Abstract
The EMG signal is affected by air humidity, sweat, and many other factors, therefore, it is not stable for long time usage. To cope with this shortage, EMG based prosthetic hands available in the market are usually controlled by the on/off strategy. However, this strategy is not sophisticated for controlling prosthetic hands with multi-DOFs. Kinematics information is frequently used in controlling lower limb prostheses instead of upper limb ones because of the object-oriented feature of the motion of the upper limb. In this paper, we propose a control strategy which identifies the fetching motion of the user from his/her shoulder joint. An experiment is conducted to show the stability and easy-use of the proposed strategy in comparison with the on/off strategy.
| Original language | English |
|---|---|
| Pages (from-to) | 667-675 |
| Number of pages | 9 |
| Journal | Zidonghua Xuebao/Acta Automatica Sinica |
| Volume | 44 |
| Issue number | 4 |
| DOIs | |
| Publication status | Published - Apr 2018 |
Keywords
- Myoelectric signal processing
- Pattern recognition
- Prosthesis control
- Shoulder joint angle