Research And Simulation On Walking Methods For Humanoid Robot In Variety Of Environments

Humanoid robot research is a hot research field. Much outstanding achievement has come out in recent years. However, humanoid robot becomes the utility for human still has a long way to go. Humanoid robot can not walking flexibility and stability in the complex environment where is human living is considered to an important reason. In this paper, a humanoid robot with 24 dof which is independent development by the Intelligent Computer Lab of South China University of Technology is discussed. Gait planning methods are proposed for the corresponding environments like plain, slope and stair through the deeply researched in humanoid robot mechanism theory, mathematical models and gait stability. The main working achievements and innovation are as follows:A design scheme of humanoid robot is given according to the proportion of the human body structure. In order to simulated the movement of human, the robot is required to define the range of motion of each joint refer to human being. According to the humanoid robot walking mechanism, a simplified model for walking is development by ignoring the lateral movement, and then a mathematical model on the forward plane is established.Planning the key points for the humanoid robot walking on the ground, upslope and upstairs based on principles of mechanical stability. Fitted a smooth curve of the joint trajectories using cubic spline interpolation function, and calculated each joint trajectory using inverse kinematics equations. Then, using the kinetic model to calculate the ZMP trajectory of humanoid robot, and judging whether the ZMP trajectory in the stable region to find the stable walking trajectory.Used OpenGL and 3DS MAX to explore a real-time three-dimensional simulation system for humanoid robot, and the realize simulation on slopes, stairs and many other environment by simulating real-world force, collision and other physical constraints. Finally, tested the walking algorithm of human robot on grounds, on slopes and on stairs through the simulation system. The experimental results approved the viability and stability of the gait planning algorithm.