We introduced a new concept of controlling powered gear shifts in automatic transmissions for passenger cars. This work was motivated by the effort to reduce calibration procedures by circumventing the need for controller tuning. The new control concept is based on high-gain feedback control in conjunction with a λ-tracking adaptor that continuously computes the control parameters, i.e. the feedback gains. By doing so, no knowledge of the system parameters is needed and there is no need for calibration, since the controller tunes itself during operation. The applicability of the adaptive control principle to a nonlinear drivetrain system like the one at hand was first demonstrated with the help of simulating a complete vehicle and drivetrain model that used the actual control software where the adaptive controller had been implemented. The promising results of the simulation were then confirmed by testing the controller in an actual vehicle drivetrain. We have shown that a control method of the proposed kind can in fact be applied to a dnvetrain system of this sort. The following conclusions can be drawn: Adaptive λ-tracking control is feasible and seems stable for this application. Therefore, the parameters that have originally been used to calibrate the shift process during this phase are no longer necessary. Thus, the proposed control concept contributes to a simplification of the calibration efforts, although this effect is still rather small at this point. The results of the simulation and the actual vehicle tests are comparable. This validates the simulation environment. The parameter set chosen for the purpose of this paper lead to satisfactory controller performance. However, the control goal is not yet achieved in all situations. The adaption process during vehicle tests is somewhat delayed compared to those in simulations. The control gain is increased slightly after the error value has left the λ-strip. This behavior is due to computational delays caused by the rapid prototyping setup. We aim the principle of λ-tracking control at reducing the calibration efforts by incorporating self-tuning feedback controllers instead of open-loop control strategies in the transmission control software. We are exploring the ability of λ-tracking controllers to reduce the calibration efforts further. For instance, other phases of the shift process may be investigated to determine whether they are feasible for λ-tracking control. Also, if a full stability analysis can be provided, the λ-tracking controllers discussed here may be expanded to other nonlinear systems that share the same properties. Thus, the findings obtained during this work are not restricted to drivetrain systems.
A novel adaptive λ-tracking approach for automotive transmission control systems
2012
16 Seiten, 7 Bilder, 1 Tabelle, 12 Quellen
Conference paper
English
Automotive radar data filtering approach for adaptive cruise control systems
Tema Archive | 2008
|Adaptive automotive speed control
Tema Archive | 1993
|Automotive power transmission systems
SLUB | 2018
|Automotive power transmission systems
TIBKAT | 2018
|