基于加速参数自调整粒子群算法的物流配送优化模型

LOGISTICS DISTRIBUTION OPTIMISATION MODEL BASED ON ACCELERATION PARAMETERS SELF-ADJUSTED PSO

提出一种加速参数随个体适应值调整的改进粒子群(PSO)算法用来解决物流配送模型优化的多峰早熟问题。首先,从算法行为分析和向量分析的角度,根据当前粒子适应值和种群最优适应值设计一种简单实用的加速参数自调整策略。其次,通过理论和数值分析进而得到算法的全局收敛条件,为算法的实际应用提供理论基础。最后,结合改进PSO算法对物流配送模型进行研究。实验表明,基于个体适应值的加速参数变化策略对于PSO算法的"深度开发"和"全局探索"两个重要进化过程具有很好的平衡作用。算法的改进方式简单,未增加算法的时间复杂性,可以有效地对物流配送模型进行优化。

We presented an improved particle swarm optimisation （PSO）algorithm with the acceleration parameters adjusted according to the individual fitness value,which is used to solve the multimodal premature problem of logistics distribution optimisation model.First,from the perspectives of algorithm behaviour analysis and vector analysis,we design a simple and practical acceleration parameters self-adjustment strategy according to current particle fitness and population optimal fitness value.Secondly,through theoretical and numerical analyses we get the global convergence conditions of the algorithm,and provide the theoretical basis for the practical applications of the algorithm.Finally,we study the logistics distribution model in combination with the improved PSO algorithm.Experiments show that the acceleration parameters self-adaptation strategy based on individual fitness value has good balance role on two important evolution processes of PSO in ＂deep development＂and ＂global exploration＂.The improvement way of the algorithm is simple,without increasing its time complexity,and can effectively opti-mise the logistics distribution model.