A Study on Hybridization of Particle Swarm and Tabu Search Algorithms for unconstrained Optimization and Estimation Problems

This paper presents a short study on the hybridization of a swarm based optimization algorithm with a single agent based algorithm. Swarm based algorithms and single agent based algorithms have each distinct advantages and disadvantages. One goal of the presented work is to combine the concepts of the two different algorithms such that a more effective optimization routine results. In particular, we used a Particle Swarm (PS) based optimization algorithm as basis and induce Tabu Search (TS) based operators. The developed hybrid algorithm is tailored such that it has the capability to adapt to the given cost function during the optimization process. The proposed algorithm is tested on a set of different benchmark problems. In addition, the hybrid algorithm is utilized for solving the estimation problem encountered when colored noise is present and the Least Squares (LS) algorithm has bias problems. Key-Words: Particle Swarm, Optimization, Tabu Search, Parameter Estimation, Colored Noise