Skip to Main Content
Human perception and processing of information is granular and multi-resolution instead of numerical and precise. Due to this multi-resolution perception-based computing, human mind can quickly evaluate (calculate) the fitness of a large subspace of the search space. Indeed, this characteristic enables human to simplify and solve very complex problems. In contrast, evolutionary optimization (EO) as one of the most applied artificial problem solvers is based on computing with numbers since a chromosome is a single point of the search space and fitness function calculation is numerical. Hence, EO is blind towards the optimization landscape and this blindness inhibits its performance when the search space is very large and complex. Inspired by human perception based reasoning, a novel approach to optimization and problem solving is proposed here. Perception-based evolutionary optimization (PEO) is fundamentally based on computing with words. In PEO, chromosomes and fitness function calculation are perception-based (granular) instead of numerical and thus PEO works with granules (subspaces) rather than single points. Also, search is performed in a multi-resolution manner.