【摘要】 为了确定使路网行程时间和尾气排放最优的禁左交叉口组合策略,以区域交叉口为例建立禁左双层规划模型。在搭建双层模型之前,分别针对单进口禁左的信号方案、转向车流的表示方法及转向延误的确定3个方面对现有禁左模型进行改进。模型中路网总阻抗和总排放量最小是上层的主要目标,并负责将禁左决策变量传到下层。下层以改进了约束条件的UE分配模型为基础,并通过Frank-Wolfe算法将求得的路段流量传到上层。模型采用遗传算法进行求解,在上下层的不停迭代中得到优化。通过对实例路网的求解,路网总阻抗和总排放加权和有效降低,验证了模型的可靠性。更多还原

【Abstract】 In order to determine the combination strategy of left-forbidden intersection with the optimal travel time and exhaust emission of the road network,a bi-level programming model of left-forbidden is established with an example of regional intersections. Before the bi-level programming model is built,the present model is improved from three aspects: the signal scheme of left-forbidden at single entrance,the expression method of turning traffic flow and the determination of turning delay. The upper layer model takes the total impedance and total emission of the system as the objective function,the left-forbidden decision variable is continuously passed to the lower level. The lower layer is based on the UE allocation model with improved constraint conditions,and the obtained link flow is transferred to the upper layer through the Frank-Wolfe algorithm. The model is solved by using genetic algorithm,which is optimized in the iterative process. The weighted sum of total impedance and total emission of road network is effectively reduced by solving the example road network,and the reliability of the model is verified.更多还原