2005 Fiscal Year Final Research Report Summary
Study on Body Dynamics Formation and Pattern Generation for Motor/Action Learning
| Project/Area Number |
14350227
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Control engineering
|
|---|
| Research Institution | Tokyo Institute of Technology |
|---|
Principal Investigator |
ITO Koji Tokyo Institute of Technology, Department of Computational Intelligence and Systems Sciences, Professor, 大学院・総合理工学研究科, 教授 (30023310)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
MATSUNO Fumitoshi The University of Electro-Communications, Department of Mechanical Engineering and Intelligent Systems, Professor, 知能機械工学科, 教授 (00190489)
SHIBATA Katsunari Oita University, Faculty of Engineering, Associate Professor, 工学部, 助教授 (10260522)
KONDO Toshiyuki Tokyo Institute of Technology, Department of Computational Intelligence and Systems Sciences, Assistant Professor, 大学院・総合理工学研究科, 助手 (60323820)
|
|---|
| Project Period (FY) |
2002 – 2005
|
| Keywords | motor adaptation / dynamic environment / internal model / sensorimotor mapping / brain nervous system / 運動制御 |
|---|
| Research Abstract |
(1) Biological system has the ability of appropriately and rapidly constraining their sensorimotor mapping in real-time, even though they are situated in unknown environments. Based on the biological evidences, we proposed a computational motor adaptation model with a context-based environmental cognition, which is called CPG-CM (Central Pattern Generators with Constraints Modulation). It is basically consists of three components ; Brain nervous system, Body and Environments. Here, the body and environment correspond to the dynamics of musculo-skeletal system and external worlds, respectively. We built up the brain nervous system (RNN and CPG) which could generate an appropriate motion pattern based on a time-series of the proprioceptive feedbacks (i.e. context-based cognition). As an example, it was applied to a redundant manipulator control for a crank rotation. It was then demonstrated that the proposed model could recognize the change of external environment through the time-series
… More
observation of the system state variables and evoke CPG parameters to maintain the crank rotation.
(2) To manipulate objects or to use tools, humans must compensate for the resultant forces arising from interaction with the physical environment. Recent studies have shown that humans can acquire a neural representation of the relation between motor command and movement, i.e. learn an internal model of the environment dynamics. Then, we can compensate for the mechanical perturbation in a feedforward manner. We investigated whether humans could identify one side of dynamics from the mixed force field in the case where humans had experienced either of them. The experimental results suggested that the orthogonality of force vectors loaded to the hand played an important role to identify the environment dynamics. Though it is necessary to verify the various combinations of force fields, these results gave topics of much interest on the spatial representation in sensory area through the somatosensory feedback and the formation of body image. Less
|
