Constraint multi-objective automated synthesis for CMOS operational amplifier

A multi-objective evolution algorithm (MOEA) is presented to automatically determine the parameters in Op-Amp synthesis where the cost functions (e.g., minimizing the power dissipation and the chip area) and the constraint functions (e.g., the user-defined specifications) can be modeled as polynomia...

Full description

Saved in:
Bibliographic Details
Published in:Neurocomputing (Amsterdam) Vol. 98; pp. 108 - 113
Main Authors: Tao, Jili, Fan, Qinru, Chen, Xiaoming, Zhu, Yong
Format: Journal Article
Language:English
Published: Elsevier B.V 03.12.2012
Subjects:
CMOS Op-Amp
Constraint handling
Elitist maintaining
Multi-objective evolution algorithm
Constraint handling
Multi-objective evolution algorithm
Elitist maintaining
CMOS Op-Amp
ISSN:0925-2312, 1872-8286
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A multi-objective evolution algorithm (MOEA) is presented to automatically determine the parameters in Op-Amp synthesis where the cost functions (e.g., minimizing the power dissipation and the chip area) and the constraint functions (e.g., the user-defined specifications) can be modeled as polynomials of the design variables. The proposed algorithm is based on MOEA which does not use weighting coefficients in converting multiple objectives into single objective. A constraint handling strategy without penalty parameters is proposed to avoid the difficulty of penalty parameter selection. Moreover, an elitist maintaining scheme is utilized to keep the evenness of the Pareto front. Simulations over several benchmark functions validate the efficiency of the proposed algorithm for the evenness of population distribution and the convergence to the Pareto front. Numerical experiments of a Miller compensated two-stage Op-Amp show that the proposed MOEA is able to achieve better performance than NSGA-II+PCH, GA+SPF and GA+PCH.
ISSN:0925-2312
1872-8286
DOI:10.1016/j.neucom.2011.09.042