TR2014-107

Governor-based Control for Rack-Wheel Coordination in Mechanically Decoupled Steering Systems


    •  Zafeiropoulos, S.; Di Cairano, S., "Governor-based Control for Rack-Wheel Coordination in Mechanically Decoupled Steering Systems", IEEE Conference on Decision and Control (CDC), DOI: 10.1109/CDC.2014.7040025, ISBN: 978-1-4799-7746-8, December 2014, pp. 4089-4094.
      BibTeX Download PDF
      • @inproceedings{Zafeiropoulos2014dec,
      • author = {Zafeiropoulos, S. and {Di Cairano}, S.},
      • title = {Governor-based Control for Rack-Wheel Coordination in Mechanically Decoupled Steering Systems},
      • booktitle = {IEEE Conference on Decision and Control (CDC)},
      • year = 2014,
      • pages = {4089--4094},
      • month = dec,
      • publisher = {IEEE},
      • doi = {10.1109/CDC.2014.7040025},
      • isbn = {978-1-4799-7746-8},
      • url = {http://www.merl.com/publications/TR2014-107}
      • }
  • MERL Contact:
  • Research Areas:

    Advanced Control Systems, Dynamical Systems, Mechatronics


A mechanically decoupled steering system enables autonomous or semi-autonomous vehicle steering by independently actuating the vehicle wheels and the steering wheel. In semi-autonomous operation the steering system should be controlled such that the vehicle wheels angle tracks a reference signal provided by the trajectory planner rapidly and safely, while guaranteeing that a certain alignment is maintained between the steering wheel and the vehicle wheels to avoid loss of "driver's panic". We develop a controller for a mechanically decoupled steering system that can achieve this by coordinating the steering column and the steering rack actuators, while enforcing constraints on the motion of the vehicle wheels, on the interaction between the steering wheel with the driver, and on the relative motion between steering wheel and vehicle wheels. Our design is based on a particular command governor, for which convergence is proven. The control strategy is simulated in closed loop with a detailed simulation model.