An Auto-Adjusting Hybrid Quantum Genetic Algorithm-Spectre platform for the multi-objective optimization of analog circuit sizing

Among the phases constituting analog circuit design, circuit sizing is considered labor-intensive, formidable, and heavily experience-dependent due to its non-linearity. Asa result, design automation coupled with effective optimization techniques has arisen as a feasible candidate to address challenges with circuit design and satisfy the increasing need for high-performance circuits. Among evolutionary algorithms, the combination of the genetic algorithm (GA) and quantum computing techniques has yielded the hybrid quantum genetic algorithm (HQGA) which has proven to be an effective optimization method in many fields due to its convergence rate and near-optimal solutions. This paper introduces an upgraded version of HQGA we call the Auto-adjusting Hybrid Quantum Genetic Algorithm (AHQGA) which avoids premature convergence and improves convergence speed through the use of an additional best-fitness-based scheme for rotation angles. In particular, this work proposes the utility of AHQGA for the multi-objective optimization of analog circuit sizing, with the two- stage Miller-compensated operational amplifier (op-amp) used as a topological case study. Additionally, for an objective evaluation, optimization results by AHQGA are compared with those by HQGA with fixed rotation angles and classical GA.