Combination of Reinforcement Learning and Dynamic Self Organizing Map for Robot Arm Control

This paper shows that a system with two link arm can obtain arm reaching movement to a target object by combination of reinforcement learning and dynamic self organizing map. Proposed model in this paper present state and action space of reinforcement learning with dynamis self organizing maps. Because these spaces are continuous. proposed model uses two dynamic self-organizing maps (DSOM) to estimate the state and action space by addition and reduction of neurons. It has been showed that DSOM has a better execution in conservation of network topology, error reduction and processing speed than the SOM, a new reinforcement learning algorithm based on these maps are presented which has variable Q-table that initialization, changing the nuerons and update the Q-table has been studied. In this paper, a new reinforcement learning algorithm based on dynamic self dynamic map has been presented that it’s Q-table is magnified over time.