Adaptive optimal multi-critic based neuro-fuzzy control of MIMO human musculoskeletal arm model

• Published in: Neurocomputing (Amsterdam) 2016-01, Vol.173, p.1529-1537
• Main Authors: Balaghi E., M. Hadi, Vatankhah, Ramin, Broushaki, Mehrdad, Alasty, Aria
• Format: Article
• Language: English
• Subjects: Adaptive optimal controller; Functional electrical stimulation; Input optimization; Musculoskeletal model; Neuro-fuzzy controller
• ISSN: 0925-2312; 1872-8286
• DOI: 10.1016/j.neucom.2015.09.026

Human bodies use the electrical currents to make the muscles move. Disconnection of the electrical signals between the brain and the muscles as a result of spinal cord injuries, causes paralysis below the level of injury. Functional electrical stimulation (FES) is used to stimulate the peripheral nerves of the disabled limbs. The level of these electrical signals should be selected so that the desired tasks are done successfully. Applying the appropriate controller which can result a human like behaviour and the accomplishment of the desired tasks has become a significant research area. In this paper, the multi-input multi-output (MIMO) musculoskeletal model of human arm with six muscles is investigated and a new Adaptive optimal critic-based neuro-fuzzy controller is proposed to control the end point of the human arm model. Computer simulations are accomplished to investigate the effectiveness of the proposed neuro-fuzzy control structure. Adaptivity and muscle force optimization are of important features of the proposed neuro-fuzzy controller. The results show satisfactory behaviour.