论文部分内容阅读
为研究营运客车换道安全性,确定营运客车换道越线时间,建立营运客车换道越线时间预测方法。利用图像识别传感器获取营运客车换道过程横向表征参数,基于6次多项式,拟合不同速度条件下的营运客车换道轨迹。同时采用数据库对比分析方法,建立营运客车换道越线时间预测模型,对比换道越线时间预测值与真实值。结果表明,采用6次多项式拟合的营运客车换道轨迹,拟合系数均在0.993以上。越线时间真实值和预测值一致率为62.5%,绝对误差均在±0.2 s以内,均值为-0.0175 s,且两者之间存在明显的正态分布关系。由预测模型得到的越线时间稍小于真实值。
In order to study the safety of lane changing for operating passenger cars, determine the cross-lane transit time of operating buses and establish the forecast method of cross-lane transit time of operating buses. The image recognition sensor is used to obtain the lateral characterization parameters of tram passenger car lane change process. Based on the 6th order polynomial, the change trajectory of operating passenger car under different speed conditions is fitted. In the meantime, the comparative analysis method of database was used to establish the forecast model of transit bus lane time for operating buses, and to compare the forecast value and the true value of lane transit time. The results show that fitting coefficients of operating passenger trains with 6 times polynomial fitting are above 0.993. The coincidence rate between the true value and the predicted value of cross-line time was 62.5%, the absolute error was within ± 0.2 s, the mean was -0.0175 s, and there was a significant normal distribution between the two. The lead time obtained from the prediction model is slightly less than the true value.