International Journal of Engineering and Technology

In this paper, the arm rehabilitation device controller based on fuzzy logic techniques is presented. Patients who has post-stroke may lose control of their upper limb. If they are treated with functional rehabilitation training, the patients can rehabilitate their motion functions and working abilities. These rehabilitation devices are used to recover the movement of arm after stroke. Many controllers had been used for the rehabilitation device and one of them is ANFIS-PID controller where Adaptive Neuro-Fuzzy Inference System (ANFIS) technique is the combination of fuzzy logic and neural network system. The objectives of this project are to develop arm rehabilitation device controller based on the ANFIS-PID technique. The development of ANFIS is purposely as an inverse model to the system and proportional-integral-derivative (PID) controller as a feedback control. EMG model is integrated to the control system as reference where Artificial Neural Network (ANN) is used to model the EMG to position relationship. Simulation is conducted using MATLAB to validate the system performance that is integrated with EMG model. Then the performance is compared between ANFIS-PID controller and PID alone controller. ANFIS–PID controller reduced more tracking error compared to PID controller and demonstrates better results when disturbance is applied to the control system.