Self adaptation of agent's behavior using GA with n-BDD

A gene expression n-BDD (binary decision diagram) is proposed in the paper. The gene expression is suitable for agent models to decide agent's behavior using information from environment. We can apply genetic algorithm to evolve behavior of agents. A simulator of a quasi-ecosystem developed using the gene expression is described. In the quasi-ecosystem, a herbivore, a carnivore and plants make a food chain. A herbivore using an n-BDD more rapidly acquire its ability to survive from a carnivore than using finite state automaton.