Application of QUBO solver using black-box optimization to structural design for resonance avoidance.

Area of Science:

• Engineering
• Computational Science
• Materials Science

Background:

• Quadratic Unconstrained Binary Optimization (QUBO) solvers offer potential for structural design optimization, including resonance avoidance.
• Current limitations in QUBO applications stem from the complexity of problem transformation.
• Black-box optimization (BBO) methods, utilizing machine learning and Bayesian approaches, have emerged to address these challenges in combinatorial optimization.

Purpose of the Study:

• To propose a novel black-box optimization (BBO) method for designing printed circuit boards (PCBs) to prevent resonance.
• To formulate the PCB design problem as maximizing natural frequency while minimizing mounting points.
• To overcome the challenge of approximating natural frequency for QUBO formulation within a BBO framework.

Main Methods:

• A factorization machine-based black-box optimization (BBO) method was developed for PCB resonance avoidance.
• The natural frequency, a key challenge for QUBO, was approximated using a quadratic model.
• Probabilistic neighbor generation and model updating were employed for efficient approximation around optimal solutions.

Main Results:

• The proposed BBO method successfully identified optimal mounting point positions for PCBs.
• The method demonstrated a shorter calculation time compared to conventional BBO techniques.
• A higher success probability in finding optimal solutions was achieved with the proposed approach.

Conclusions:

• The developed BBO method offers an effective solution for PCB resonance avoidance.
• This research enhances the applicability of QUBO solvers in structural design.
• The findings suggest broader potential for QUBO in various structural engineering applications.