Research On The Hierarchical Optimization Control Method For A Group Of Intelligent And Connected Hybrid Electric Vehicles

Energy saving,safety,comfort and environmental protection are the main themes of the automotive industry.Focusing on these themes,contemporarily,the most popular research areas include the intelligent vehicles,connected vehicles and electric vehicles.In this dissertation,the intelligent and connected hybrid electric vehicle(HEV)is chosen as the research object.Aiming at improving the fuel economy,mobility,comfort and safety of the vehicles,model predictive control(MPC)is employed as the main research method.A hierarchical predictive control architecture for a group of connected HEVs is proposed.In addition,the solution to achieve the real time control of the hierarchical control architecture is put forward and vehicle-in-the-loop experiment and hardware-inthe-loop simulation are performed.Furthermore,the fuel efficient control considering system errors,the management optimization based on the predicted information and the fuel efficient control incorporating vertical vibration are studied.The main work of this dissertation is summarized as follows:A hierarchical control architecture for intelligent and connected HEVs considering efficiencies feedback is studied.The control architecture consists of two levels.The higher level controller presents a MPC controller based on the fuel economy,relative distance,SPAT information and longitudinal acceleration to optimize the target velocities of a group of vehicles.The lower level controller focuses on the A-ECMS based energy management optimization to improve the fuel efficiency mainly with battery SOC sustainability constrained.Meanwhile,the average propulsion and recuperation efficiencies are periodically calculated and fed back to the higher level controller to correct the fuel consumption model associated with the efficiencies.The real-time velocity prediction and energy management optimization for the intelligent and connected HEVs based on the hierarchical control architecture are presented.Focusing on the real structure of the MPC optimization problem,the KKT function is solved by Newton method and approximation-based methods are used to solve the Newton iteration function,and therefore a F-MPC based method is achieved to predict the optimal velocities in real time.The lower level controller linearized the ECMS by designing WL-ECMS based on the Willans Line model of the power components to realize the real-time energy management control.A closed-loop coordinated control strategy is researched considering the random errors of the control system of the connected HEVs.Based on the hierarchical control architecture,the higher level controller considers the random errors of the control system and uses Markov decision chain to model the random errors and the probability transition matrices.The random errors are handled by SMPC and scenario tree is employed to simplify the SMPC problem to reduce the computational intensity.The lower level controller adopts A-ECMS for the energy management of the HEVs.A simplified HEV model is embedded in the higher level codes and the propulsion and recuperation efficiencies are modeled using measurable variables and the post-processing files in Autonomie Software,and then the efficiencies are fed back to the higher level controller to optimize the velocities of the next time step.The energy management optimization method of the connected 4WD vehicles based on the predicted information is presented.A rule-based control strategy for the 4WD HEV is introduced and then ECMS is designed based on rule-based strategy.According to the predicted velocity information,the equivalent factor is optimized both globally and segmentally.In addition,DP is employed for the global optimization of the energy management of the 4WD HEVs.A fuel efficient control strategy incorporating vertical vibration is studied.The longitudinal dynamics of the vehicle and the vertical dynamics of the suspension system are seamlessly integrated based on the characteristics that the state of the suspension is correlated to both the vehicle velocity and position.A MPC model that compromises fuel efficiency,traffic mobility,tracking distance,vertical ride comfort and longitudinal ride comfort is established based on HEV velocity and SPAT information obtained in the intelligent and connected vehicle environment,and thus the vertical ride comfort is improved while guaranteeing the advantages of the connectivity concept-based MPC and the fuel efficiency.In order to reduce the computational burden,F-MPC is employed to optimize the velocity profiles of the vehicles.