A novel intelligent hyperball fuzzy agent for intelligent inhabited environments

Generally, the occupants in a same inhabited environment may have different preferences. Moreover, the occupants' preferences change with varying space, time, etc, which desires that the intelligent agents have evolvement function. In this paper, a novel intelligent hyperball fuzzy control agent is proposed, which is embedded in the inhabited environment where it learn the user behavior in a non intrusive mode and control the environment on the user behalf to realize the intelligent ambience. First, an improved FCM fuzzy cluster method is given to determine the number of fuzzy rules. Second, the one-dimensional membership functions are replaced by the multi-dimensional membership functions. Third, a fuzzy inference algorithm is presented to calculate the initial values of fuzzy rule outputs. The proposed approaches can reduce the number of fuzzy rules and simplify the network calculation. Moreover, the fuzzy rules base can be modified online when the users' preferences changed. The simulation results demonstrated the effectiveness of the proposed approaches.