Indicator-based particle swarm optimization with local search

An indicator-based particle swarm optimization algorithm with local search (IBPSO-LS) is proposed. IBPSO-LS with O(nN²) computational complexity integrates preference information of the decision maker into multi-objective PSO, and the local search is used to approach the Pareto-optimal solutions quickly so as to yield a computationally efficient and convergent procedure. Meanwhile, a mutation operator is adopted to avoid premature convergence and improve the exploratory capabilities of IBPSO. Simulations on several multi-objective benchmark instances indicate that IBPSO-LS has favorable performance with respect to different performance measures.