ; Evolutionary algorithms ; Gaussian noise (electronic) ; Multiobjective optimization ; Packaging ; Surface analysis

Packaging, General:694.1 ; Engineering Research:901.3 ; Optimization Techniques:921.5 ; Probability Theory:922.1 ; Mathematical Statistics:922.2 ; Materials Science:951

Web of Science

Southern Univ Sci & Technol, Sch Syst Design & Intelligent Mfg, Shenzhen, Peoples R China

Dai, Weijing,Wang, Zhenkun,Xue, Ke. Multi-objectives design optimization based on multi-objectives Gaussian processes for System-in-Package[C]. 10662 LOS VAQUEROS CIRCLE, PO BOX 3014, LOS ALAMITOS, CA 90720-1264 USA:IEEE COMPUTER SOC,2021:2151-2157.