【摘要】 交叉口车流量随机性与不确定性导致车辆信息采集结果的差异化明显。以传感器信息融合为技术支撑，提出交叉口车流量多时段控制方法。利用多传感器采集交叉口交通信息，预处理采集到的数据。采用由指标层与目标层构成的信息融合模型，完成交通信息融合。根据信息之间的关联性聚类所有交通信息，明确分类数并划分控制时段。综合考量交叉口通行效率与环境等因素，结合排队长度、平均延误及尾气排放量建立多目标多时段控制模型，由层次分析法明确各目标权值后，采用改进萤火虫算法进行求解，实现交叉口车流量多时段控制。实验结果表明，该方法能有效改善交叉口的拥堵情况，提升通行效率，降低车流饱和度与延误时间，应用优势显著。更多还原

【Abstract】 The randomness and uncertainty of intersection traffic flow lead to obvious differences in vehicle information collection results. Based on sensor information fusion, a multi period control method of intersection traffic flow is proposed. The intersection traffic information is collected by multi sensors, the collected data are preprocessed, and the traffic information fusion is completed by using the information fusion model composed of index layer and target layer. All traffic information is Clustered according to the correlation between information, the classification number is clarified and the control period is divided. The factors such as intersection traffic efficiency and environment being onsidered, combined with queue length, average delay and exhaust emission, a multi objective and multi period control model is established. After the weight of each target is determined by analytic hierarchy process, the improved firefly algorithm is used to realize the multi period control of intersection traffic flow. The experimental results show that the proposed method can effectively alleviate the congestion of the intersection, and improve the traffic efficiency, reduce the traffic saturation and delay time, and it has significant application advantages.更多还原