基于极限学习机的VANET路段逻辑连通性算法研究  

Research on VANET Link Logical Connectivity Algorithm Based on Extreme Learning Machine

作　　者：胡敬文[1] 聂承静[2] 刘世明[3] 

机构地区：[1]佛山广播电视大学,广东佛山528000 [2]河北经贸大学公共管理学院,石家庄050061 [3]中南大学资源与安全工程学院,长沙410083

出　　处：《科技通报》2017年第12期179-182,193,共5页Bulletin of Science and Technology

基　　金：河北省自然科学基金项目(D2015207002)

摘　　要：VANET是一种应用在道路交通环境中的无线通信模式,数据的传输性能从很大程度上都依靠道路交通流的状态。针对大量因素影响着VANET逻辑连通性,在对其建模时无法只应用单纯的数学分析方法,本文通过采用数据挖掘与仿真分析两种策略有效结合的方法系统性的研究和分析了VANET逻辑连通性。提出了基于极限学习机的VANET路段逻辑连通性模型,进而明确了路段逻辑连通概率与道路交通状态参数之间的直接关联。仿真实验结果表明,相比较于非线性回归模型,本文提出的基于极限学习机的模型因其设置的维隐含层节点更多一些,所以其连通概率估计误差值更小。VANET is a kind of wireless communication mode, which is applied in the road traffic environment. The transmission performance of the data depends on the state of the road traffic flow. For a large number of factors affecting VANET logical connectivity, the modeling can not only use simple mathematical analysis method, this paper systematic research methods of effective combination of the two strategies and analysis of the VANET logical connectivity by using simulation data mining and analysis. A VANET link logic model is proposed based on the extreme learning machine, and the direct correlation between the link probability of the link and the state parameters of road traffic is defined. The simulation results show that, compared with the nonlinear regression model, extreme learning machine model because of its setting dimension of hidden layer nodes based on some more, so the connectivity probability estimation error is smaller.

关 键 词：VANET 极限学习机 路段逻辑连通概率 非线性回归 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]