Font Size: a A A

Research On Intelligent Decision Theory And Application Based On Rough Sets

Posted on:2006-10-28Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y Q HeFull Text:PDF
GTID:1116360152989417Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
The intelligent decision theory and application based on rough sets is studied in this dissertation. The basic problems in rough sets, the rough sets decision analysis methods for different decision systems, and its applications are studied. The main contents of the dissertation are as follows: The entropy methods to measure the uncertainty and the fuzziness in rough sets are proposed. To the uncertainty in rough sets, new modified rough entropy is proposed to get a proper measure depending on the analysis for disadvantage of roughness and rough entropy. To the fuzziness in the rough sets, new fuzzy entropy is proposed. Moreover, the modified rough entropy and the fuzzy entropy based on equivalence relation is extended to the generalized modified rough entropy and the generalized fuzzy entropy. Finally, the example shows that the new methods for measuring uncertainty and fuzziness in rough sets are more proper and more effective than the classical methods. The discretization of real value attributes is one of the most main problems in rough sets theory. To solve this problem, new oblique hypersurface discretization method is presented. The hypersurface expression and the way of hypersurface obtained by support vector machine are given. The method is applied to the support of Air Force air materiel, and the result shows that more simple decision rules can be obtained. The classical rough sets based on indiscernility relation is extended. To decision system with incomplete and multi-value information, the tolerance rough sets is discussed, and a tolerance rough sets method for modifying spares demand of military aircraft is proposed. To decision system with preferential information, the dominance rough set based on the dominance relation is discussed, and the method is applied to preference selection of Air Force air materiel supply site. To decision system with preferential and incomplete information, the extended dominance rough set based on extended dominance relation is presented. The integration of rough sets and fuzzy sets is studied to the decision system with fuzzy information. Considering that the decision system may be subject to random noise, the rough-fuzzy sets and the fuzzy-rough sets are extended. The compatibility rough-fuzzy set is proposed as the condition attribute values are certain, and the decision attribute values are fuzzy. The compatibility fuzzy-rough set is proposed as the condition attribute values and the decision attribute values are fuzzy. A numberical example was given to illustrate effectiveness of the methods respectively. A rough neural network method is proposed to the decision system with interval numbers. The topologic structure and learning algorithm of the rough neural network are given. The approximation theorem of the rough neural network is proved. Then the method is applied to the fault verification of fighter plane, the simulation results show that the rough neural network is efficient. Finally, the two rough set methods for evaluating self-repairing flight control system effectiveness are proposed to two different evaluation information systems. The integration method of rough and fuzzy sets is proposed as the condition attribute values are continuous,certain, and the decision attribute values are certain. The compatibility rough-fuzzy set method is proposed as the condition attribute values are continuous,certain, and the decision attribute values are fuzzy. Moreover, the two methods are used in effectiveness evaluation for practical self-repairing flight control system successfully.
Keywords/Search Tags:Intelligent Decision, Rough Sets, Extended Rough Sets, Fuzzy Entropy, Rough Entropy, Discretization, Fault Verification, Effectiveness Evaluation of Self-repairing Flight Control System
PDF Full Text Request
Related items