An Improved Particle Swarm Algorithm and Its Application in Low NOx Combustion Optimization of Coal-fired Utility Boiler

摘要

To prevent premature convergence frequently appeared in the particle swarm optimization (PSO) for the complex and highly nonlinear relationship between the system inputs and output(s), an improved particle swarm optimization was proposed, named particle swarm optimization with invasive weed (IW-PSO). In IW-PSO, PSO and IWO were integrated in parallel form based on the powerful local search ability of the invasive weed optimization (IWO), and after some iterations, the more diversiform and adaptive invasive weeds were introduced in particle swarm for improving the diversity, the worse particles comparing the fitness of the best previous position were partly mutated for assistantly local searching in invasive weeds. Simulation results show that the method has better searching ability and stability for some complex multimodal functions. Then, IW-PSO was introduced into low Nox combustion optimization, comparison results indicate that the proposed method is a fast and effective one to reduce NOx emissions.