Dasher

Humanoid robotics : Dasher

In the humanoid robotic scenario, MDH is performing research on biped locomotion working on a test case of a running machine, and performing experiments on control parameters optimizations using the Dasher Running Robot (www.dasher.se). The robot is bio inspired, hydraulic actuated, and provides intrinsic compliance in the ankles joints due to Össur elastic prosthesis mounted behind the knee joints. Linear extensible actuators coupled with intrinsic compliant lower limbs provide a non-conventional biped locomotion platform able to exploit the adaptive stiffness concept proper of humans’ manoeuvres while running and jogging

Dasher

The dasher robot is a human sized 7 Dof humanoid robot hydraulic actuated. Each leg consists of two hydraulic cylinders fitted inside the thigh. Each cylinder is equipped with a potentiometer connected to be co-axial that measures the instant cylinder extension. The oil flow to each cylinder is controlled by MOOG 30 series servo proportional valves. The knee joints house the robot feet, two “Flex-Foot Cheetah” prosthesis from Össur, especially designed to generate propulsion for a 40 kg body that is the robot overall weight. Strain sensors are used to measure the load on the ankle joints and to retrieve estimate the ground reaction forces.
The hydraulic system consists of a brushless DC motor that runs a pump to maintain 100 Bars hydraulic pressure within the entire system. A reservoir filled with nitrogen gas works as a pressure capacitor and maintains a constant pressure eliminating fluctuations. The proportional valves and the cylinders and are tightly assembled inside the legs to lower the robot centre of gravity and to reduce the trapped oil between the valves and the cylinders improving the performance of the control loop and eliminating the death band and spring effect created by oil trapped between the valves and the cylinders chambers.
The global stability sensors are placed inside the head. Two MT9P03 5 Mega-Pixel CMOS image sensors send image information to an also FPGA located inside of the head. The FPGA uses Stephen and Harris combined corner and edge detection to find interests points and edges to be used as global reference. Also a three axial inertial sensor is used and located inside the head. The sensor values (X Y gyros and inclinations) are processed into the FPGA and sent to the main controller board located in the torso via CAN interface. Also the CAN bus is used to send the position and torques set points to the low-level control boards located on each joint.
The arms are designed to swing according to a stabilizing strategy during the robot running motion. They are driven by two electrical motors, which are located on each shoulder. The arms cards control the angular position and speed. The angular rotations are measured through optical encoders.

Comments are closed.