74

transition
info

0.1
0.1
0.1
0.1

DOF:

1

0
0
0
0

2:02:1

3

80
80
55
55

4

-95
-95
-65
-65

5

45
45
25
25

6:06:17:07:1

8

55
55
80
80

9

-65
-65
-95
-95

1011:0 11:1

S1
S2
S3
S4

0
0
0
0

0
0
15
15

-45
-45
-25
-25

0
0
0
0

0
0
0
0

15
15
0
0

25
25
45
45

-25
-25
-45
-45

0
0
0
0

state

Figure 4.17 - Robo-bird base PCG for running

Figure 4.18 - A running robo-bird

While both the robo-bird and human models are bipeds, their dynamics are different.

When

standing and walking, the human model is supported by the rigidity of its straight stance leg,

which typically has little or no freedom in the vertical direction.

In theory, no actuator forces are

required to prevent the human model from collapsing. The leg structure of the robo-bird model is

such that actuator forces must constantly be applied even to remain standing upright.

This means

there is a significant amount of vertical compliance when moving, leading to "bouncy" motions.

As well, the base of the robo-bird model is wide enough to make standing on one foot for any

length of time difficult.

for this model.

These factors make a dynamic running motion more natural than a walk

4. 6

Conclusions

In

this

chapter

we

have

presented

evidence

that

the

discrete

limit

cycle

control

technique

introduced in Chapter 3 can be used to generate balanced walking motions for the human model

and a running motion for the robo-bird model.

Two choices of RV, based on the up-vector and

the swing-COM vector, were successfully applied to this end.

A torso servo was shown to be

useful in improving the aesthetics of the human model's motion.

Finally, a running motion for