交通网络中多路径优化选择算法的研究

Study on Multi-path Optimization Routing Algorithm in Traffic Network

指出了传统的多路径问题及其算法存在的缺陷,给出了多路径优化算法的几个定义,在此基础上,提出基于重叠惩罚的多路径选择算法,设计出了重叠惩罚函数数学模型结构以及路径相似性评价函数定义,对惩罚函数中的协调因数敏感性进行分析,给出了合适的协调因数。该算法能有效地解决候选路径重叠相似性问题,而平均总旅行时间比率基本不变,从而更适合驾驶人员的实际选路需求。与传统k最短路径算法进行了试验比较,结果表明该算法明显优于传统k最短路径算法。

The traditional multi-shortest path algorithms are based on a mathematical perspective, and alternate routes have very similarity. On the basis of several definitions, a multiple path selection algorithm based on overlapping penalty is proposed, overlapping penalty function and route similarity evaluating function are designed, harmony factor sensitivity are analysed and suitable harmony factor value is presented. Experiment results show that the proposed, algorithm solves candidate route overlapping effectively, while average total travel time ratio does not change. It satisfies driver＇ s routing path demand and obviously superior to traditional k- shortest path algorithms by comparing to experimental data.