A torque-based nonlinear predictive control approach of automotive powertrain by iterative optimization was developed in this paper. Simultaneously, the control scheme of torque demand is proposed to meet the torque requirement of drivers. From the experimental results and analysis, an NPC (nonlinear predictive control) controller like adaptive PID can be used to control the automotive powertrain directly by torque in the test bench. By use of the iterative optimization, the NPC algorithm can be extended from a two-step to a multi-step prediction of system states instead of increasing the online computation, which makes it possible to realize the application of the multi-step NPC in the real-time control of engineering practice. Through the experimental analysis, it can be concluded that the dynamic performance of the PI observer needs to be improved due to the lack of the differential. In summary, the estimated load can track the actual load during the dynamic operation of the powertrain, and the tracking performance of the steady state is good. After validation in the test bench, the NPC approach will be applied to a real vehicle in the next step. Simultaneously, except for the NPC approach, other control methods can also be used to implement the torque-based control of powertrain, e.g. adaptive PID and other ordinary optimization. Comparative experiments between NPC and other control methods hence should be conducted in the future.
A torque-based nonlinear predictive control approach of automotive powertrain by iterative optimization
2012
10 Seiten, 6 Bilder, 20 Quellen
Aufsatz (Zeitschrift)
Englisch
SAGE Publications | 2012
|