Biped robots have better mobility than conventional wheeled robots, but they tend to tip over easily. Thus, in this paper, we generate a stable walking pattern and propose a real-time posture control method.
The walking pattern can be created by a hip trajectory and an ankle trajectory. We generated a foot trajectory by 3rd order spline interpolation using two parameters - the maximum height of the ankle joint and the step length. And we derived the hip trajectory that makes the robot walk stably by fitting an actual ZMP trajectory of a biped robot to the target ZMP trajectory. With the walking pattern, we accomplished stable walking on even surface.
We use two tilt sensors to get the inclination of the ground - one for front to rear inclination, the other for left to right inclination. Adjusting the appropriate joint angle of the robot according to the sensor information, we made robot to control its posture stable to the inclination to all directions of the ground.