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

66

Figure 4.9 shows the nominal range plots and hip plots for a series of trials using swing-COM

RVs and stance hip control perturbations with each of the three sampling strategies (F-L, L-F and

SP). The trials explore the useful range of the forward component of Qd
, with uniform sampling.

The lateral component of Qdis fixed at 0.0 for all trials. Torso servoing is not applied.

are a maximum of 60 steps long.


All trials

A few of observations can be made based on Figure 4.9and on the corresponding animated

motion:

  1. The swing-COM RVs generate walking motions generally similar in nature to those

generated using up-vector RVs.

As with the up-vector results, there is a "nominal

range" of Qdvalues which produce the most successful walks.

  1. In contrast to the up vector case, different

sampling

strategies

give

significantly

different results. SP sampling generates the most stable motions and F-L sampling the

least stable.

  1. Unlike the up vector walks, which generally move straight forward, many swing-

COM walks demonstrate a tendency of the biped to twist sideways about its own

vertical axis.

This tendency is most prevalent when using the F-L sampling strategy

and is almost entirely absent for the SP sampling strategy.

[CONVERTED BY MYRMIDON]