Balance of a Robotic Leg on Uneven Terrain

Bipedal balancing is a complex subject as the process of balancing is difficult to define. As robot designs are becoming increasingly humanoid and human-robot interactions increasingly common, algorithms must be developed to replicate natural movement in non-ideal circumstances. This project implements such an algorithm on a robotic leg to attempt to balance on uneven surfaces and terrain. A pseudoinverse Jacobian was implemented to handle positional movement, and a linear quadratic regulator was used to shape smooth motion to desired balance points. They were then developed into an Arduino code which controls the motion of the robot in attempts at balance. Due to the limited stall torque of the chosen servos the robot could not achieve balance in practice. And although the position and control codes work well with the chosen microcontroller, the balance equations could be more versatile and accurate, so future work should focus on these aspects of the robot design process