基于智能代理的交通分配建模
|
摘要
针对现有交通分配研究对驾驶员有限理性、道路交通动态随机性的忽视,提出一种基于智能代理的动态交通分配模型IATAM,其以驾驶员为代理,依据驾驶员的路线偏好,综合驾驶员信息处理的模糊随机过程,考虑邻居驾驶员影响,提出神经网络结构的驾驶员智能路线决策机制。实验中IATAM模型检测点车流量与真实车流量的平均相对误差减小到6.42%,表明IATAM模型实验精度更高,更符合复杂多变的交通环境。基于智能代理进行交通分配,更能反映驾驶员的异构性和路线决策的模糊随机性,并有效提高交通分配精度。
On account of the neglect of drivers' bounded rationality and traffic dynamic and random in the existing traffic assignment study,the paper proposes the traffic assignment model based on intelligent agent( IATAM). Each driver is defined as an agent and has different route preference. Based on the composition of fuzzy stochastic information processes of all drivers,as well as the effect of neighbor drivers,the agent route decision-making mechanism is constructed which is like neural network. Average Relative Error( ARE) of IATAM model and actual road network in checkpoint volume is reduced to 6. 42%,indicating that the IATAM model has higher prediction precision and conforms more to the complex traffic environment. Traffic assignment based on intelligent agent reflects the isomerism of drivers and the vagueness and randomness of route decision-making,which improves the accuracy of traffic volume significantly.
On account of the neglect of drivers' bounded rationality and traffic dynamic and random in the existing traffic assignment study,the paper proposes the traffic assignment model based on intelligent agent( IATAM). Each driver is defined as an agent and has different route preference. Based on the composition of fuzzy stochastic information processes of all drivers,as well as the effect of neighbor drivers,the agent route decision-making mechanism is constructed which is like neural network. Average Relative Error( ARE) of IATAM model and actual road network in checkpoint volume is reduced to 6. 42%,indicating that the IATAM model has higher prediction precision and conforms more to the complex traffic environment. Traffic assignment based on intelligent agent reflects the isomerism of drivers and the vagueness and randomness of route decision-making,which improves the accuracy of traffic volume significantly.
引文
[1]陈彦如,蒲云.用遗传算法解决固定需求交通平衡分配问题[J].西南交通大学学报(自然科学版),2000,35(1):44-47.
[2]Lu Shiqiang,Dai Shoufeng,Liu Ximin.A discrete traffic kinetic model-integrating the lagged cell transmission and continuous traffic kinetic models[J].Transportation Research Part C Emerging Technologies,2011,19(2):196-205.
[3]周溪召.驾驶员实时动态路径选择行为组合模型[J].上海理工大学学报,2002,24(2):109-112.
[4]王卫卫,赵小梅,李新刚,等.VMS对驾驶员路径选择影响的实证研究与建模[J].交通运输系统工程与信息,2013,13(3):60-64.
[5]李强,缪立新,段进宇.基于驾驶员路径选择的动态交通仿真模型[J].公路交通科技,2001,18(6):65-69.
[6]朱大奇.人工神经网络研究现状及其展望[J].江南大学学报(自然科学版),2004,3(1):103-110.
[7]昌霞,孙瑜,冉婕,等.基于时态的模糊描述逻辑初探[J].微型机与应用,2010,29(6):75-77.
[8]伍建国,王峰.城市道路交通数据采集系统检测器优化布点研究[J].公路交通科技,2004,21(2):88-91.
[2]Lu Shiqiang,Dai Shoufeng,Liu Ximin.A discrete traffic kinetic model-integrating the lagged cell transmission and continuous traffic kinetic models[J].Transportation Research Part C Emerging Technologies,2011,19(2):196-205.
[3]周溪召.驾驶员实时动态路径选择行为组合模型[J].上海理工大学学报,2002,24(2):109-112.
[4]王卫卫,赵小梅,李新刚,等.VMS对驾驶员路径选择影响的实证研究与建模[J].交通运输系统工程与信息,2013,13(3):60-64.
[5]李强,缪立新,段进宇.基于驾驶员路径选择的动态交通仿真模型[J].公路交通科技,2001,18(6):65-69.
[6]朱大奇.人工神经网络研究现状及其展望[J].江南大学学报(自然科学版),2004,3(1):103-110.
[7]昌霞,孙瑜,冉婕,等.基于时态的模糊描述逻辑初探[J].微型机与应用,2010,29(6):75-77.
[8]伍建国,王峰.城市道路交通数据采集系统检测器优化布点研究[J].公路交通科技,2004,21(2):88-91.