Eco-speed optimization model for active rear-end collision avoidance of connected and automated vehicles on freeways

Abstract

This paper proposed an eco-speed optimization model for active rear-end collision avoidance of connected and automated vehicles on freeways in the scene of dangerous sources downstream, and the optimal speed curve was obtained to guide connected and automated vehicles to travel or stop safely and smoothly to avoid rear-end collision accidents. To implement the eco-speed optimization model, a traffic state prediction model based on confidence interval was proposed for forecasting the state of the front low-speed traffic flow. To improve the computational efficiency, the original optimal control problem was decomposed into several sub-problems based on the hyperbolic tangent function and only the leading vehicle was optimized. To eliminate the deviation caused by traffic variation during the period of solving the eco-speed optimization model, a prediction-based optimization strategy was adopted. To verify the performances of the eco-speed optimization model, the control strategy of in-vehicle driving alert systems was taken as the study case for comparison. The results show that the approximate model can basically ensure the computational accuracy and significantly improve the computational efficiency, that the proposed traffic state prediction model can ensure traffic safety, and that the speed control strategy obtained in this paper can make the vehicle decelerate in advance to avoid rear-end collision, smooth the vehicle’s speed curve, save the fuel consumption by 24.08%, restrain traffic oscillation and improve the safety of the traffic flow.