Sign In
Sign in
English

Walking

w1.png
Walking is the essential requirement of a humanoid robot. We have been exploring how the robot can walk stably and how we can control the robot. The walking algorithm of the HUBO series consists of two categories: walking pattern generation and walking control

■ Walking Pattern Generation

1) Parametric Methods

 Walking pattern could be genrated empirically by simple parametic functions.

w2.png

 

2) Analytic Solution Methods

 Using analytic solution of an stability crieteria function, walking pattern
    (COM of the robot ) could be generated.

w3.png

■ Walking Control

The walking control algorithm of the HUBO series consists of three categories: the Walking Pattern Control (WPC) , the Real-Time Balance Control (RTBC) and Predictive Motion Control (PMC).

 

1) Walking Pattern Control (WPC)

 

This control scheme changes walking pattern gradually according to the behavior  of the robot and the load conditions during walking.

 

2) Real-Time Balance Control (RTBC)

 

This control scheme are necessary for the conservation of the walking stability in real-time.

 

3) Predictive Motion Control (PMC)

 

The predicted motion control is based on the expected movements of the robot, and it is needed to prevent abnormal movements. This is because it is too late to stabilize the robot after occurrence of abnormal movements.

w4.png t_1.png