摘要：This study proposes a Particle Swarm Optimization-based Augmented Lagrangian (PSOAL) algorithm which combines particle swarm optimization technique with a non-stationary penalty function method to solve constrained optimization and engineering design problems. A set of novel strategies are developed based on the particle feasibility to adaptively update critical parameters and a point-based local search procedure is embedded within the algorithm framework to improve the convergence property of the proposed algorithm. The 13 well-known constrained benchmark problems are solved and the obtained results are compared with other state-of-the-art algorithms. The results demonstrate that, the proposed PSOAL achieves higher accuracy compared to other considered algorithms. In addition, as an added benefit, PSOAL can also easily find out the Lagrange multipliers, which have great value for sensitivity analysis in practice but are almost not considered in most intelligent algorithms designed for constrained problems.
