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

85

torso rotation causes the biped to rotate about its own vertical axis more quickly than it can move

along a circular path. As a result, it turns to face the centre of the circle.


In the F-L and L-F sampling trials, the highest turning rates eventually cause the biped to turn

around completely and walk almost straight backwards. The turning perturbation is ineffective in

the backward walking cases because the model walks primarily on its heels.

never obtains sufficient ground reaction forces to fix the foot's orientation.

As a result, the foot

IMAGE Imgs/thesis.final.w6170.gif

IMAGE Imgs/thesis.final.w6171.gif

(a) - F-L sampling

(b) - L-F sampling

IMAGE Imgs/thesis.final.w6172.gif

(c) - SP sampling

Figure 5.7- Hip plots for the most successful
over a range of scaling factors. Q
dturning perturbation trials
= [.25,0] for all trials.

Not all values of Qdwork equally well as Figure 5.8demonstrates.


Linearly


varying


the

perturbation scaling factor still yields smooth behaviour, but not the desired results of Figure 5.7.

While this result makes it clear that some care must be taken in choosing parameters, it also serves

to illustrate the fact that reasonable choices can often be made without exploring a large parameter

[CONVERTED BY MYRMIDON]