Hierarchical Optimization Algorithm for the Automatic Design of Analog Integrated Circuit

In this work, we have explored the use of hierarchical optimization technique to automatically design low dropout regulator. When we use genetic algorithm for multi-objective optimization, the increase of design variables will greatly increase the difficulty of optimization. Therefore, we divide the whole circuit into different levels and only upload the sub circuits’ combination of design parameters to reduce the design variables and shorten the optimization time. In our work, we have explored the optimization results under different constraints, and compared them with the direct optimization results systematically.