|
|
Robot Control in Terms of Hamiltonian Mechanics
Záda, V. ; Belda, Květoslav
The paper deals with a mathematical modeling of robot motion and control. Instead of frequently used Lagrangian formulation of robot dynamics, this paper presents robot dynamics by Hamiltonian formulation. This formulation leads to different physical descriptive quantities considered for control design. In the paper, as a comparative control approach, PD control with gravity compensation is considered. The control approach considering Hamiltonian formulation is demonstrated for simplicity on two-mass robot-arm system. However, the explained modeling approach is general and it can be applied, e.g., to usual industrial articulated robots-manipulators with multiple degrees of freedom.
|
|
|
Control and synchronization drive unit of leg of walking robot
Houška, P. ; Březina, Tomáš ; Singule, V.
The contribution explores several approaches in control and synchronization of leg drives. It further deals with the selection of suitable trajectory for the movement of leg end point. Kinematics of walking robot leg movement is essentially influenced by used drive units and its control. The speed of leg endpoint relocation and amount of energy consumed in particular movements depends mainly on the drive control. Leg control is proposed in such way that the leg can be considered as independent system.
|
|
|
Using Q-learning for four-legged robot walking control
Březina, Tomáš ; Houška, P. ; Singule, V.
Significant goal in walking robot design is realization of autonomous system which is capable of motion in unknown environment. One of the possibilities how to reach necessary adaptability of control system without modeling of complex and unpredictable cases of system behaviour is the application of machine learning. From various learning paradigms the Q-learning is particularly attractive.
|
|
| |
guest ::
