Gait Planning Of Legged Robot Based On Central Pattern Generator

Legged robot is a important manner which explore motion mechanism of human and animal, has simulate natural walking of human and animal being completely. The dissertation using biped robot and quadruped robot as a study platform, takes advantage of CPG (central pattern generator) which generate curve as gait of biped robot and quadruped.Following the description of the significance, contents and background of biped robot, quadruped robot, and leg walk robot, and description of the make use of CPG apply to this field. This dissertation gives detailed introduction of the system composition of biped robot and quadruped, and then focuses on research of CPG gait planning.Due to complex modeling, non-unique solutions and single-period planning, the traditional model-based motion-controlling approach is disadvantage to produce quick steady motion and to adapt to unstructured environment for a leg walk robot. The dissertation uses CPG to control gait of robot. Based on of nerve cell, the oscillating model is established, by which the CPG control network of CPG of the biped robot and the quadruped robot is developed. The impact on the output by parameters of CPG model, and the CPG control network parameters of the robot is set by use GA (genetic algorithm) to optimize.Finally, we validate that using control network of CPG can generate gait of biped robot through simulation, and CPG model has apply to our biped robot. The gait of walk and trot of quadruped robot has planned by CPG through simulation.