Summary: | 碩士 === 國立中山大學 === 電機工程學系研究所 === 91 === The distribution system containing numerous protective facilities and switch equipment ranges over a wide boundary . The most urgent problem the dispatcher has to tackle right after the breakdown of distribution system in how to resume as soon as possible power supply for the none-faulted out-of-service area. Therefore, distribution system service restoration is an important and practical subject.
During the process of maintaining and operation ,the distribution district collected a lot of data. These data includes substation-related automatic operating data , switch section of doing a feeder , load data of high and low tension customers, and historical distribution system planning information concerning load transfer while operation, etc. The useful resource could help the dispatcher to get the best load transfer on distribution system service restoration.
Datum-oriented researcher refers to the exploration of the concealing regulations and knowledge and the techniques regarding comprehensible model. This thesis manages to collect the maintaining and operation data of the distribution district ; it makes use of the information exploration techniques and ID3 decision tree and fuzzy theory to get the load transfer rule and knowledge , and to establish the load transfer model. After the fault occurred, it would help the dispatcher to get the best tactics for the load transfer on distribution system service restoration under the constraint condition.
This thesis chooses the underground distribution system in sijhih of the Keeling District Offices of Taiwan Power Company as a sample. That underground distribution system comprising 8 distribution feeder ,73 distribution spaces,16 high tension customers, 28 feeder section, 4 feeder-tie-switch and 19 lateral-tie-switch will be simulated on computers to verify the proposed method for the distribution system service restoration.
|