Brussee, A.J. (2004) Pro-active trajectory formation using a biomimetic and a biomechanical model. Master's Thesis / Essay, Artificial Intelligence.
|
Text
msc-ai-thesis-alex-brussee-2004.pdf - Published Version Download (781kB) | Preview |
Abstract
This MSc thesis describes biomechanically inspired models for ego-motion trajectory formation in robotics. The first model uses direct muscle control as a model for ballistic trajectory formation in ego-motion. The second model is based on optimality constraints on known limb movement found in nature [23, 41]. In other words, two models for jump and catch in 2D are introduced. The notion of an internal motor command model for trajectory formation is based upon findings in humans and animals [35, 26, 41]. Traditional AI relies on reactive models. In such models, trajectory formation is heavily dependent on its surrounding environment [42, 37]. Biomechanical systems often possess many reactive components, but these components offer no complete explanation for their trajectory formation. To explain these differences between theory and reality, the two models investigate possible explanations that may give us a better understanding of the rules underlying trajectory formation in biomechanical systems. The two models differ in their perspective, but both try to relate findings in biomechanical systems to these rules. The direct muscle control model models the human muscle and emulates neural motor commands to form a trajectory. The second model develops a theory using constraints and rules that simulates empirically found trajectory formation. The trajectory from the direct muscle control model is transformed to steer a differentially controlled robot. This path is then fitted to the trajectory generated with the empirical model. The resulting trajectory can thus be used in pro-active robotic behaviors. The model was tested for relevant trajectories. It generated good results for trajectories with a low curvature.
| Item Type: | Thesis (Master's Thesis / Essay) |
|---|---|
| Degree programme: | Artificial Intelligence |
| Thesis type: | Master's Thesis / Essay |
| Language: | English |
| Date Deposited: | 15 Feb 2018 07:29 |
| Last Modified: | 15 Feb 2018 07:29 |
| URI: | https://fse.studenttheses.ub.rug.nl/id/eprint/8665 |
Actions (login required)
 |
View Item |