The calculation of the optimal trajectory of the stepped top-down robot was made using a genetic algorithm and a computational torque controller. First, the total energy efficiency was minimized using the Red-Cold Generic Algorithm (RCGA) consisting of reproductive, cross, and mutation. The reproducibility condition related to the position assembly of the start and end of the stride and the joints, angles, and angular velocities are linear constraints. Next, the unequal constraint accompanies the condition for preventing the collision of the swing leg at the corner with the outer surface of the stairs, the condition of the knee joint for preventing kinematic peculiarity, and the condition of no moment in safety in the traveling direction. Finally, the angular trajectory of each joint is defined by fourth-order polynomial whose coefficient is to approximate chromosomes. This is to approximate walking. In this study, the energy efficiency of the optimal trajectory was analyzed by computer simulation through a biped robot with seven degrees of freedom composed of seven links.

ABSTRACT
 1. 서 론
 2. 이족보행로봇 모델
 3. 에너지 최적화 보행궤적의 연구(유전 알고리즘 이용)
  3.1 에너지 최적 보행궤적을 찾기 위한 유전 알고리즘
  3.2 등가 구속 조건(Equality constraints)
  3.3 부등가 구속 조건(Inequality constraints)
  3.4 최적화 방법
 4. 시뮬레이션
  4.1 계단의 상향 경우
  4.2 계단의 하향 경우
  4.3 에너지 효율의 비교
 5. 결 론
 References

• 이정익(INHA Technical College, Dept. of Mechanical Design) | Jeong-Ick Lee Corresponding Author