一种优化的城市交通状态判别方法  

An Optimized Method for Urban Traffic State Discrimination

作　　者：黄佳慧 黄鹤[1,2] 杨澜 王会峰 茹锋[1,2] HUANG Jiahui;HUANG He;YANG Lan;WANG Huifeng;RU Feng(School of Electronic and Control Engineering,Chang'an University,Xi'an,Shaanxi 710064,China;Xi'an Key Laboratory of Intelligent Expressway Information Fusion and Control,Xi'an,Shaanxi 710064,China)

机构地区：[1]长安大学电子与控制工程学院,陕西西安710064 [2]西安市智慧高速公路信息融合与控制重点实验室,陕西西安710064

出　　处：《复旦学报(自然科学版)》2025年第1期1-13,共13页Journal of Fudan University:Natural Science

基　　金：国家重点研发计划(2021YFB2501200);国家自然科学基金(52172324,52172379);中央高校基本科研业务费资助项目(300102325501);西安市智慧高速公路信息融合与控制重点实验室(长安大学)开放基金(300102323502)。

摘　　要：在对交通参数数据进行聚类分析时,传统K均值聚类(KMC)存在聚类中心初始化过程随机性较大,聚类边界划分不清晰且迭代速度慢、易陷入局部最优解等问题。针对这些问题,提出了一种基于多策略自适应繁衍传递鹈鹕算法(MARPOA)优化的KMC交叉迭代聚类算法(MARPOA-KMC),实现对交通运行状态的准确划分。首先,设计了一种聚拢映射法,解决了KMC随机初始化引起的交通状态聚类结果不稳定问题;然后,通过多策略自适应繁衍传递来修正当前代最优解,解决了鹈鹕算法搜索路径单一带来的全局寻优能力差和搜索精度不足的问题;最后,将MARPOA引入KMC优化寻找聚类中心的过程,提高了聚类精度。利用标准测试函数对MARPOA、POA、SSA、GWO、MFO算法进行比较,由性能指标可以看出,提出的MARPOA相较于其他比较算法,在收敛速度和精度等方面都表现最佳。由PeMSD8公开交通数据集上的验证结果可知,相对于比较算法,提出的MARPOA-KMC算法能够更快速、准确地划分交通运行状态。In the classification process for traffic state discrimination,traditional K-Means Clustering(KMC)suffers from issues such as the randomness in the initialization of cluster centers,unclear division of clustering boundaries,slow iteration speed,and susceptibility to local optima.To address these challenges,a KMC algorithm optimized by a Multi-strategy Adaptive Reproduction Pelican Optimization Algorithm(MARPOA),known as MARPOA-KMC,has been proposed to accurately categorize traffic operation states.The approach involves the following steps.Firstly,a convergence mapping method is designed to resolve the instability in traffic state clustering results caused by the random initialization of KMC.Secondly,the MARPOA is used to correct the optimal solutions of the current generation,addressing the issues of poor global optimization capability and insufficient search precision due to the singular search path of the Pelican Algorithm.Finally,the MARPOA is integrated into the KMC process to optimize the search for cluster centers,thereby enhancing clustering accuracy.Standard test functions are used to compare the performance of the improved MARPOA with other optimization algorithms such as Particle Optimization Algorithm(POA),Sparrow Search Algorithm(SSA),Grey Wolf Optimization(GWO),and Moth Flame Optimization(MFO).Performance metrics indicate that MARPOA outperforms the other algorithms in terms of convergence speed and precision.The California Department of Transportation's Performance Measurement System(PeMSD8)public traffic dataset is used for experimental validation.The results show that,compared to the comparative algorithms,the proposed MARPOA-KMC algorithm can more quickly and accurately categorize traffic operation states.

关 键 词：鹈鹕优化算法 K均值聚类 智能交通系统 交通状态判别 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]