2018 年 56 巻 5 号 p. 198-208
Surface electrodes are commonly used for functional electrical stimulation (FES) as they are easily placed with low burden on the user's body. However, the development of a simple control method for human fingers, which have certain redundant degrees of freedom, is hindered by the difficulty of locally and selectively stimulating the target muscle. On the contrary, the central nervous system does not individually control each muscle by inverse kinematics, but simultaneously controls agonist-antagonist muscle groups. The control of muscular movements through specific control parameters—in this case, the stiffness and angle of each joint—is known as the equilibrium-point hypothesis of voluntary joint movements. This hypothesis suggests that finger joint movements, which are realized by the complex cooperation of many muscles, can be controlled by only a few parameters, without the need for discrete muscle movement commands. In this study, we placed surface electrodes on the forearm under an isometric condition, and measured the characteristics of finger joint movements by FES. The fingertip force output plotted against the electrical agonist-antagonist muscle ratio input was adequately estimated by a second order system in conjunction with the dead time. The identified parameters of the finger joint movements were compared with those of the elbow and ankle joints movements. The results suggested that the identified parameters were affected by both muscle size and the length ratio of the muscle belly to the muscle tendon. Although the current results are qualitative rather than quantitative, they validate the model by roughly providing the expected output. The proposed model may also control minute fingertip forces.