developed and implemented for Online Short-term Load Forecasting System for Korea Electric Power Corporation (KEPCO).
forecast the 24-hour total electric power system load nationwide one-day ahead
ANN method : forecasting the work days (having similar patterns each other)
Fuzzy relation method : foresting weekend and holidays (having specific patterns)
ANN are trained with many shapes of impedance loci for various fault conditions (input parameters of ANN) and their fault distances (output parameters of ANN). The proposed method improves the speed of fault detection without accuracy degradation
By using the new indicators, voltage unbalance rate and harmonic distortions of current, the proposed method overcomes the limitations of previous islanding detection methods on small variation before/after the main grid interruption
The proposed method changes the decision criteria for overcurrent detection and enhance the performance of overcurrent relay which are usually hindered by the variation of wind turbine generation since WT supply the fault current to the faulted circuit.
The 1st try in Korea for developing the Korean large-scale Wind Turbines (750kW) in 2001
Assessment of the effect on power system operation and power quality due to interconnection of the WT
developed the integrated monitoring system for WT test and research
The projects design the pricing and tariff structure for distribution network (under 22.9kV)
The interconnection capacity of DG was limited with 10MW in Korea Distribution network(22.9kV), and the projects proposed increasing to 10 ~ 30MW depending on the network condition