Title: Automated knowledge-based filter synthesis using modified Legendre approximation and optimisation of summed sensitivity

Authors: Vlastimir D. Pavlović; Maja M. Lutovac; Miroslav D. Lutovac

Addresses: Faculty of Electronic Engineering, University of Niš, Niš, Serbia. ' Lola Institute, Belgrade, Serbia. ' State University of Novi Pazar, Novi Pazar, Serbia

Abstract: The paper presents the automated design of active RC and programmable filters. The approximating function is derived using Legendre orthogonal polynomial with the appropriate weights and usage of two Legendre multiplication factors at the origin and the pass-band edge frequency. Detailed analysis is done for the frequency response and sensitivity analysis of active RC filters. Optimisation is performed using symbolic manipulation of expressions inputted into computer algebra system and numeric solvers. The code is in the form of template notebook, and the user should specify the minimal number of numeric values, such as the number of second-order filter sections, the pass-band edge frequency, the reflection coefficient, and the preferred values of component values.

Keywords: active filters; modified Legendre approximation; circuit structures; circuit functions; computer algebra; knowledge representation; knowledge-based filter synthesis; optimisation; frequency response; sensitivity analysis.

DOI: 10.1504/IJRIS.2012.046488

International Journal of Reasoning-based Intelligent Systems, 2012 Vol.4 No.1/2, pp.17 - 27

Published online: 10 Apr 2012 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article