受到部分交通设施限载和限高的影响,货车司机需要合理的导航路径规划,该路径规划既要满足全路径交通设施限载和限高,又要实现运行的最小成本,目前缺乏高效的算法流程,因此本文对基础路网进行拓扑,提出了将道路上的货车限制信息相匹配,...受到部分交通设施限载和限高的影响,货车司机需要合理的导航路径规划,该路径规划既要满足全路径交通设施限载和限高,又要实现运行的最小成本,目前缺乏高效的算法流程,因此本文对基础路网进行拓扑,提出了将道路上的货车限制信息相匹配,形成面向货车的约束路网,考虑道路约束信息对路径规划的相互关系,通过计算通行成本,标记合适道路,构建面向超载超限货车的导航系统路径规划方法。最后本文以浙江大学紫金港校区区域为例,构建了与地图相匹配的路径规划算法,展示了货车路径规划工作原理的工作逻辑。Affected by some traffic facilities with load and height restrictions, truck drivers need reasonable navigation path planning, which should satisfy the full path traffic facilities with load and height restrictions, and also realize the minimum cost of operation, so this paper topologizes the basic road network, proposes to match the truck restriction information on the road to form a truck-oriented constrained road network, and considers the interrelationship of road constraint information on path planning. By calculating the access cost and marking the suitable roads, the path planning method of the navigation system for overloaded and overloaded trucks is constructed. Finally, this paper takes the Zijingang campus area of Zhejiang University as an example, constructs the path planning algorithm matching with the map, and shows the working logic of the working principle of truck path planning.展开更多
为探究公交站点之间的关联度并对公交客流进行更精准的实时预测,本文提出基于Attention的交通预测核心算法(Traffic Forecast Model Based Attention,TFMA),结合数据预处理和站点信息编码完成基于站点实时关联度的短时公交客流预测方法...为探究公交站点之间的关联度并对公交客流进行更精准的实时预测,本文提出基于Attention的交通预测核心算法(Traffic Forecast Model Based Attention,TFMA),结合数据预处理和站点信息编码完成基于站点实时关联度的短时公交客流预测方法。该方法首先创新性地提出了站点实时关联度,可实现对目标站点客流量更精准的预测;其次,在公交站点的编码信息中融入线路站点信息、客流变化率、天气、日期等关联因素;接着,该方法依靠Attention机制计算站点实时关联度;核心算法中使用multi-headed机制、增加通道和残差连接进一步提升预测能力;最后,以苏州市公交数据进行验证。结果显示:在准确率上,对比多元线性回归的53.8%、GRU(Gated Recurrent Unit)的66.9%和LightGBM(Light Gradient Boosting Machine)的81.2%,本文提出的基于站点实时关联度的短时公交客流预测方法的准确率在90%以上,表明该方法具备优秀的短时公交客流预测能力。展开更多
文摘受到部分交通设施限载和限高的影响,货车司机需要合理的导航路径规划,该路径规划既要满足全路径交通设施限载和限高,又要实现运行的最小成本,目前缺乏高效的算法流程,因此本文对基础路网进行拓扑,提出了将道路上的货车限制信息相匹配,形成面向货车的约束路网,考虑道路约束信息对路径规划的相互关系,通过计算通行成本,标记合适道路,构建面向超载超限货车的导航系统路径规划方法。最后本文以浙江大学紫金港校区区域为例,构建了与地图相匹配的路径规划算法,展示了货车路径规划工作原理的工作逻辑。Affected by some traffic facilities with load and height restrictions, truck drivers need reasonable navigation path planning, which should satisfy the full path traffic facilities with load and height restrictions, and also realize the minimum cost of operation, so this paper topologizes the basic road network, proposes to match the truck restriction information on the road to form a truck-oriented constrained road network, and considers the interrelationship of road constraint information on path planning. By calculating the access cost and marking the suitable roads, the path planning method of the navigation system for overloaded and overloaded trucks is constructed. Finally, this paper takes the Zijingang campus area of Zhejiang University as an example, constructs the path planning algorithm matching with the map, and shows the working logic of the working principle of truck path planning.
文摘为探究公交站点之间的关联度并对公交客流进行更精准的实时预测,本文提出基于Attention的交通预测核心算法(Traffic Forecast Model Based Attention,TFMA),结合数据预处理和站点信息编码完成基于站点实时关联度的短时公交客流预测方法。该方法首先创新性地提出了站点实时关联度,可实现对目标站点客流量更精准的预测;其次,在公交站点的编码信息中融入线路站点信息、客流变化率、天气、日期等关联因素;接着,该方法依靠Attention机制计算站点实时关联度;核心算法中使用multi-headed机制、增加通道和残差连接进一步提升预测能力;最后,以苏州市公交数据进行验证。结果显示:在准确率上,对比多元线性回归的53.8%、GRU(Gated Recurrent Unit)的66.9%和LightGBM(Light Gradient Boosting Machine)的81.2%,本文提出的基于站点实时关联度的短时公交客流预测方法的准确率在90%以上,表明该方法具备优秀的短时公交客流预测能力。
