Quotient space-based boundary condition for particle swarm optimization algorithm

Abstract

To control particles flying inside the limited search space and deal with the problem of slow search speed and premature convergence, this paper applies the theory of topology, and proposes a novel quotient space-based boundary condition by using the properties of quotient space and homeomorphism. In this new method, named QsaBC, Search space-zoomed factor and Attractor are introduced to enhance the performance of PSO according to analyzing the dynamic behavior and stability of particles. QsaBC proposed in this paper not only reduces the subjective interference and enforces the capability of global search, but also enhance the power of local search and escaping from an inferior local optimum. Four CEC'2008 benchmark functions are selected to evaluate the performance of QsaBC. Comparative experiments show that QsaBC is more effective to do with the boundary condition, furthermore, can get the satisfactory solution with fast convergence speed, easy calculation and good robustness than other experienced methods.