换乘导向的轨道交通网络发车时间优化研究
Transfer-Oriented Dispatching Optimization of Rail Transit Network

针对大城市轨道交通网络换乘站在高峰时段换乘乘客聚集造成的瓶颈，带来的运营安全性能差、效率低等问题，基于非线性规划方法，构建了以网络总换乘等车时间最短为目标的数学模型。该模型考虑了乘客在换乘站的走行时间，以列车发车时间的调整变化量为决策变量，利用调整轨道交通列车在起点站的发车时间及时刻表，协调了列车在换乘站的换乘衔接时间。用模拟退火算法进行求解，得到优化的列车发车时间及时刻表。对13个换乘站及5条线路组成的深圳市轨道交通网络进行优化，结果表明，早高峰小时的优化方案使网络乘客总的换乘等车时间减少了689 h，提升系统换乘效率22%。
： As bottlenecks of rail transit network due to its high volume of transfer passengers during peak periods in big cities, transfer stations are facing problems of increasing safety risk, operation inefficiency, etc. To solve these problems, a mathematical model is developed using the nonlinear programming method to minimize the total transfer time in the rail transit network. Taking passengers' walking time in transfer into consideration, this model selects the adjustment amount of train dispatching time as the decision variable to optimize the transfer coordination time, by adjusting the departure time of trains at the start terminal and their timetables. Then, optimal solutions of train departure time and timetables are obtained using a simulated annealing algorithm. As a case study, the proposed methodology was applied to optimization of a real-life metro network with five lines intersecting at thirteen stations. The results show that compared to the existing operation plan, the total transfer waiting time of the optimized schedule in the morning peak hour is reduced by 689 h, and the system transfer efficiency is increased by 22%