Walking motion planning method for biped walking robot

Abstract

The invention provides a walking motion planning method for a biped walking robot. The walking motion planning method takes an omnidirectional moving trolley as a model, takes a condition that the trolley does not overturn as a dynamics constraint condition, considers the dynamics constraints in footprint transform, uses a three-dimensional linear inverted pendulum model between partial discrete footprints, generates a hip trajectory online, and obtains driving joint trajectories through inverse kinematics; and finally, the walking motion planning method adopts a multi-rigid-body dynamics model, adopts a ZMP theory to determine stability, verifies the effectiveness of the joint trajectories, optimizes parameters of the omnidirectional moving trolley and the three-dimensional linear inverted pendulum model, and finally realizes biped walking robot walking motion planning constrained by dynamics under a fast walking condition.

Description

technical field [0001] The invention belongs to the field of motion planning of a biped walking robot, in particular to a walking motion planning method for a biped walking robot. Background technique [0002] The existing biped walking motion planning is difficult to balance stability and efficiency, mainly because walking motion is far more complicated than wheeled movement, and there are more than a dozen related degrees of freedom. In order to prevent the robot from falling, it must strictly comply with its dynamic constraints. Fast walking usually has poor stability and is prone to falls; while walking with high stability requires sacrificing walking speed. [0003] The current research on walking motion planning mainly focuses on gait planning and footprint planning. Gait planning is walking motion planning without environmental constraints, and footprint planning is walking motion planning considering the environment. [0004] At present, the relatively mature gait ...

AI Technical Summary

Problems solved by technology

[0007] The above method only considers dynamic constraints in the stage of gait planning to obtain joint trajectories, and does not consider the connection constraints between strides in the footprint planning stage, so it is possible to plan a footprint plan that is very unreasonable for the dynamic characteristics of the robot.

Following such a footprint plan may generate joint trajectories that violate stability constraints and eventually fall

The existing method to solve this problem is to reduce the walking speed. When walking at a slow speed, the probability of falling will be greatly reduced. Even if the gait planning is not reasonable, it is not easy to fall. can achieve the purpose

But this method did not realize the potential of the robot, the walking speed was very slow

Images