Design of optimal IIR digital filter using Teaching-Learning based optimization technique

In this paper an Enhanced Teaching-Learning Based Optimization (ETLBO) algorithm is employed to design stable digital infinite impulse response (IIR) filter using Lp-norm error criterion. The original TLBO algorithm has been remodeled by merging the concept of opposition-based learning and migration for selection of good candidates and to maintain the diversity, respectively. The multiobjective IIR digital filter design problem considers minimizing the Lp-norm approximation error and minimizing the ripple magnitude simultaneously while satisfying stability constraints on the coefficients of the filter. Weighted sum method and p-norm method are applied to solve the multicriterion optimization problem . Best weight pattern is searched using evolutionary search method that minimizes the performance criteria simultaneously. The validity of the method is demonstrated for the design of low pass (LP), high pass (HP), band pass (BP) and band stop (BS) IIR filters. The comparison of simulation results with other existing methods show that the proposed ETLBO algorithm is superior in terms smaller L1-norm error, L2 -norm error and smaller pass band and stop band ripples. Key-Words: IIR filter, TLBO, magnitude response, stability, Lp-approximation error.