Predecessor speed prediction-based predictive cruise control of connected autonomous vehicle in platoon with multiple-human-driven-vehicles

Abstract

This paper investigates the connected cruise control issue for the connected autonomous vehicle (CAV) in a mixed platoon consisting of a CAV, consecutive human-driven vehicles (HDVs), and a connected HDV considering mixed traffic of human-autonomous vehicles from the low penetration of autonomous vehicles. To deal with the uncertainty of traffic flow and vehicle platooning when the uncertain number of HDVs enter or leave the mixed platoon, a CAV predictive cruise control strategy based on the prediction of the preceding vehicle’s speed is designed for the CAV with the help of the information of connected vehicles ahead and Signal Phase and Timing (SPaT) information through vehicle-to-everything (V2X) communication. A stochastic speed prediction method combining a conditional linear Gaussian speed prediction model and a backpropagation neural network is proposed, which improves the prediction accuracy of the future speed of the predecessor vehicle. The CAV’s target speed is planned based on the network information so that the CAV can pass the intersection without stopping. The fuel efficiency driving problem is transformed into the target speed tracking problem, and the optimal solution is carried out in the model predictive control (MPC) framework, which improves fuel economy while ensuring safety. Compared with existing other schemes verify the effectiveness and advantage of the designed strategy.