Bottleneck算法的改进与仿真

Improvement and Simulation of Bottleneck Ramp-Metering Algorithm

分析了Bottleneck算法的不足,指出在路段较为拥挤的情况下该算法中因不具备预测机制而产生的时滞会引起控制误差以及主线拥挤时会造成交通流大幅波动等,产生不足的原因是算法的两个限制条件过于严格,进而提出了改进方案。改进算法中对两个限制条件适当放宽,在此基础上以宁杭高速(江苏段)一期为背景进行了仿真,并与原算法进行了比较。仿真结果表明,改进后的算法能更快更平稳地使交通流达到稳定状态,且具有较强的鲁棒性。

The Bottleneck Ramp- metering Algorithm is discribed briefly. Some shortcomings of the algorithm are exposed due to lack of prediction mechanism, including time- lag which leads to governing error, and wide fluctuation of traffic flow when the mainline have traffic jam. Virtually, these shortcomings result from the two limiting conditions of the algrithm, which are too restrictive. This paper tries to .soften the oonditions to get an＂Extended Bottleneck Ramp - metering Algorithm＇. At last, simulation is carried out and eontrol effects are compared between the origin and extended Bottleneck Rampmetering Algorithm. The simulation results show that the extended algorithm is more robust and can stabilify the traffic flow more quickly and smoothly.