2003 Fiscal Year Final Research Report Summary
Design Principles of Autonomous Robots based on Coevolutionary Dynamics.
| Project/Area Number |
14580418
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | Nagoya University |
|---|
Principal Investigator |
AKITA Takaya Nagoya University, Graduate School of Information Science, Professor, 情報科学研究科, 教授 (40202759)
|
|---|
| Project Period (FY) |
2002 – 2003
|
| Keywords | Coevolution / Artificial Life / Evolutionary Computation / Swarm Intelligence / Constructive Approach / Autonomous Robots / Emergence / Complex Systems |
|---|
| Research Abstract |
The field of evolutionary computation based on the algorithms inspired from biological evolution has been achieving the remarkable results recently. Most of the algorithms adopted in previous successful researches use static fitness functions. However, natural evolution is not so simple. Fitness of individuals depends on the complex interactions with the individuals in the same population or other populations. Furthermore, regarding the level of an individual, contribution of a trait to fitness depends on dynamic interactions with other traits. This study focused on such kind of "coevolution" in a broad sense and investigated the dynamics of it. At the same time, We aimed to apply the mechanisms and the mechanisms of the coevolution to design autonomous robot systems. First, we constructed 3 minimal models regarding coevolutionary phenomenon. Next, we examined coevolution of morphology and control of robots in order to construct design methodoloogy of autonomous robots. Then we investigated emergence of swarm intelligence, emotions in robots, and adaptivity of misperception in order to grasp the mechanism of emergence in the robot systems based on coevolution.
|
