Multi-resolution state-space discretization for Q-Learning with pseudo-randomized discretization

A multi-resolution state-space discretization method with pseudo-random griding is developed for the episodic unsupervised learning method of Q-Learning. It is used as the learning agent for closed-loop control of morphing or highly reconfigurable systems. This paper develops a method whereby a state-space is adaptively discretized by progressively finer pseudo-random grids around the Regions Of Interest within the state or learning space in an effort to break the Curse of Dimensionality. Utility of the method is demonstrated with application to the problem of a morphing airfoil, which is simulated by a computationally intensive computational fluid dynamics model. By setting the multi-resolution method to define the Region Of Interest by the goal the agent seeks, it is shown that this method with the pseudo-random grid can learn a specific goal within ±0.001, while reducing the total number of state-action pairs needed to achieve this level of specificity to less than 3000.