Abstract:
The main focus of this thesis is, design of an adaptive controller to control the speed of an elastic
system with Backlash such as Automotive Powertrain using sampled data techniques. Backlash
occurs in mechanical transmission links due to the clearance gap in gears, when the driving gear
is not exactly connected with the driven gear. It causes problem in drivability of the vehicles and
also motion lost during the braking, acceleration and deceleration phenomenon.
Lot of techniques is present in literature to tackle the problem of backlash in transmission links.
This thesis deals with two techniques. First one is the model-reference-adaptive-control (MRAS)
also known as model-reference-adaptive-system and the second one is self tuning regulator.
Automotive Powertrain will be used in this thesis for implementation of the Elastic System with
Backlash.
The Automotive Powertrain model used as the elastic system with backlash is a nonlinear model
that swings between two linear modes, the backlash-mode and the contact-mode. When the
driving gears are exactly connected to the driven gears, the system is in Contact-mode and
whenever the driving gears are not exactly connected to the driven gears, the system is in
Backlash-mode.
In model reference adaptive control system two separate MRAS controllers and two reference
models for the respective Contact and Backlash modes will be used. In Self Tuning Regulator the
parameters of the system will be estimated through switched Kalman Filter and using these
estimated states a state feedback controller will be developed to control the speed of Automotive
Powertrain. At the end Sampled-Data-System techniques will be used for Digital implementation
of above said controllers.
