抄録
In the present paper, coordinating functions among antagonistic pairs of the mono- and the bi-articular muscle in the human upper extremity were analyzed in terms of electromyographic (EMG) kinesiology, and control properties of a two-joint rink model equipped with an antagonistic pair of bi-articular muscles were analyzed experimentally as well as theoretically. EMG studies: Subjects employed were 5 healthy young male adults, and muscles tested were deltoid anterior (Da: f_1) and posterior (Ds: e_1) portions, brachialis (Br: f_2), biceps brachii long head (Blo: f_3), and triceps brachii lateral (Tla: e_2) and long (Tlo: e_3) heads. EMGs were recorded during isometric arm push and pull movements in all directions (360°) with maximal effort in the sagittal plane. All round force directions were divided into six ranges with three crossing lines at the wrist joint (W). Directions b and e were pushing up and pulling down along the forearm, and directions c and f, pulling up and pushing down, respectively, parallel with the upper arm. Directions a and d were passing through shoulder joint (S) and point W. All subjects showed almost the same EMG patterns, where the Blo and the Tlo reversed their activity levels in the ranges between a and b and between d and e, the Da and the Ds, in the ranges between f and a and between c and d, and the Br and the Tla, in the ranges between b and c and between e and f. Thus, two pairs of the antagonistic mono-articular muscles as well as the pair of bi-articular muscles showed crisscross activity patterns in each pair of the opposing ranges. In the other ranges than the opposing pair of ranges where the criss-cross patterns appeared, one muscle of the antagonistic pair of mono- as well as the bi-articular muscles showed full activity level and the other antagonist of the pair showed almost nothing. This essential activity pattern never changed even though angular distributions of the ranges of force directions changed widely with the postural changes. The model analyses: 1) Based on the EMG results, coordinating patterns of contractile forces exerted at the muscles incorporated on the model were postulated as follows: uf_1+ue_1=100%, uf_2+ue_2=100%, uf_3+ue_3=100%. Distributions of the maximum forces developed at the end point of the model were very similar to the one obtained in the human experiments. 2) The shape and inclination of the stiffness ellipse exerted at the end point of the model can be controlled independently in the model with the bi-articular muscles, but they cannot be controlled independently without the bi-articular muscles. Effects of forces applied at the end point of the model from outside were examined theoretically as well as experimentally. The end point can be moved along the direction of the outside force in the model with the bi-articular muscles, but the end point moved in the shifted direction away from the outside force in the one without the bi-articular muscles. A robot with these properties could control its position, force and stiffness without a complex and hard robot controller, and these properties will be useful to an assembly robot, human-friendly robot, orthotics or prosthetics.
