The integrated methodology of rough sets theory, fuzzy logic and genetic algorithms for multisensor fusion

The strong qualitative analysis ability of the rough sets theory is used to deal with the multisensor datum in order to extract a hierarchy rule set for fusion. Even in the absence of incomplete measures, the hierarchy rule set can also derive satisfied results. However, the rough sets theory processes the discrete datum, so discretization of the continuous valued attributes of raw sensor datum to intervals must be performed first. In the rough sets theory, the dependency factor represents the consistency of a decision system. So it is used as the fitness function of genetic algorithms to derive the optimal cut points of intervals in order to ensure the maximum consistency of the discrete datum. However, normal interval lacks the robustness and continuity, so at the same time it is fuzzified and fuzzy inference is used to make decision in order to enhance the robustness