摘要
现有停车诱导系统中采用的停车场选择及路径诱导方法多局限于提供用户出行前多目标最优的停车场选择方案及路径诱导方案,无法实现用户行进过程中动态的多目标停车场及路径优化选择,此外由于现有多目标优化算法性能受限,无法快速获得真正高维多目标最优的决策方案。针对上述问题,提出一种综合用户出行前静态的和行进中动态的高维多目标停车场选择及路径诱导模型,并设计了高维多目标优化算法KS-MODE保证模型的高效求解。实验结果表明,KS-MODE在4~15目标优化问题上的收敛性能相比较于现有多种算法具有明显优势,基于KS-MODE的模型求解能够在城市交通网中实现出行前及行进中的五目标最优的动态停车场选择及路径诱导,证明了高维多目标优化算法是求解停车场选择及路径诱导的有效方法,能够提高现有停车诱导系统的诱导精度及智能化程度。
Most of the existing research on multi-objective parking lot choice and route guidance could only provide drivers approximate optimal schemes before trip, which couldn＇ t realize dynamic choice optimization during the trip. Besides, due to the limitation on accuracy and speed, most of the commonly used optimization algorithms eouldn＇ t achieve real many-objective optimal decision schemes quickly. In order to overcome the above-mentioned difficulties, this paper proposed a novel parking lot choice and route guidance model, in addition, it designed a novel many-objective evolutionary algorithm （MOEA） based on improved K-dominated sorting （KS-MODE） to solve the model. Simulation results prove the superiority of KS-MODE in solving 4-15 objective optimization problems compare with several state-of-the-art algorithms, the model solved by KS-MODE can also provide drivers 5-objective optimal schemes accurately and dynamically in city transportation network. The method propose in this paper proves that, MOEA is efficient in solving parking lot choice and route guidance, it can improve accuracy and intelligent level of the parking guidance system.
出处
《计算机应用研究》
CSCD
北大核心
2015年第7期2009-2013,2026,共6页
Application Research of Computers
基金
黑龙江省博士后基金资助项目(LBH-Z12073)
辽宁省博士科研启动基金资助项目(201205118)
辽宁省教育厅科学技术研究一般项目(L2012458)
辽宁省交通高等专科学校优秀人才计划专项基金资助项目(Lnccrc02)
关键词
停车诱导系统
停车场选择
动态路径诱导
高维多目标优化
parking guidance system
parking lot choice
dynamic route guidance
many-objective optimization
