PSC Ship-Selecting Model Based on Improved Particle Swarm Optimization and BP Neural Network Algorithm

摘要

PSC targeting model has drew much attention recent years. Based on the analysis of PSC targeting mechanisms and algorithms of primary MOU organizations in the maritime society, as 2009/16/EC NIR for instance, a more scientific mathematical targeting model relying on intelligent optimization algorithms is developed in this paper. This algorithm exploits the improved particle swarm-BP neural network mechanism, confronting the weakness of neural network which is easy to drop in local minimum. It could adaptively adjust inertia weights, update speed and position according to premature convergence degree as well as individual fitness value, by exploring improved PSO algorithm to train BP network. The effectiveness and reliability of the algorithm applied to PSC ship-selecting is validated, based on the real cases obtained from the THETIS Inspection database of Paris-MoU. The testing results demonstrate that the proposed PSC ship-selecting model could improve the performance not only on speed of convergence, but also the precision of convergence.