Abstract Filtering of motorized two-wheelers (MTWs) is a common practice in dense urban heterogeneous traffic environments where they often tend to navigate through the available lateral spaces (commonly termed as ‘pore’ in the literature) described by the vehicles in-front. Considering the increased vulnerability of MTW riders in dense urban systems, proper evaluation and modeling of pore acceptance/rejection predictions of MTWs can essentially provide a safer driving environment to MTW riders and the surrounding vehicles in a cognitive architecture, augment the reliability and predictability of microsimulation models and ameliorate the overall traffic flow phenomena. The current study therefore investigates the applicability of Raff’s method, binary logit model and support vector machines (SVM) in predicting the pore acceptance decisions of MTW riders during filtering in urban mid-block sections. The results of the study suggested that SVM technique could be considered as a potential tool for estimating the pore acceptance and rejection predictions of MTW riders, which can be further implemented in a cooperative intelligent transport systems environment for an overall safe yet smooth flow of traffic.

中文翻译：

---

城市中段过滤过程中机动两轮车的孔隙接受度预测

摘要 机动两轮车 (MTW)​​ 的过滤是密集城市异构交通环境中的常见做法，在这些环境中，它们通常倾向于通过前方车辆描述的可用横向空间（在文献中通常称为“孔隙”）。考虑到 MTW 骑手在密集城市系统中的脆弱性增加，对 MTW 的孔隙接受/拒绝预测进行正确评估和建模可以在认知架构中为 MTW 骑手和周围车辆提供更安全的驾驶环境，增强微观模拟的可靠性和可预测性模型并改善整体交通流现象。因此，目前的研究调查了 Raff 方法的适用性，二元 logit 模型和支持向量机 (SVM) 在预测城市中段过滤期间 MTW 车手的孔隙接受决策中。研究结果表明，SVM 技术可以被视为一种潜在的工具，用于估计 MTW 车手的孔隙接受和拒绝预测，可以在协作智能交通系统环境中进一步实施，以实现整体安全而顺畅的交通流。