Modeling Of Mechatronic Systems: Cosimulation

Mechatronic systems in engineering consist of mechanical, control, hydraulic, pneumatic subsystems and so on. It is difficult to find a single software to model all these subsystems.LMS Virtual.Lab Motion can build 3D mechanical model and carry on kinematic and dynamic analysis. AMESim is a 1D simulation software developed by Imagine which is a company in France. There are a lot of libraries in AMESim and it can simulate the combustion process. By cosimulation between Virtual.Lab Motion and AMESim, we can model the complex mechatronic systems. The feasibility of cosimulation is studied in this thesis for the drivability simulation in the future.The thesis starts from a simple inverted pendulum. The 3D inverted pendulum model is built in Virtual.Lab Motion. We use two methods to control the inverted pendulum: (1) by the PID controller in Virtual.Lab Motion; (2) by the PID controller in AMESim. In the cosimulation process, Virtual.Lab Motion provides the angle between pendulum and plumb line and AMESim provides the force applied on the vehicle. After many experiments, the proper PID parameters are selected to stabilize the inverted pendulum system. The results between Vitual.Lab Motion simulation and cosimulation are compared. The cosimulation result is believable and the cosimulation time is longer than the Virtual.Lab Motion simulation time.A cosimulation engine model is built based on the 3D engine model in Virtual.Lab Motion and AMESim six cylinder engine model. Virtual.Lab Motion provides the position and velocity of each piston and AMESim calculates the combustion force on each piston. The cosimulation model is verified by driving experience and engine movement principle. The cosimulation model is correct.Finally, the future work on cosimulation is discussed. On one hand, we can increase the complexity of the model. For example, we can mount the engine model on the full vehicle model to get a better result in the drivability simulation. Also, we can increase the complexity of the interface. For example, we can provide the throttle angle by Virtual.Lab Motion to simulate the acceleration and deceleration. On the other hand, there is a trade off between simulation time and fidelity. The simulation time can be shorted by improving the mechanism of the data communication and the algorithm of the solver. Meanwhile, the simulation fidelity in engineering should be guaranteed.