| 171 | 11 | 205 |
| 下载次数 | 被引频次 | 阅读次数 |
针对城市快速路交通事件检测问题,提出了一种基于自适应遗传算法与神经网络相结合的自动检测算法。通过改进的自适应遗传算法优化神经网络结构和权值参数,保证了神经网络能以较小规模和最优的权值参数来描述事件发生与交通参数间的映射关系,从而提高检测效果。利用PARAMICS交通软件模拟了北京市京通快速路从大望桥到四惠桥路段间的一组交通数据,仿真结果表明,该算法同现有的典型算法相比较,具有较高的检测率和较快的检测速度。
Abstract：An automatic incident detection algorithm based on an adaptive genetic algorithm combined with neural networks was proposed for urban freeway traffic.The neural structure and parameters were optimized by the adaptive genetic algorithm,which could describe the relation between traffic incident and the traffic parameters and improve the detection results.A set relevant traffic data was obtained with PARAMICS traffic software by simulating the road section between DAWANG bridge and SIHUI bridge of Jingtong freeway in Beijing city.The simulation results indicated that the given algorithm has better detection rate and detection speed for urban freeway traffic,compared with some classical detection algorithms.
[1]Cheu R.L.,Ritchie S.G.Automated detection of lane-blocking freeway incidents using artificial neural networks[J].Transportation Research Part C,1996,3(6):371–388.
[2]Srinivasan.D,Sanyal.S and Sharma.V.Freeway incident detection using hybrid fuzzy neural network[J].IEEE Transportion Intelligent Transportion System,2007,1,(4):249–259.
[3]Thomas.K,Dia.H.Comparative evaluation of freeway incident detection models using field data[J].IEEE Transportion Intelligent Transportion System,2006,153(3):230-241.
[4]Chen Shuyan;Wang Wei,Qu Gaofeng,et al.Application of neural network ensembles to incident detection proceedings of the 2007[C].shenzhen:IEEE International Conference on Integration Technology,2007,388-393.
[5]姜紫峰,刘小坤.基于神经网络的交通事件检测算法[J].西安交通大学学报,2000,20(3):67-73.
[6]吕琪,王慧.基于动态神经网络模型的交通事件检测算法[J].公路交通科技,2003,20(6):105-108.
[7]樊小红,荆便顺.基于遗传算法的交通事件检测[J].长安大学学报,2005,(4):70-72.
[8]王清,马广富等.一种基于遗传算法的神经网络控制方法研究[J].系统仿真学报,2006,18(4):1070-1073.
基本信息:
中图分类号:U491.116
引用信息:
[1]田秋芳,陈阳舟,张利国.城市快速路交通事件检测的自适应算法研究[J],2010,8(04):99-103+125.
基金信息:
国家自然科学基金(60904069)