Design, analysis and passive balance control of a 7-DOF biped robot

Abstract

Biped robots have many advantages than traditional wheeled or tracked robots. They have better mobility in rough terrain and can travel on discontinuous path. The legs can also provide an active suspension that decouples the path of the trunk from the paths of the feet. Furthermore, the legs are able to step over considerably bigger obstacles compared to wheeled robots. However, it is difficult to maintain the balance of biped robots because they can easily tip over or slide down. To be able to walk stably, it is necessary for the robot to walk through a proper trajectory, which is the goal of this research. In this research, a complete 7-DOF biped walking trajectory is planned based on human walking trajectory by cubic Hermite interpolation method. The kinematics and dynamic model of the biped are derived by Denavit-Hartenberg (D-H) representation and Euler-Lagrange motion equations, respectively. The zero moment point of the robot is simulated to check the stability of the walking trajectory. The setpoint sampling method and sampling rate for trajectory tracking control are investigated by studying sinusoidal curve tracking on a single link robot arm. Two control sampling time selection methods are introduced for digital controllers. A 7-DOF biped is designed and built for experiments. Each joint has its own independent microcontroller-based control system. PD controllers are used to control the biped joints. Simulations are performed for the walking trajectory and zero moment point. Simulation results show that the walking trajectory is stable for the 7-DOF biped. Experiment results indicate that the sampling time is proper and the PID controller works well in both setpoint control and trajectory tracking. The experiment for the marching in place shows the trajectory is stable and the biped can balance during the marching process.