2020/11/18 by Gurcan Comert, Comert, Gurcan, Negash Begashaw +1
Engineering · Social Sciences · #Applications (stat.AP) #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #FOS: Electrical engineering #Machine Learning (stat.ML) #Robotics (cs.RO) #Systems and Control (eess.SY) #Traffic control and management #Transportation Planning and Optimization #Vehicle emissions and performance #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2011.09370
openalex publication_date 2020/11/18 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
Estimation models from connected vehicles often assume low level parameters\nsuch as arrival rates and market penetration rates as known or estimate them in\nreal-time. At low market penetration rates, such parameter estimators produce\nlarge errors making estimated queue lengths inefficient for control or\noperations applications. In order to improve accuracy of low level parameter\nestimations, this study investigates the impact of connected vehicles\ninformation filtering on queue length estimation models. Filters are used as\nmultilevel real-time estimators. Accuracy is tested against known arrival rate\nand market penetration rate scenarios using microsimulations. To understand the\neffectiveness for short-term or for dynamic processes, arrival rates, and\nmarket penetration rates are changed every 15 minutes. The results show that\nwith Kalman and Particle filters, parameter estimators are able to find the\ntrue values within 15 minutes and meet and surpass the accuracy of known\nparameter scenarios especially for low market penetration rates. In addition,\nusing last known estimated queue lengths when no connected vehicle is present\nperforms better than inputting average estimated values. Moreover, the study\nshows that both filtering algorithms are suitable for real-time applications\nthat require less than 0.1 second computational time.\n