Cell-state-space-based search

The cell-state-space concept is a promising computational method for designing, evaluating, training, adapting, and tuning digital control systems. Over the last decade, we have continually improved and refined some of these cell-state-space methods. This effort has culminated in an automated controller optimization algorithm called incremental best estimate directed search (IBEDS). IBEDS starts from an initial training set obtained through the sampling of the control surface of a controller with poor performance. IBEDS is based on the assumption that if the training data set is optimized, the fuzzy logic controller trained therefrom will also be optimized.