Electroencephalograph based control and feature extraction algorithms for automated powered exoskeleton

Abstract

Mobility and the posture are the major concern for people suffering from endo skeleton disorders. For such people, exoskeleton provides great support and protection. Mostly Exoskeletons are used for the rehabilitation purposes and can improve the life of an individual who has lost his body part and thus help in assistance. In this article, various technologies which are mainly used for the implementation and performance of exoskeleton are reviewed. The systems are compared based on their accuracy and work reliability. The work discusses the machine used, the movement for which the system is tested, the training algorithms used by different researchers, the input control for the system, controlling element for the system, and their advantages. Various training algorithms are also discussed based on the feature extraction techniques which are used to obtain features from the pre-processed brain signals such as EEG, iEEG, EMG. Applying these features in the training model and robot system and testing of the system in real time or pre obtained data. This research paper is written with the intention to differentiate various techniques used and to analyse the best possible technique for conducting the biomechanical operations processes on the robot and to get the user with the most accurate and reliable system which can be used in daily life. These exoskeletons have their application in the field of medicine, military and even help firefighters for various purposes.