Toward affective cognitive robots for human-robot interaction

摘要

We present a brief overview of an architecture for a complex affective robot for human-robot interaction. Implementation of an Affective Architecture Affect might have several functional roles in agent architec-tures (Scheutz 2004). Some have argued that affect serves the purpose of integration and management of multiple pro-cesses and is thus required for the effective functioning of an autonomous system (Ortony, Norman, & Revelle forth-coming). Specifically, affect allows for motivational signals originating not from changes in the external environment de-tected via sensors, but from components within the architec-ture itself (e.g., from deliberative subsystems). Such signals can then influence various other parts of the architecture and modify goal management, action selection, and learning.