Automatic locomotion mode control of wheel-legged robots
Ilkka Leppänen
- Year
- 2007
- Citations
- 9
- Access
- Open access
Abstract
Traditional manipulator-type robots are mostly mounted in a fixed position and no locomotion is needed. During the past two decades, robots have slowly moved from factories to more open areas, such as metro stations, hospitals or mines. These mobile robots, called field and service robots, have to move and act on partly closed working sites or in ordinary environments, therefore the significance of their locomotion has increased. The main land locomotion principles, wheeled and legged locomotion, can be combined as hybrid to help find an optimal solution for greatly varying ground conditions. Hybrid locomotion can also be called rolking (rolling-walking). The wheel-legged robot can move using different locomotion modes; it can drive with wheels, rolk (use wheels and legs at the same time) or even walk.\n\nIn this work, an automatic locomotion mode control for the wheel-legged robot has been developed. The developed locomotion mode control can utilize two locomotion modes, wheeled and rolking, depending on terrain properties. The developed generic solution is valid for wheeled locomotion systems with a 2-dof active suspension system for the wheel.\n\nThe basis for the automatic locomotion mode control is to measure characteristics of vehicle-terrain interaction, such as energy consumption, slipping and drawbar force of a wheel and terrain parameters, such as terrain slope and roughness. Algorithms for sensing these in real time, utilizing the locomotion system as a sensing system, have been developed. The characteristics of vehicle-terrain interaction and terrain parameters are used as criteria for indicating the functioning of a wheel in wheeled mode and terrain trafficability in rolking mode. These criteria are used to determine if a mode change is required. The criteria are mainly based on sensing forces and energy consumption by utilizing the sensor system used for the locomotion control; a minimum number of additional sensors is needed.\n\nThe automatic locomotion mode control was successfully implemented and verified in the wheel-legged service robot, WorkPartner. It can use rolking mode on soft terrain or in negotiating obstacles and use wheeled mode for moving faster on flat terrain automatically. In the future, the developed sensing methods can also be used in developing new locomotion modes, such as free gait-based moving in which the motion control switches a single wheel to a different mode, depending on terrain properties.
Keywords
Related papers
Statistical Learning Theory
Yuhai Wu, Vladimir Vapnik
1999
Artificial intelligence: a modern approach
1995
Applied Nonlinear Control
Jean-Jacques Slotine, Weiping Li
1991
A new optimizer using particle swarm theory
R.C. Eberhart, James Kennedy
2002