Organization and selection methods of composite behaviors for artificial creatures using the degree of consideration-based mechanism of thought

Abstract

This paper proposes organization and selection methods of composite behaviors for artificial creatures. Using the degree of consideration-based mechanism of thought (DoC-MoT), each pre-defined atom behavior is evaluated by the fuzzy integral of the partial evaluation values of atom behaviors over the artificial creature’s wills and external contexts, with respect to the fuzzy measure values representing its degrees of consideration (DoCs). Based on these evaluation values of atom behaviors, a composite behavior is organized as a set of atom and composite behaviors which are connected by the relationships of ‘parallel,’ ‘choice’ and ‘sequence.’ However, in the organized composite behavior, the behaviors connected by ‘choice’ relationship can not be generated at the same time, and therefore, only one atom or composite behavior is randomly remained in each set of atom or composite behaviors connected by ‘choice’ relationship. The effectiveness of the proposed scheme is demonstrated by simulations carried out with an artificial creature, “DD” in the 3D virtual environment. The results show that the diversity of the generated behaviors is increased fourfold compared to the behavior selection without the organization process of composite behaviors. Moreover, the generated composite behaviors satisfy the artificial creature’s wills more and the logical connectivity of them is increased compared to the method without the process.