作者: Jih-Gau Juang
DOI: 10.1109/3477.865178
关键词:
摘要: In this paper, a learning scheme using fuzzy controller to generate walking gaits is developed. The uses combined with linearized inverse biped model. provides the control signals at each time instant. algorithm used train "backpropagation through time". model error for backpropagation instants. Given prespecified constraints such as step length, crossing clearance, and speed, can gait that satisfies these constraints. Simulation results are reported five-link robot.