Search

—This paper proposes two real-time energy-oriented driving strategies to minimize the energy consumption for electric vehicles on highways with varying slopes. First, a novel strategy, called normalized-energy consumption minimization strategy (NCMS), adopts a designed kinetic energy conversion factor to convert the vehicle kinetic energy change into the equivalent battery energy consumption. By minimizing the total normalized energy consumption, the energy-orientated vehicle control sequence is calculated. In addition, a logic car-following algorithm is developed to enhance NCMS for avoiding collisions with the potential preceding vehicle on the journey. Second, an improved model predictive control (MPC) is developed with a hierarchical framework, which achieves a balance between optimization and computational efficiency. In the upper level, a global, coarse-grained, iterative dynamic programming is employed to penalize the MPC terminal state, while the lower level performs online rolling optimization of the vehicle within a moderate time step. Thirdly, the performance of the proposed driving strategies is verified through a traffic simulation to evaluate the energy efficiency improvement and processor computation time compared to dynamic programming and constant speed strategy. Finally, a vehicle-in-the-loop test is carried out to validate the feasibility of the proposed two novel driving strategies. • A normalized energy consumption model is presented, combining two energy usage. • A novel eco-driving strategy is developed based on the proposed energy consumption. • The algorithm suits both free traffic and car-following scenarios with safety. • The characteristics of the four eco-driving strategies are compared. • The proposed strategy is verified by both simulation and vehicle-in-the-loop tests. [ABSTRACT FROM AUTHOR]

The emergence of the intelligent transportation system and cloud computing technology has brought the available traffic information and increasing computing power, which lead to a significant improvement in driving performance. In order to enhance energy economy and mobility simultaneously, a hierarchical eco-driving strategy is proposed in this paper, which is comprised of the cloud-level controller and the vehicle-level controller. The dynamic programming-based cloud-level controller optimizes the velocity and battery state-of-charge utilizing the global traffic information obtained from the intelligent transportation system. However, the global traffic information suffers from uncertainties, which deteriorates the effectiveness of the cloud-level controller. The vehicle-level controller is constructed on the model predictive control framework, aiming to cope with the uncertainties, improve fuel economy and reduce travel time. Besides, a transfer learning-based particle swarm optimization algorithm is presented for solving the optimization problem in model predictive control, which can achieve great control performance utilizing the knowledge from the cloud-level controller. To validate the effectiveness of the proposed strategy, simulation tests are conducted. The results demonstrate that the proposed strategy can achieve near-global-optimal performance in fuel economy and mobility. Moreover, the real-time performance of the proposed strategy is validated through the hardware-in-loop test. • A hierarchical eco-driving strategy is designed using vehicle-to-cloud connectivity. • The uncertainties in traffic information are considered and tackled. • A novel transfer learning-based partial swarm optimization algorithm is proposed. • Coordinative optimization for vehicle velocity and power distribution. • The proposed strategy is validated in both simulation and hardware-in-loop test. [ABSTRACT FROM AUTHOR]

This paper proposes a real-time dynamic predictive cruise control (PCC) system to minimize the energy consumption for electric vehicles (EVs) under integrated traffic situations with synthetic driving scenarios, considering both constraints from the preceding vehicle and the influence of traffic signal lights. The proposed PCC system is working based on the bi -level model predictive control (MPC) algorithm. The Signal Phase and Timing (SPaT)-oriented MPC calculates a desired acceleration command as the optimal control signal at each sampling step based on the forthcoming SPaT information with the purpose of passing the nearest signalized intersection during the green light interval without stop. The car-following-oriented MPC executes preceding vehicle tracking task through maintaining a safe inter-distance using a customized variable time headway (VTH) strategy. The instantaneous energy consumption for EV in different traffic scenarios was quantified by a data-driven model. The developed system was validated through comparison with IDM and human driver's maneuver in both suburban and urban areas road in the city of Fukuoka, Japan, during off-peak and peak hours, using the real traffic system and SPaT data. To further evaluate the performance of the proposed PCC system in high speed driving situation, another case study with transitions from highway to urban road was conducted. The simulative results showed that the proposed PCC system can realize the energy-saving rates by 8.5%–15.6%. And it was working well and robustly under high speed driving situation. • A real-time PCC system was developed to minimize the energy consumption of EVs. • Two scenarios of car-following and signal anticipation were evaluated. • An ANN model was developed to predict the energy economy performance of the EV. • The proposed PCC can realize 8.5%–15.6% of the energy saving in the urban areas. [ABSTRACT FROM AUTHOR]