Plug-in hybrid electric buses feature long battery life and low comprehensive energy consumption,and have broad development prospects in the context of pure electric mileage anxiety.The energy economy of a hybrid power train depends to a large extent on the strengths and weaknesses of its control strategy.For a hybrid configuration like a coaxial series-parallel hybrid configuration,the rotational speeds of the power sources on the drive shaft are coupled.The vehicle energy management strategy can be turned into an optimal control problem of torque distribution.At the same time,in the context of the current complexity of automotive systems,the development process based on model design has become a trend.In this dissertation,the plug-in hybrid bus with coaxial hybrid configuration is taken as the research object.Combined with the characteristics of the system configuration,the optimal energy allocation problem of different power sources is studied.Firstly,the working mechanism of the power train and the parameters of real vehicle acquisition are expounded,the high-fidelity mathematical model of each component of the vehicle is established,and the operation mode of the plug-in coaxial hybrid configuration and the advantages of the hybrid system are analyzed.Furthermore,the forward simulation model for real veh;icle application is built,and the vehicle-oriented forward simulation architecture is completed.Based on the logic threshold and flow chart,the core control strategy of rule-based energy management is formulated.Embedded code based on code generation technology for rule-based energy management strategy is developed and vehicle test are conducted.Secondly,under the background of intelligent network such as car networking,the research significance of predictive control on vehicle energy management issues appears.This paper carries out the feasibility verification of predictive control in the development of energy management strategy.The complex coaxial series-parallel hybrid system model is approximated,and the continuous system model oriented to control is established by surface fitting and curve fitting.The nonlinear predictive control problem of hybrid system energy management is transformed into the optimal control based on the Pontryagin's minimum principle;the equations corresponding to the optimization necessary conditions are obtained through the Hamilton equation,and discrete processing is performed by numerical iterative method.The solution of the control variables is obtained by solving the equations.The control-optimized predictive control architecture is realized by establishing a simplified continuous model.It is proved that the model predictive control can control the nonlinear process while processing the variable constraints,and the feasibility of the hybrid control system energy management strategy based on predictive control is verified which lays the foundation for the development of energy management strategies based on nonlinear model predictive control with the potential of real-time application.Then,the nonlinear predictive energy management control strategy is implemented for a more accurate nonlinear discrete model of hybrid system.The torque prediction secant method is proposed to improve the torque exponential prediction method and used to predict the expected driving torque in predictive horizon.Considering the fixed bus line of the city and the strong statistical regularity,the battery based on the station mileage is proposed.The SOC prediction method is compared with the optimal calibration method of the reference SOC curve such as dynamic programming.Then the optimal control framework is completed according to the nonlinear discrete state space model and objective function of the controlled dynamic system,and the evolutionary algorithm is used as the numerical algorithm to solve the nonlinear programming problem to obtain the optimal control variables.The results show that the satisfying performance of the proposed NMPC strategy is partly due to the consideration of the predicted torque and the reference SOC curve,and the energy optimization potential of the hybrid vehicle is explored.The dissertation has a total of 60 pictures,6 tables,and 56 references. |