摘要
随着经济和科技的迅猛发展,人民的生活水平逐渐提高,城市人口和机动车保有量急速增长,交通拥堵问题在世界各个大中城市普遍存在。该文在对算法的发展进行了解后,选择模糊C均值聚类分析(FCM)、集对分析和BP神经网络3种算法,进行交通状态判别研究。通过Matlab程序对实际交通数据进行仿真,从准确率和运行时间2方面对算法效果进行了对比分析,在所用的实际交通流数据和运行程序下,集对分析算法是相对较好的算法。
With the rapid advancement of economy and technology,people's living standards have gradually increased,urban population and motor vehicle possession have grown rapidly,and traffic congestion is widespread in all major cities in the world.After understanding the development of the algorithm,fuzzy C-means clustering analysis(FCM),set pair analysis and BP neural network were selected to study the traffic state discrimination in this paper.The actual traffic data is simulated by Matlab program,and the algorithm effects are compared and analyzed from the aspects of accuracy and running time.The set pair analysis algorithm is a relatively good algorithm under the actual traffic flow data and running procedures used.
作者
高林
李琪琪
GAO Lin;LI Qiqi(College of Automation and Electronical Engineering,Qingdao University of Science and Technology,Shandong Qingdao 266061,China)
出处
《工业仪表与自动化装置》
2020年第2期8-11,共4页
Industrial Instrumentation & Automation
基金
山东省自然科学基金“基于大数据技术的交通OD分析算法研究”(ZR2014FL018)
青岛科技大学博士科研基金“数据挖掘技术在智能交通系统中的应用”(010022530)。
关键词
聚类分析
集对分析
BP神经网络
cluster analysis
set pair analysis
BP neural network
