摘要
基于传统模拟退火算法,通过引入记忆函数、结合GIS手段,运用SPSS聚类分析来确定初始化状态种群、多种群并行机制和新状态的产生.依据种群规模采用不同产生算法来改进算法,并将改进算法应用于城市物流中确定的多目标车辆路径优化问题,验证了算法的可行性与实用价值.此外,改进算法分别与传统模拟退火算法和GIS图解法相比较,优化效率和准确率都得到了很大的提高.
Based on traditional simulated annealing algorithm, this study combines mem- ory function, GIS and the cluster analysis of SPSS to determine the population initialization, multiple population of parallel mechanism and the initial states. The new sate has differ- ent produce algorithm depended on the population size. And the improved algorithm is certified for the feasibility and practical value by being applied to optimizing the certainty multi-objective vehicle routing problem in city logistics. Besides, it has higher optimization efficiency and accuracy compared with the traditional simulated annealing algorithm and the diagrammatizing model on GIS.
出处
《数学的实践与认识》
北大核心
2016年第2期105-113,共9页
Mathematics in Practice and Theory
基金
国家自然科学基金(71302005)
关键词
模拟退火
路径优化
车辆控制
多目标优化
simulated annealing
routing optimization
vehicle control
multi-objective op- timization
