EMG sensor controlled five axis robotic arm and robotic hand design and implementation

Semi-autonomous or manual-controlled electromechanical equipments is usually controlled via control panels or through simple software. The first step for using these equipments's in harsh field conditions, such as military/medical applications, is the quick, simple and accurate retrieval of input signals from the operator. However, employment of trained technical staff or coding of the software leads to time and money loss. It is also very difficult to develop a software that adapts to all kinds of field conditions. It is important that these operations are done correctly in situations where the operator's has little or no training. For this purpose, the use of human's natural physical movements or neural signals as an input signal may be preferred as a technique. Within the scope of this thesis, as a first application, a five-axis robotic arm designed to provide the functionality of the human arm was designed and prototype was produced. For real-time control of the robot arm, the surface electromyography signals taken from six different muscles are used in the healthy individual's arm. 81.7% success was achieved in the experiments. In the second application, data from six different movements were obtained. Using the collected data, an artificial intelligence application was realized. When the support vector machine (SVM) classifier was used in the developed method, the accuracy rate was found to be 92.77%.