Homotopy-Inspired Cat Swarm Algorithm for Global Optimization
Shi Da Yang, Ya Lin Yi, Yan Ping Lu · Advanced materials research · 2012
Based on the concepts of homotopy, a novel cat swarm algorithm, called a homotopy-inspired cat swarm algorithm (HCSA),is proposed to deal with the problem of global optimization. Proceeding from dependent variables of optimized function,it traces a path from the solution of an easy problem to the solution of the given one by use of a homotopy--|a continuous transformation from the easy problem to the given one.This novel strategy enables the cat swarm algorithm (CSA) to improve the search efficiency. Theoretical analysis proves that HCSA converges to the global optimum. Experimenting with a wide range of benchmark functions, we show that the proposed new version of CSA, with the continuous transformation, performs better, or at least comparably, to classic CSA.