An Improved PSO Algorithm Based on Chaos Theory and Used in Design Flood Hydrograph

Abstract: The deficiency of basic Particle Swarm Optimization（PSO） lie in the ubiquitous prematurity and the incapability in seeking the global optimal solution while optimizing complex high dimensional functions. To over come such deficiencies, COSPSO algorithm was established by introducing the chaos optimization mechanism and global particle stagnation-disturbed strategy. In the improved algorithm, the chaotic movement was adopted in particle’s initial movement locus to replace the former stochastic movement and chaos factor was used to lead particle’s flying. When the global particle stayed in, disturbing strategy was used to keep the particle from stagnation. Five benchmark optimizations were introduced to test COSPSO and proved COSPSO could remarkably improve the efficiency in optimizing complex functions. Finally, the application of the COSPSO in calculating the design flood hydrograph gave a good result.