基于理想模型和遗传算法的模拟电路自动化设计研究

Research of automated design of analog circuit based on idealized model and genetic algorithm

受模拟电路人工设计启发,以 MOSFET 电路为例提出一种模拟电路的自动化设计方法。首先以 MOSFET 的理想模型为基础,利用遗传算法(GA)产生电路拓扑并优化其参数;然后用实际元件替换其理想模型,通过少量调整即可得到最终电路。GA 在电路拓扑生成和参数优化方面具有优势,理想模型可有效缩小算法的搜索空间,因而所提方法在最优电路拓扑生成和加快电路设计速度两方面具有更为明显的优势。通过对三次方运算电路的设计,证实了所提方法的有效性。

Inspired from manual design of analog circuits, this paper proposed an approach of automated design for analog circuits which was introduced using MOSFET circuits. Firstly, the system synthesizes circuits using idealized elements by a genetic algorithm （GA）, and then idealized elements in a generated circuits are replaced by MOSFETs. GA takes the advantages of evolving circuit topologies and transconductances of idealized elements to achieve the given specifications, and the use of idealized elements effectively reduces search space and makes the synthesis efficient. Experiments of synthesis of a cubing circuit is demonstrates that this methed is effective, can rapidly generate the result circuit, and is expected to be applied in practice.