考虑减排的公交线网分时差异化票价动态优化

Dynamic Optimization of Time-dependent Differential Fare for Urban Public Transport Network Considering Emission Reduction

为在降低公交线网系统社会成本的同时实现低碳绿色出行,本文考虑OD间出行需求在工作日与周末的动态变化、出发时间差异以及出行者的有限理性行为,分别以降低碳排放与降低社会成本为目标函数,以分时差异化票价为优化变量,构建动态多目标优化模型。进一步,为使价格政策的实施对需求环境的变化具有预判性,将基于BP神经网络的种群分布预测算子、OD矩阵均衡算法以及具有优秀性能遗传算子的静态多目标粒子群算法进行有机结合,构建新的集群智能动态多目标优化算法,求解分时差异化票价。算例及实例研究发现:在线网整体出行需求水平较低时,下调公交票价有助于降低碳排放水平;相对于固定票制,分时差异化票价可有效降低社会成本与碳排量,使不同线路上的客流更加均衡。

To achieve low-carbon travel and reduce the social cost of public transport network systems,this paper considers the dynamic changes in travel demand between weekdays and weekends,differences in departure times,and travelers'bounded rational behavior.We construct a dynamic multi-objective optimization model with the objective of reducing carbon emissions and social costs.This model takes the time-dependent differential fare as the decision variable.To ensure that fare implementation can adapt to changes in demand,we use a population distribution prediction operator based on the BP neural network,an OD matrix equilibrium algorithm,and a static multi-objective particle swarm optimization algorithm with excellent genetic operators.By combining these approaches,we design a new cluster intelligent dynamic multi-objective optimization algorithm to solve the time-dependent differentiated fare.To evaluate the effectiveness of our approach,we provide calculation examples and conduct case studies.The results reveal the following findings:(1)When the overall travel demand level of the bus network is low,reducing bus ticket prices is more effective in reducing carbon emissions.(2)Compared to a fixed ticket system,the time-dependent differentiated fare can reduce social costs and carbon emissions while also making the passenger flow across different lines more balanced.