Rear-end crashes are among the most common crash types at signalized intersections. To examine the risk factors for the occurrence of this crash type, this study involved the analysis of nine years of intersection cra...Rear-end crashes are among the most common crash types at signalized intersections. To examine the risk factors for the occurrence of this crash type, this study involved the analysis of nine years of intersection crash records in the state of Wyoming. With that, the contributing factors related to crash, driver, environmental, and roadway characteristics, including pavement surface friction, were investigated. A binomial logistic regression modeling approach was applied to achieve the study’s objective. The results showed that three factors related to crash and driver’s attributes (commercial vehicle involvement, speeding, and driver’s age) and four factors related to environmental and roadway characteristics (lighting, weather conditions, area type, whether urban or rural and pavement friction) are associated with the risk of rear-end crash occurrence at signalized intersections. This study provides insights into the mitigation measures to implement concerning rear-end crashes at signalized intersections.展开更多
深入分析交通事故数据可以为规避事故发生、降低事故严重程度提供重要理论依据,然而,在事故数据采集、传输、存储过程中往往会产生数据缺失,导致统计分析结果的准确性下降、模型的误判风险上升。本文以芝加哥2016—2021年的101452条追...深入分析交通事故数据可以为规避事故发生、降低事故严重程度提供重要理论依据,然而,在事故数据采集、传输、存储过程中往往会产生数据缺失,导致统计分析结果的准确性下降、模型的误判风险上升。本文以芝加哥2016—2021年的101452条追尾事故数据为研究对象,将原始数据按照7∶3随机分为训练集和测试集。在训练集数据上,利用生成式插补网络(Generative Adversarial Imputation Network,GAIN)实现对缺失数据的填补。为对比不同数据填补方法的效果,同时选择多重插补(Multiple Imputation by Chained Equations,MICE)算法、期望最大化(Expectation Maximization,EM)填充算法、缺失森林(MissForest)算法和K最近邻(K-Nearest Neighbor,KNN)算法对同一数据集进行数据填补,并基于填补前后变量方差变化比较不同填补算法对数据变异性的影响。在完成数据填补的基础上,构建LightGBM三分类事故严重程度影响因素分析模型。使用原始训练集数据,以及填补后的训练集数据分别训练模型,并使用未经填补的测试集数据检验模型预测效果。结果表明,经缺失值填补后,模型性能得到一定改善,使用GAIN填补数据集训练的模型,相较于原始数据训练的模型,准确率提高了6.84%,F1提高了4.61%,AUC(Area Under the Curve)提高了10.09%,且改善效果优于其他4种填补方法。展开更多
传统的部署MATALB Web App的方法是使用MATLAB App Designer设计应用的界面和功能,利用MATLAB Web App Server将设计好的应用打包部署到Web端,但是使用这种方式在设计时存在功能扩展复杂的问题,在部署时存在应用程序加载缓慢,部分浏览...传统的部署MATALB Web App的方法是使用MATLAB App Designer设计应用的界面和功能,利用MATLAB Web App Server将设计好的应用打包部署到Web端,但是使用这种方式在设计时存在功能扩展复杂的问题,在部署时存在应用程序加载缓慢,部分浏览器版本不兼容等问题,降低了处理效率与使用体验。为了改善以上情况,提出利用HTML(Hyper Text Markup Language)与Vue设计前端应用界面和后端连接的RESTful API(Representational State Transfer Application Programming Interface),然后用Python构建后端应用接口用于函数计算,再使用Nginx将前端界面部署到Web端,实现一种远程部署MATLAB应用的新方法。网页端FIR(Finite Impulse Response)低通与高通滤波器设计的测试结果表明,上述方法与MATLAB生成的滤波器一致,部署简单且高效,能够较好解决上述问题,同时为MATLAB Web App的托管与共享方式提供了新思路。展开更多
文摘Rear-end crashes are among the most common crash types at signalized intersections. To examine the risk factors for the occurrence of this crash type, this study involved the analysis of nine years of intersection crash records in the state of Wyoming. With that, the contributing factors related to crash, driver, environmental, and roadway characteristics, including pavement surface friction, were investigated. A binomial logistic regression modeling approach was applied to achieve the study’s objective. The results showed that three factors related to crash and driver’s attributes (commercial vehicle involvement, speeding, and driver’s age) and four factors related to environmental and roadway characteristics (lighting, weather conditions, area type, whether urban or rural and pavement friction) are associated with the risk of rear-end crash occurrence at signalized intersections. This study provides insights into the mitigation measures to implement concerning rear-end crashes at signalized intersections.
文摘深入分析交通事故数据可以为规避事故发生、降低事故严重程度提供重要理论依据,然而,在事故数据采集、传输、存储过程中往往会产生数据缺失,导致统计分析结果的准确性下降、模型的误判风险上升。本文以芝加哥2016—2021年的101452条追尾事故数据为研究对象,将原始数据按照7∶3随机分为训练集和测试集。在训练集数据上,利用生成式插补网络(Generative Adversarial Imputation Network,GAIN)实现对缺失数据的填补。为对比不同数据填补方法的效果,同时选择多重插补(Multiple Imputation by Chained Equations,MICE)算法、期望最大化(Expectation Maximization,EM)填充算法、缺失森林(MissForest)算法和K最近邻(K-Nearest Neighbor,KNN)算法对同一数据集进行数据填补,并基于填补前后变量方差变化比较不同填补算法对数据变异性的影响。在完成数据填补的基础上,构建LightGBM三分类事故严重程度影响因素分析模型。使用原始训练集数据,以及填补后的训练集数据分别训练模型,并使用未经填补的测试集数据检验模型预测效果。结果表明,经缺失值填补后,模型性能得到一定改善,使用GAIN填补数据集训练的模型,相较于原始数据训练的模型,准确率提高了6.84%,F1提高了4.61%,AUC(Area Under the Curve)提高了10.09%,且改善效果优于其他4种填补方法。
文摘传统的部署MATALB Web App的方法是使用MATLAB App Designer设计应用的界面和功能,利用MATLAB Web App Server将设计好的应用打包部署到Web端,但是使用这种方式在设计时存在功能扩展复杂的问题,在部署时存在应用程序加载缓慢,部分浏览器版本不兼容等问题,降低了处理效率与使用体验。为了改善以上情况,提出利用HTML(Hyper Text Markup Language)与Vue设计前端应用界面和后端连接的RESTful API(Representational State Transfer Application Programming Interface),然后用Python构建后端应用接口用于函数计算,再使用Nginx将前端界面部署到Web端,实现一种远程部署MATLAB应用的新方法。网页端FIR(Finite Impulse Response)低通与高通滤波器设计的测试结果表明,上述方法与MATLAB生成的滤波器一致,部署简单且高效,能够较好解决上述问题,同时为MATLAB Web App的托管与共享方式提供了新思路。
