Operational Amplifier Design Automation Based on Multi-Objective Optimization

This paper proposes an operational amplifier (OpAmp) design automation approach based on multi-objective optimization. The design of the OpAmp is introduced as a multi-objective optimization problem and the non-dominated sorting genetic algorithm (NSGA-II) is adopted to solve this problem. Applying graph-pair decision diagram to get accurate symbolic design equations, and using proposed initial population generation method to fully automate the OpAmp design. Taking the CMOS Miller-compensated OpAmp as an example, the results show that proposed design automation approach is able to automate and accelerate the OpAmp design.