AN IMPROVED IMPORTANCE REDUCTION ALGORITHM APPLICATION IN REGIONAL POWER GRID FAULT DIAGNOSIS
A large number of incomplete information will flood into the dispatching center due to the refusal or misoperation of protec-tion devices after the failure of distribution network.These missing,redundant and wrong information may lead to inaccurate diagnosis results or even serious errors.Based on the original classical dependency algorithm,this thesis proposes an improved importance reduc-tion algorithm to reduce the decision table with incomplete,redundant and wrong information.The decision attributes of the reduced de-cision table are consistent with the original decision table,but the conditional attributes are less than the original decision table.Finally,it calculates the importance of inconsistent decision table after distribution network failure,and reduces the original decision table by using the reduction algorithm of improved importance to get the decision table after removing redundant attributes with an example of re-gional distribution network,which solves the problem of misjudgment of decision attributes caused by lack of information in inconsistent decision table and verifies the correctness of the algorithm.
importanceattribute reductionregional power gridfault diagnosis
万方数据
维普
