Design and implementation of a device to control a robotic arm by EMG signal

---

Tezin Türü: Yüksek Lisans

Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Elektrik ve Elektronik Mühendisliği Bölümü, Türkiye

Tezin Onay Tarihi: 2013

Öğrenci: GÖRKEM KANDEMİR

Danışman: BEHÇET MURAT EYÜBOĞLU

Özet:

In this study, an electromyogram (EMG)based human machine interface system is designed and implemented. System acquires EMG signals and processes them to generate commands to control a robotic arm. Different signal processing methodologies are investigated, realized, and compared. The thesis study includes design and implementation of an 8 channels electromyogram data acquisition system which is used to record raw EMG signals from operator’s muscles. The system transfers raw EMG data to control software via USB. Control software processes the raw EMG data accordingly and generates control commands which are used to control the robotic arm. The robotic arm “AL5D” and its actuator driver which are used in thesis are not designed in the scope of this study. They are commercial products which are purchased from Lynxmotion Company. There are five servo motors placed on the robotic arm to realize movements. Communication with the robotic arm is achieved via USB so that the system may be interfaced with any PC. In the near-real time implemented system, the operator is expected to contract his/her predefined muscles. Each muscle is used to control a specific servo on the robotic arm. After operator is trained to run the designed HMI (Human Machine Interface) system, he is able to drive the robotic arm with full functionality. Different time domain and frequency domain signal processing algorithms are investigated in this study.However, only time domain algorithms are implemented in the designed near-real time HMI system. On the other hand, frequency domain algorithms are studied in post-processing environment. Time domain algorithms are based on different time domain features of the raw EMG signal.All time domain features aresuccessfully processed to drive the corresponding servo on the robotic armwith 100% accuracy.