The invention discloses an intersection signal control multi-period division method based on k-shape, and the method comprises the steps: carrying out the similarity matching of shapes during the division of traffic control periods through employing a k-shape algorithm, enabling the method to adapt to time sequences with different lengths, and determining the clustering number through a Clinski-Harabasz (CH) index and a contour coefficient method, calculating the distance between each time sequence and the new cluster center by adopting k-shape clustering, redistributing the time sequence to the nearest cluster according to the minimum distance, and updating the cluster center again until the cluster center is unchanged or reaches a preset number of iterations; and performing trend comparison on the output visual clustering result and the data, and dividing each group of data into the cluster to which the data belongs. According to the method, the defect that the k value needs to be determined manually in the prior art is overcome, a new thought is provided for signal timing multi-period division of the intersection, the crowdedness degree of the road intersection is relieved in different periods, and the operation efficiency of the intersection is improved.
本发明公开了一种基于k‑shape的交叉口信号控制多时段划分方法,通过采用k‑shape算法在交通控制时段划分时进行形状的相似性匹配以及适应不同长度的时间序列,并由Calinski‑Harabasz(CH)指数和轮廓系数法确定聚类数目,采用k‑shape聚类计算每个时间序列与新簇心之间的距离,根据最小距离重新分配时间序列到最近的簇,再次更新簇心至不变或达到预定的迭代次数;将输出后的可视化聚类结果与数据进行趋势对比,将各组数据划分到所属的簇中。本发明解决了传统需要人工确定k值的缺陷,为交叉口的信号配时多时段划分提供了新思路,且在不同时段缓解道路交叉口的拥挤程度,提高了交叉口的运行效率。
Intersection signal control multi-period division method based on k-shape
基于k-shape的交叉口信号控制多时段划分方法
2024-06-07
Patent
Electronic Resource
Chinese
IPC: | G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS |
European Patent Office | 2020
|Trans Tech Publications | 2012
|British Library Conference Proceedings | 2013
|European Patent Office | 2021
|TRAJECTORY DATA-BASED SIGNAL CONTROL PERIOD DIVISION METHOD
European Patent Office | 2021
|