*4.2.1. Control method of Saga university prosthetic arm*

Saga University prosthetic arm is developed for the realization of 5DoF upper limb motions for a transhumeral amputee. The hand is controlled using a combination of an EMG based controller (EBC) and a task oriented kinematic based controller (KBC). Figure 17 shows an experimental setup of an EMG signal based controller of a Saga prosthetic arm. In a trans‐ humeral amputee a part of the biceps and triceps are remaining. EMG signals of the amputee's biceps and triceps are used as the input information for the EBC to control elbow flexion/ extension and hand grasp/release. Forearm supination/pronation, wrist flexion/extension and ulnar/radial deviation get controlled from the KBC. Motion intention of the amputee is identified via a task classifier using shoulder and prosthesis elbow kinematics. For the scope of this context only the EBC will be considered. EMG based fuzzy controller is the base for EBC. It proportionally controls the torque of the elbow and hand actuator according to the amount of the EMG signal. The activation of biceps generates the elbow flexion and the activation of triceps generates the elbow extension. Hand grasp is realized when both triceps and biceps are activated simultaneously. The release position of the hand is achieved when the both muscles are not working.

**Figure 17.** Experimental setup of EMG signals based Controller of Saga prosthetic arm [22]

of the exoskeleton (hand force and forearm torque) robot [35] are used as input information to the controller. This fuzzy-neuro control method consists of a flexible fuzzy control and adaptive neural network control which is used to obtain natural and flexible motion assist. Fuzzy if-then rules have been constructed to determine the required torque to the motor according to the motion intention of the human. In total, nine fuzzy-neuro controllers are used and this allows operating the exoskeleton robot flexible with EMG signals. Depending on the subject and nature of power-assistance, training of the fuzzy-neuro controller is performed.

The main drawback of this kind of control method is the difficulty of defining the fuzzy if-then control rules when the controller is applied for exoskeleton robots with higher DoF. Further, training of the controller is essential even when the physical and psychological conditions

This section reviews recent EMG based controllers of upper-limb prostheses. Controlling of a prosthetic device using EMG signals is cumbersome task compared to controlling an exoskel‐ eton device using EMG signals, since already the person who wears the device has lost the best site of the body to get the required EMG signals for the controlling of the prosthesis. It makes the number of inputs for the control system to be less and obviously causes to under‐ perform a conventional control system. So far researchers [21-23] have proposed different control methods to control prosthetic devices. In addition, the introduction of more advanced technologies such as targeted muscle reinnervation and implantable electrodes marks new boundaries in prosthesis controlling. The ensuing subsections review EMG based control methods of upper-limb prostheses. The logic for selecting a particular control method is its

Saga University prosthetic arm is developed for the realization of 5DoF upper limb motions for a transhumeral amputee. The hand is controlled using a combination of an EMG based controller (EBC) and a task oriented kinematic based controller (KBC). Figure 17 shows an experimental setup of an EMG signal based controller of a Saga prosthetic arm. In a trans‐ humeral amputee a part of the biceps and triceps are remaining. EMG signals of the amputee's biceps and triceps are used as the input information for the EBC to control elbow flexion/ extension and hand grasp/release. Forearm supination/pronation, wrist flexion/extension and ulnar/radial deviation get controlled from the KBC. Motion intention of the amputee is identified via a task classifier using shoulder and prosthesis elbow kinematics. For the scope of this context only the EBC will be considered. EMG based fuzzy controller is the base for EBC. It proportionally controls the torque of the elbow and hand actuator according to the amount of the EMG signal. The activation of biceps generates the elbow flexion and the activation of triceps generates the elbow extension. Hand grasp is realized when both triceps and biceps are activated simultaneously. The release position of the hand is achieved when

**4.2. EMG based control methods of upper-limb prostheses**

*4.2.1. Control method of Saga university prosthetic arm*

change in the user.

256 Electrodiagnosis in New Frontiers of Clinical Research

significance and novelty.

the both muscles are not working.

Information from the raw EMG signals is extracted by taking the RMS value of the raw EMG signals. The RMS is determined as,

$$RMS = \sqrt{1 / N \sum\_{i=1}^{N} v\_i^2} \tag{3}$$

where *vi* is the voltage value at *i th* sampling and *N* is the number of samples in a segment.

The EBC is provided with four EMG RMSs. Three kinds of fuzzy linguistic variables for each input were defined. Ten kinds of fuzzy if-then rules for elbow joint torque control and seven kinds of if-then rules for hand torque control were prepared. In addition, details on the KBC can be found in [22].

Even though the prosthetic arm is capable of catering to the human motion intension to a certain degree, still some improvements can be made. Since the KBC is trained to offline Vicon data for a given set of daily activities, the orientations of the limbs other than those trained for cannot be achieved using the KBC. Therefore, an inertial measurement unit (IMU) can be fixed to the stem arm of the amputee and used as an interface to read the real time kinematic data. By using the real time kinematic data as an input to the controller it can be improved to reach almost all the orientations of daily activities. In addition a hybrid control method - EMG coupled with EEG - can be used to enhance the performance of the controller by obtaining correct human motion intention.
