Optimal Control Strategies for Legged Locomotion
Guimaraes, Fernando Pereira
Legged locomotion belongs to the category of Non-
linear hybrid systems, which makes it particularly in-
teresting. To adequately design a controller for such
systems, requirements as ability to ensure integrity, ar-
ticulation of long time and short time horizons, Scala-
bility in time and space, Coordinated decentralization
of the decision and control system, Adaptivity and Ro-
bustness arise. Optimal control results have been ex-
plored, once they can provide the required control syn-
thesis. This led to the selection of a Model Predictive
Control scheme. After presenting the theoretical back-
ground, use-cases with different degrees of complexity
have been formulated and implemented.
The results show that optimal control techniques are
very powerful, but also tricky and time consuming in
case of higher complexity.