Research On Eco-driving Control Strategy Of Connected Automoted Vehicle At Signalized Intersections

Aiming at the eco-driving problems of vehicles at signalized intersections,a research targeting on connected and autonomous vehicles(CAV)is presented.The research ia carried out in various aspects,including the analysis of energy-oriented optimal control law,optimization of green wave speed at multiple signalized intersections,safety and energy-saving control at single signalized intersection,and the construction of test platform for CAV.This paper focuses on the energy-oriented optimization of driving speed at signalized intersections in order to improve traffic efficiency and reduce energy consumption of electric vehicles.The specific research content is as follows:Firstly,the powertrain model of EV incorporating vehicle longitudinal dynamics,energy efficiency model of in-wheel motor(IWM)as well as battery dynamics is established.Two energy consumption models are analyzed which are ‘wheel-mileage’ model and ‘batterymileage’ model.The energy-saving potential of EV speed control and the control law of optimal speed under constrained control inputs are analyzed from the perspective of ‘wheel-mileage’model.Secondly,based on the ‘battery-mileage’ model,the energy-oriented passing strategy and speed optimization at multiple signalized intersections are studied.Taking road slope and signal phase and timings(SPa T)of traffic lights into consideration,the road model and traffic model are constructed respectively.The green wave passing strategy at multiple signalized intersections based on feasible window planning method is presented.On that basis,dynamic programming(DP)is ultilized to solve the multi-objective optimization problem concering energy consumption as well as battery aging,so as to realize the efficient and energy-saving passing of vehicles at intersections.The preceding control strategy puts particular emphasis on green-wave speed passing,without carefuly considering the constraints of vehicle passing at single signalized intersection,especially when influenced by the traffic flow ahead.Thus,a controller aiming at safety and energy consumption is designed using model predictive control(MPC).An optimal analytical solution is derived using pontryagin’s minimum principle(PMP)and the indirect adjacency method is ultilized to handle the inequality constraints to achieve real-time speed optimization under the safe distance constraint,safe speed constraint and signal phase constraint.Finally,the test platform for CAV is built and the experimental validation of green wave passing strategy at multiple signalized intersections is carried out.The CAV involves three aspects: low-level hardware modification,high-level software design and the processing and analysis of the data from sensors.The low-level hardware is modified on the basis of the original chassis structure.Robot operating system(ROS)is used for distributed development of highlevel software design.Sensor data parsing and processing includes combined navigation systems,millimeter-wave radar and on-board equipments for V2 X,etc.