An Innovative Framework to Evaluate the Performance of Connected Vehicle Applications: From the Perspective of Speed Variation-Based Entropy (SVE)

Recently, a significant number of Connected Vehicle (CV) applications have emerged to offer solutions to some of the problems caused by increasing traffic demands. To better understand the influence of different CV applications on traffic in a relatively unified manner, a novel measure of effectiveness (MOE) called speed variation-based entropy (SVE) is proposed. An analytical study was carried out on different types of CV applications by relating SVE with other conventional MOEs on safety, mobility, and environmental sustainability. Four applications–High Speed Differential Warning (HSDW), Lane Speed Monitoring (LSM), Eco-Speed Harmonization (ESH), and Eco-Approach and Departure (EAD), representing safety-, mobility- and environment-focused CV applications–were selected for detailed evaluation. Results from the sensitivity analysis on technology penetration rate and congestion level reveal that: 1) SVE can accurately represent the speed variation of individual vehicles and the overall traffic; 2) the proposed SVE distribution can be used as an MOE for CV applications in a more holistic way and at different scales; and 3) the overall SVE has a strong positive correlation with conventional MOEs (i.e., average conflict frequency and average fuel consumption) especially under freeway scenarios. Therefore, conventional MOEs can be estimated and explained by SVE under a variety of scenarios.