本文提出一種變係數下垂控制策略應用於儲能系統及太陽能光電系統並聯運行的情況下，利用儲能系統之實功率/角頻率P-w下垂控制方程式進行反推的方式來估測下垂係數，以此改善微電網實功率及頻率在負載變化時之暫態響應。此外，為了更有效改善微電網實功率及頻率在負載變化下之暫態響應，本文提出了具有線上訓練能力的柴比雪夫派翠模糊神經網路(Chebyshev Petri Fuzzy Neural Network, CPFNN)以用於取代傳統的比例積分(Proportional-Integral, PI)控制器，並且本文詳細推導所提出的CPFNN之網路架構和線上學習策略。最後，本文利用模擬及實驗以驗證所提出之變係數下垂控制策略結合CPFNN於微電網中改善實功率及頻率之暫態響應的有效性;A variable coefficient droop control strategies applied to the parallel connection of energy storage system and solar system in this study, and uses the real power/angular frequency P-w droop control equation of the energy storage system to estimate the droop coefficient , so as to improve the transient response of the real power output and frequency of the microgrid when the load changes. In addition, in order to more effectively improve the transient response of the real power output and frequency of the microgrid under load changes, this paper proposes a Chebyshev Petri Fuzzy Neural Network (CPFNN) with online training capabilities for replaces the traditional Proportional-Integral (PI) controller, and this paper deduces the network architecture and online learning strategy of the proposed CPFNN in detail. Finally, this paper uses simulation and experiments to verify the effectiveness of the proposed variable coefficient droop control strategy combined with CPFNN to improve the transient response of real power and frequency in microgrids.
