Design And Realization Of A Baby-mimic Crawling Robot With Flexible Spine

The carrying capacity, velocity and stability of traditional legged robots are constrained greatly because that most of them has only rigid components. Research shows that it can simplify the dynamic motion control and improve the walking stability and coordination of robots when adding compliant components to a rigid robot.In this paper, a baby-mimic quadruped robot with a global flexible spine of variable cross-section stiffness was proposed aiming at solving the problem of unstable body posture when the quadruped robot is walking. The locomotion control method based on biological neural control mechanism was employed and the dynamic simulations were carried out. An experimental platform was set up and the walking experiments and performance evaluations of the baby-mimic quadruped robot were conducted. This paper is organized as follows.A baby-mimic quadruped robot mechanism scheme with6active degrees of freedom was proposed.A global flexible spine with variable cross-section stiffness and an arc elastic leg were designed. The mechanical parameters of the flexible spine and the arc elastic leg were calculated using pseudo-rigid-body-model (PRBM).A center pattern generator(CPG) model based on biological neural control mechanism was built as the motion controller for the baby-mimic quadruped robot.The virtual prototype of the baby-mimic quadruped robot was built and comparative simulations between the baby-mimic quadruped robot with a spine of bigger yaw flexibility and a rigid spine was accomplished.The prototype of the baby-mimic quadruped robot and its control system were constructed. The comparative walking experiments between the rigid-spine baby-mimic quadruped robot and compliant-spine quadruped robot were carried on and the influence of the flexible spine to the walking performance of the robot in the direction of roll and yaw was evaluated.The result shows that adding the baby-mimic quadruped robot with a global flexible spine of a variable cross-section stiffness has a positive effect on reducing the swing of the body posture (head). Thereby, it is effective for improving walking posture stability and walking trajectory accuracy for legged robots.