Deep Learning Method for Power System Generator Tripping Assessment

• Main Authors: Lai, Kuan-Ting, 賴冠廷
• Other Authors: Huang, Pei-Hwa
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/68qw2v

碩士 === 國立臺灣海洋大學 === 電機工程學系 === 106 === The main purpose of this thesis is to use Deep Learning method to determine that which load shedding can let the system recover stable and decide the better load to be shed. When the generator tripping occurs in the system may be affected the voltage, frequency and rotor angle of the system. If the system loses excessive amounts of output power, the system frequency decrease rapidly. To avoid system frequency drop to cause system breakdown, we must shed the loads appropriately to let the system recover stable. This thesis uses the power system state value as the feature to train the Deep learning network that is based on the Levenberg-Marquardt algorithm. The results show that the accuracy rate of the Deep Learning network to decide the better load to be shed can reach 88%.