The automotive industry has long relied on testing powertrain components in real vehicles, which causes the development process to be slow and expensive. Therefore, hardware in the loop (HIL) testing techniques are increasingly being adopted to develop electronic control units (ECU) for engine and other components of a vehicle.
In this thesis, HIL testing system is developed to provide a laboratory testing environment for continuously variable transmissions (CVTs). Two induction motors were utilized to emulate a real engine and vehicle. The engine and vehicle models, running in real-time, provide reference torque and speed signals for input and output dynamometers, respectively. To design torque and speed tracking controllers, linear models of the motor and drive systems were firstly identified from the test results. Feedforward controllers were then designed according to the inverse dynamics of the identified models. Because of the existence of unstable zeros in the model, design effort was focused on the stability and causality of the inverse process. Digital preview filters were formulated to approximate the stable inverse of unstable zeros as part of the feedforward controller. Normally, future information of input trajectory is required when implementing the digital preview filters, which makes the feedforward controller non-causal. Since the engine and vehicle model require current information to calculate the next output and no future value can be provided in advance, the application of non-causal digital controllers was limited. A novel method is proposed here to apply non-causal digital controllers causally. Robustness of the controllers is also considered when the two motors are coupled and the gear ratio between them was changed.
The proposed coupled control method was tested and verified experimentally by using a manual gearbox before recommending its use for a CVT testing. A multifrequency
test signal as well as simulation results of a whole vehicle model were used as torque and speed demand signals in the experiments. A HIL testing case was also presented. Frequency and time domain results showed the effectiveness of the method under both testing procedures to fully compensate for the dynamics of both actuators.
|Date of Award||1 Dec 2009|
|Supervisor||Mehmet Sahinkaya (Supervisor)|
Causal Tracking Control of a Non-Minimum Phase HIL Transmission Test System
Wang, P. (Author). 1 Dec 2009
Student thesis: Doctoral Thesis › PhD