机动车交通冲突技术研究综述

对机动车-机动车交通冲突定义、交通冲突度量指标、冲突严重性判定以及交通冲突评价与预测等的国内外研究进展做了归纳和综述。分析表明,目前交通冲突研究存在以下主要问题:传统的冲突度量指标存在各自的局限性和不一致性,冲突潜在碰撞的后果严重性研究处于初始阶段,交通事件分级模型和相关性需要进一步研究和确认,冲突严重性判定存在一致性问题,交通冲突产生和发展的微观机理研究不足,缺乏真实环境下整个区域相互作用车辆精确连续轨迹追踪的数据获取手段。建议未来从以下方向进行优化:复合的改进度量指标比单一指标更为科学合理,并要考虑冲突潜在碰撞的后果严重性;统一和规范交通冲突度量指标的使用也有一定意义;需要针对多个参与者的"区域连锁冲突"进行更深入的研究;可通过采集区域多车辆连续时空轨迹大数据,得到区域多车辆冲突时空演变模型,并进行交通冲突实时预警和干预研究;另外可通过大量精确数据量化和统一区分各级交通冲突的阈值,并验证交通冲突技术的有效性;在不同设施对象及冲突的空间特征等方面的探索等也会丰富交通冲突研究体系;最后,以上所有的改进方向离不开高精度大范围的采集方法和高效精确的处理手段,故急需真实环境下、长时间的整个区域相互作用车辆精确连续轨迹追踪的大数据获取手段。

Review of Research on Traffic Conflict Techniques

In this paper, the definition, measurement index, judgment of severity, and evaluation of vehicle-vehicle traffic conflicts were summarized and reviewed. The following main problems were found in the study of traffic conflicts:Limitations and inconsistencies existed in traditional conflict measurement indexes. Studies on the severity of potential collisions were in the initial stages. The classification model and correlation of traffic events needed to be studied further and confirmed. Consistency of conflict severity determination existed. There was insufficient research on the microcosmic mechanisms of traffic conflicts. There was a lack of data acquisition methods for accurate, continuous tracking of vehicles while traveling. It is suggested to optimize in the following directions in the future:Compound improvement measures may be more scientific and reasonable than single measures in considering the severity of the potential collision impact of traffic conflicts. It is also meaningful to unify and standardize the use of the traffic conflict measurement index. More in-depth research is needed on the "regional chain conflict" of multiple traffic participants. The time and space evolution model of regional multi-vehicle conflicts can be obtained by collecting big data from the continuous high-precision trajectories of regional multi-vehicle traffic, and real-time early warning and intervention research of traffic conflicts can be conducted. In addition, accurate big data can be used to quantify and uniformly distinguish the thresholds of traffic conflicts at all levels and verify the effectiveness of traffic conflict technology. The exploration of different aspects of traffic infrastructure and spatial characteristics of accidents will also enrich the study of traffic conflicts. Finally, all these strategies for improvement are inseparable from high-precision and large-scale collection methods and efficient and accurate means of processing data. Therefore, it is urgent to acquire big data for accurate and continuous tracking of vehicle conflicts in the region over time in real environments.