Model based control design and integration of automotive cyber-physical systems
Shang, Di
:
2013-04-01
Abstract
Cyber-physical systems, such as automotive vehicles, are very challenging to design because of the tight interactions between physical dynamics, computational dynamics and communication networks. In addition, the evaluation of these systems in the early design stages is very crucial and challenging. Model-based design approaches have been applied in order to manage the complexities due to interactions. In this thesis, a study to demonstrate the systematic design, analysis and evaluation of an integrated automotive control system is presented. In detail, following the model based design method, a vehicle lane keeping controller is designed and implemented on a time triggered platform which is running on real time Linux. Then the control design is verified by both of Simulink simulation and hardware-in-the-loop simulation. After that, a vehicle integrated control system is built by the integration of two independently designed controllers, lane keeping controller (LKC) and adaptive cruise controller (ACC). During the integration, the interaction and conflict between different controllers are analyzed then a supervisor control is designed to coordinate LKC and ACC under the situation when there is a conflict. Simulink simulations are performed to proof the necessity of supervisor controller. Finally, the integrated system is deployed on a hardware-in-the-loop simulator for evaluation under realistic scenarios. By the HIL simulation result, the control effects introduced by the real time platform and communication network are also observed. The efficiency of the approach is also present experimental results that demonstrate.