• 제어로봇시스템학회논문지
• /
• 제16권2호
• /
• pp.107-113
• /
• 2010

In this paper, a prototype of cooperative work model for multi-robots system is introduced and the evolutionary approach is applied to generate the motions for the cooperative works of multi-robots system using genetic algorithm. The cooperative tasks can be performed by a humanoid robot and a mobile robot to deliver objects from shelves. Generation of the humanoid motions such as pick up, rotation, and place operation for the cooperative works are evolved. Modeling and computer simulation for the cooperative robots system are executed in Webots environments. Experimental results show the feasible and reasonable solutions for humanoid cooperative tasks are obtained.

• 전기학회논문지
• /
• 제65권9호
• /
• pp.1551-1556
• /
• 2016

A combined CPG (Central Pattern Generator) based foot trajectory and GP (Genetic Programming) based joint compensation method is presented for the adaptive humanoid walking. The CPG based foot trajectory methods have been successfully applied to basic slops and variable slops with slow rates, but have a limitation for the steep slop terrains. In order to increase an adaptability of humanoid walking for the rough terrains, a GP based joint compensation method is proposed and combined to the CPG (Central Pattern Generator) based foot trajectory method. The experiments using humanoid robot Nao are conducted in an ODE based Webots simulation environmemt to verify a stability of walking for the various aslope terrains. The proposed method is compared to the previous CPG foot trajectory technique and shows better performances especially for the steep varied slopes.

• 한국지능시스템학회논문지
• /
• 제23권4호
• /
• pp.360-365
• /
• 2013

본 연구에서는 발끝 궤적을 미리 설계하지 않고, CPG(Central Pattern Generator)를 이용하여 동적으로 생성할 수 있는 기법을 제안한다. 생성된 발끝 궤적은 CPG 의 진동적인 출력에 따라 가변적인데, 이는 발끝 궤적이 CPG 진동적인 출력 신호의 맵핑 함수로 주어지기 때문이다. 이를 통해 환경에 적응적인 궤적을 생성할 수 있는 토대를 마련할 수 있다. 제안된 기법의 효율성을 검증하기 위해서, Webots 시뮬레이션을 통해 휴머노이드 로봇 Nao에 대한 실험을 수행하고, 성능과 동작 특성을 분석한다.