The basic principles of GPS (Global Positioning System) and DGPS (Differential GPS) are described. The principle and structure of vehicle navigation systems, and its application to the urban traffic flow guidance are ...The basic principles of GPS (Global Positioning System) and DGPS (Differential GPS) are described. The principle and structure of vehicle navigation systems, and its application to the urban traffic flow guidance are analyzed. Then, an area coordinated adaptive control system based on DGPS and a traffic flow guidance information system based on DGPS are put forward, and their working principles and functions are researched. This is to provides a new way for the development of urban road traffic control systems.展开更多
文中提出了一种基于路段关联度的城市交通流量Apriori-LSTM(Apriori-long short term memory network)预测模型.处理卡口检测器数据,统计交通量并提取车辆轨迹,采用Apriori算法分析预测时段内目标路段与关联路段的时空相关性,计算关联...文中提出了一种基于路段关联度的城市交通流量Apriori-LSTM(Apriori-long short term memory network)预测模型.处理卡口检测器数据,统计交通量并提取车辆轨迹,采用Apriori算法分析预测时段内目标路段与关联路段的时空相关性,计算关联路段支持度;求解预测时段内关联路段到目标路段的流入量,构建LSTM预测的输入矩阵、并采用LSTM预测路段短时流量.采用实例进行验证,对迭代次数、隐藏层神经元个数和步长进行参数灵敏度分析,并与单一的LSTM预测结果进行比较.结果表明:Apriori-LSTM的平均绝对误差降至3.8%,平均绝对百分误差和平均均方误差均有显著降低,均等系数有所提高,模型稳定性更好,达到了更好预测效果.展开更多
文摘The basic principles of GPS (Global Positioning System) and DGPS (Differential GPS) are described. The principle and structure of vehicle navigation systems, and its application to the urban traffic flow guidance are analyzed. Then, an area coordinated adaptive control system based on DGPS and a traffic flow guidance information system based on DGPS are put forward, and their working principles and functions are researched. This is to provides a new way for the development of urban road traffic control systems.
文摘文中提出了一种基于路段关联度的城市交通流量Apriori-LSTM(Apriori-long short term memory network)预测模型.处理卡口检测器数据,统计交通量并提取车辆轨迹,采用Apriori算法分析预测时段内目标路段与关联路段的时空相关性,计算关联路段支持度;求解预测时段内关联路段到目标路段的流入量,构建LSTM预测的输入矩阵、并采用LSTM预测路段短时流量.采用实例进行验证,对迭代次数、隐藏层神经元个数和步长进行参数灵敏度分析,并与单一的LSTM预测结果进行比较.结果表明:Apriori-LSTM的平均绝对误差降至3.8%,平均绝对百分误差和平均均方误差均有显著降低,均等系数有所提高,模型稳定性更好,达到了更好预测效果.