1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135

39

  1. When multiple RVs are used, independence or near-independence of the RVs can be

useful.

RV interdependencies can be complex (e.g. non-smooth), making it more

difficult to find parametric perturbations which cause only smooth variations over a

wide range of RV values. For our bipedal control, we assume near-independent RVs.

IMAGE Imgs/thesis.final.w651.gif

IMAGE Imgs/thesis.final.w652.gif

IMAGE Imgs/thesis.final.w653.gif

(a)

(b)


centre of mass

(c)

IMAGE Imgs/thesis.final.w654.gif

Figure 3.12 - Balance RV vectors
(a)swing-COM vector
(b)stance-COM vector
(c)up vector

In this thesis, we experiment with three choices of RVs, based on the vectors shown in Figure

3.12.

The first, the up-vector, is based on the notion of torso "uprightness".

The up-vector is

fixed to and runs along the length of the torso in the human model and the head in the robo-bird

model. The swing-centre of mass (swing-COM) vectordescribes the position of the COM of the

biped with respect to the current swing foot.

The stance-centre of mass (stance-COM) vector

indicates the position of the COM with respect to the stance foot.

The sampling time for all three

types of RV are at the end of states S3 and S6. For the purpose of computing RVs, the swing and

stance legs do not exchange until after the last base PCG state of the step.

The leg which is the

swing leg for most of the current step is used to compute the swing-COM RV.

RV is treated in a similar fashion.

The stance-COM

[CONVERTED BY MYRMIDON]