Simulation-aided walking pattern generation for a humanoid robotic platform

Bipedal locomotion is one of the main research topics in humanoid robotics due to the complexity of generating dynamically stable gait patterns in robots that are generally very unstable. Good walking performance requires diverse and complex control strategies based on a mathematical model of the robot. Here, a methodology to easily generate stable bipedal locomotion patterns in a humanoid robotic platform is presented. Using a simulation environment and a simple programming method, the robot gait, stability and general behaviour are first simulated and then transferred to the actual machine. Three gait patterns are developed according to the biomechanics of human walking and validated using a humanoid robot platform. The robot exhibits good performance in terms of stability, displacement and velocity. The proposed methodology allows the study of locomotion in biped robots and can serve as a basis for developing controls in closed-loop schemes.