This paper will develop a model simulation mixing traffic flow in the signal controlintersection. Applying the model, the schemes of traffic channeling and controlling in the typical intersection are analyzed and opti...This paper will develop a model simulation mixing traffic flow in the signal controlintersection. Applying the model, the schemes of traffic channeling and controlling in the typical intersection are analyzed and optimized and the optimal scheme make the average stop frequency be reduced by 31.8% and average stop time be reduced by 27.7%.展开更多
窄路段作为交通场景中不可避免的瓶颈路段,其短时车流量预测对优化路径规划、改善交通状况具有重要意义。针对窄路段的时效性,同时考虑适用模型的准确度,提出一种基于佳点集初始化种群、非线性参数控制及柯西变异扰动的改进鲸鱼优化算法...窄路段作为交通场景中不可避免的瓶颈路段,其短时车流量预测对优化路径规划、改善交通状况具有重要意义。针对窄路段的时效性,同时考虑适用模型的准确度,提出一种基于佳点集初始化种群、非线性参数控制及柯西变异扰动的改进鲸鱼优化算法(IWOA)-门控循环单元(GRU)的窄路短时车流量预测模型,以SUMO(Simulation of Urban Mobility)仿真数据进行了实证研究。对比实验结果显示,IWOA具有较好的全局性、收敛速度且更加稳定。基于IWOA-GRU的窄路短时车流量预测模型,均方根误差(RMSE)指标相较于WOA-GRU、PSO-GRU、长短期记忆神经(LSTM)网络分别降低10.96%、28.71%、42.23%,平均绝对百分比误差(MAPE)指标分别降低13.92%、46.18%、52.83%,有较为显著的准确性和稳定性。展开更多
文摘This paper will develop a model simulation mixing traffic flow in the signal controlintersection. Applying the model, the schemes of traffic channeling and controlling in the typical intersection are analyzed and optimized and the optimal scheme make the average stop frequency be reduced by 31.8% and average stop time be reduced by 27.7%.
文摘窄路段作为交通场景中不可避免的瓶颈路段,其短时车流量预测对优化路径规划、改善交通状况具有重要意义。针对窄路段的时效性,同时考虑适用模型的准确度,提出一种基于佳点集初始化种群、非线性参数控制及柯西变异扰动的改进鲸鱼优化算法(IWOA)-门控循环单元(GRU)的窄路短时车流量预测模型,以SUMO(Simulation of Urban Mobility)仿真数据进行了实证研究。对比实验结果显示,IWOA具有较好的全局性、收敛速度且更加稳定。基于IWOA-GRU的窄路短时车流量预测模型,均方根误差(RMSE)指标相较于WOA-GRU、PSO-GRU、长短期记忆神经(LSTM)网络分别降低10.96%、28.71%、42.23%,平均绝对百分比误差(MAPE)指标分别降低13.92%、46.18%、52.83%,有较为显著的准确性和稳定性。
