Modelling and Control Human Arm with Fuzzy-Genetic Muscle Model Based on Reinforcement Learning: The Muscle Activation Method

Abstract

Background: The central nervous system (CNS) is optimizing arm movements to reduce some kind of cost function. Simulating parts of the nervous system is one way of obtaining accurate information about the neurological and treatment of neuromuscular diseases. The main purpose of this paper is to model and control the human arm in a reaching movement based on reinforcement learning theory (RL).Methods: First, Zajac’s muscle model is improved by a fuzzy system. Second, the proposed muscle model applied to the six muscles which are responsible for a two-link arm that move in the horizontal plane. Third, the model parameters are approximated based on genetic algorithm (GA). Experimental data recorded from normal subjects for assessing the approach. At last, reinforcement learning algorithm is utilized to guide the arm for reaching task.Results: The results show that: 1) The proposed system is temporally similar to a real arm movement.  2) The reinforcement learning algorithm has the ability to generate the motor commands which are obtained from EMGs. 3) The similarity of obtained activation function from the system is compared with the real data activation function which may prove the possibility of reinforcement learning in the central nervous system (Basal Ganglia). Finally, in order to have a graphical and effective representation of the arm model, virtual reality environment of MATLAB has been used.Conclusions:  Since reinforcement learning method is a representative of the brain's control function, it has some features, such as better settling time, not having any peek overshoot and robustness.