Research On Gait Planning And Walking Control For Humanoid Robot

Humanoid robot is one of the most representative of today’s robotics research, which combines multi-discipline expertise and a number of high-tech in one, brought together electrical, computer, materials, sensors, control technology and artificial intelligence, and other cutting-edge technology disciplines. Humanoid robot to complete a basic human behavior class action, it must be dynamically stable bipedal walking, walking on the basis of stable dynamic stability margin is planning a large gait; due to the humanoid robot is a high degree of freedom, change structure, strong coupling of complex nonlinear systems, so walking humanoid robot controlled study has important theoretical significance and practical value. In this paper, a humanoid robot walking gait planning and control for a more in-depth study, research contents are as follows:(1)Online gait planning based on a pre-humanoid robot control conceptSince calculating for large time-consuming offline gait planning shortcomings, we propose a ZMP (Zero Moment Point) concept of online pre-planning gait control method based on the dynamics of humanoid robots analyze expand the table-car models the first method according to the given gait parameters, using cubic spline interpolation function of the ankle joint trajectory planning, which then plan out the desired ZMP trajectory of the robot in the case of a given initial state, based on the extended table-car models pre-view controller uses information and future expectations ZMP of the table-error output values and expectations and feedback ZMP robot car models get centroid track status, so hip centroid trajectory tracking, finally solving robot inverse kinematics equations can be obtained each joint angle trajectories; once again pre-constructed concept of ZMP controller, using estimates of future information ZMP error between the desired ZMP trajectory of hip correction, achieve the purpose of improving stability. This method can calculate in advance the parameters needed to calculate the amount of its online is not without loop optimizations can get a stable gait, so you can plan based on real-time feedback information online gait.(2)Use of ankle angle correction walk with arms swinging comprehensive control Because of ankle and arm joints relevance small can decoupling control, so both separately controlled. For ankle angle correction, inverted pendulum model by analyzing the dynamics equation, ankle joint torque, angle, the relationship among the ZMP position to construct a linear quadratic optimal controller to correct the ankle joint angle trajectories, thus ZMP moves to control purposes; for swing arm joint control, given the parameters of the first arms swinging tracks, swing generates mathematical relationship between torque and angle of the swing arms according to completely offset the yaw moment as the goal, seeking inverse trigonometric angle trajectory obtained arms swinging. In walking humanoid robot control, in addition to supporting leg ankle angle trajectory correction, but also to control the swing angle of trajectory tracking arms generated. Simulation results show that this method can guarantee good ZMP better tracking performance can offset the humanoid robot walking in yaw moment.At the end of this dissertation, the main research is summarized. It makes out the main innovations and research achievements, and also points out the problems and issues which need to further research.