System analysis of interactive genetic algorithm

Interactive genetic algorithm (IGA) was analyzed from the viewpoint of system, which help to understand IGA better. As a system, IGA is composed of user-subsystem, evolutionary-subsystem, and two information transfer interfaces. User-subsystem has the feature of fluctuation of preference and evolutionary-subsystem has the feature of fitness noise. These two features make some negative influence in the knowledge immigration from genetic algorithm to IGA. The main dynamic source of evolution is the interaction between subsystems, and the information-entropy of user preference and search space will increases during the evolution. The worth of IGA system ties in design, exploration and optimization. The above research of IGA established the necessary base for future studies.