Connected and automated vehicles (CAVs) can communicate with other CAVs through vehicle-to-vehicle (V2V) communication technology which can improve the stability and safety performance of the platoon. However, the V2V communication could be unreliable as some factors such as hardware damage, packet loss, and cyber-attack, may lead to communication delay or failure. Considering that the communication topology of the platoon can vary dynamically due to these factors, it may result in instability and collision risk on the platoon. To mitigate the impact of communication delay or failure, the study introduces a dynamic weights optimization-based CAV following model (DWOC). It adopts the multi-predecessor following (MPF) mode with which vehicles can obtain the dynamic information of three predecessors and the corresponding information weights can be adjusted adaptively. The study formulates the string stability conditions of the model and validates the stability and safety impact on the platoon through a series of numerical experiments. The results indicate that DWOC can effectively improve the stability and safety performance under different communication delays. The findings suggest that the CAV manufacturers should account for the possible changes of the communication topology of the platoon. The proposed DWOC can serve as a reference in designing the CAV controller to ensure stability and improve safety of the platoon in reality under different communication delays.

The quality of vehicular collision data is crucial for studying the relationship between injury severity and collision factors. Misclassified injury severity data in the crash dataset, however, may cause inaccurate parameter estimates and consequently lead to biased conclusions and poorly designed countermeasures. This is particularly true for imbalanced data where the number of samples in one class far outnumber the other. To improve the classification performance of the injury severity, the paper presents a robust noise filtering technique to deal with the mislabels in the imbalanced crash dataset using the advanced machine learning algorithms. We examine the state-of-the-art filtering algorithms, including Iterative Noise Filtering based on the Fusion of Classifiers (INFFC), Iterative Partitioning Filter (IPF), and Saturation Filter (SatF). In the case study of Cairo (Egypt), the empirical results show that: (1) the mislabels in crash data significantly influence the injury severity predictions, and (2) the proposed M-IPF filter outperforms its counterparts in terms of the effectiveness and efficiency in eliminating the mislabels in crash data. The test results demonstrate the efficacy of the M-IPF in handling the data noise and mitigating the impacts thereof.