| Summary: | 碩士 === 大同工學院 === 資訊工程學系 === 84 === AbstractThe Applications of dynamic fuzzy reasoning model to
Chinese chess and decision making problem are investigated in
this thesis. The Chinesechess is one of the most popular
examples of two-person zero-sum game,that is, a game in which
one player's loss is his opponent's gain. Wepropose a useful
fuzzy approach to help Chinese chess player to decide the best
next move. The novel technique called "guard heuristic" with
fuzzy cognitive map is exploited to dynamically determine the
guard value by correlating relationship between each pair of
intersections on the board. This can in turn impact more than
one moves of guard values.The best legal move is reasoned by
fuzzy rule based model. We use several boards to simulate best
legal move. In decision making problems, the problem of talent
selection is approached from the fuzzy method. To provide
thedecision maker a better way to find the best winner, we
consider multiperson and multicriteria decision making models to
satisfy the real situations.Both vertical and crabwise models
are presented for multilayer decision model. Besides, we
dynamically change the weights of leaves of decision making
tree, use new scoring method for subtree, and adapt the width of
membership functions,to compare the differences. A number of
simulation results are shown tovertify the effectiveness of the
proposed work.
|
|---|
