부산대학교 도서관

The growing share of renewable energy (RE) leads to increasing the number of feed-in converters. The standard parametrization method of converters merely takes into account one’s steady state of point of common coupling (PCC). This massive clustering of the converter with standalone parametrization can provoke resonance between the controlloop, which destabilizes the system. This paper presents an adaptive approach, dynamic assistance system, to improve the joint stability in a converter dominated distribution grid. The operating point of the cluster is estimated by employing the artificial neural network (ANN), which enables each converter to use their operating-point-associated parametrization adaptively. Simulation results demonstrate an improvement of system stability.