Research on a Fuzzy Agent that Makes Autonomous Decisions (II)

2005 
In this paper, we apply software agents, which use fuzzy logic and make autonomous decisions according to state-transitions, to car driving environments. We carry out an experiment on the intelligent car driving in terms of real-time reactive agents. Inference techniques for constructing the real-time reactive agents consider the settings with the max-product inference, N-fuzzy rules, and N-associatives. We then perform defuzzification processes, extract a central value, and work out inference processes. We also show some algorithms and diagrams for the agent system. We evaluate our experiment with various language variable membership values on the same car (Case A). The fuzzy inference for driving control of the car adopts multiple inputs and single output method. In this experiment, we process language variables as 5 × 5 matrixes. We also evaluate our experiment with the same input language variables, distance, and output language variables, and with the different scope of language variables (Case B).
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    1
    References
    8
    Citations
    NaN
    KQI
    []