離島系統智慧型防禦機制之研究

Research on Intelligent Defense Mechanism for Offshore Island Power Systems

詹博雄(Bo-Shiung Jan)

指導教授 ： 吳元康

國立中正大學/工學院/電機工程學系/碩士(2014年)

摘要

本論文將針對離島獨立系統進行智慧型防禦機制以及後續系統保護的規劃，並建置一套防禦以及告警機制。本論文以金門與澎湖系統為例，驗證本研究所提出之防禦系統的可行性以及預測的精確度。本論文首先以金門和澎湖系統實際運轉資料為例，利用PSS/E批次模擬系統於不同運轉下所產生的各項模擬數據，並進而建立系統資料庫。接著，本論文藉由系統模擬獲得發生跳機事故後的頻率響應數據，並且利用倒傳遞類神經網路設計智慧型預測系統，有效的預測當系統發生各種擾動時的頻率響應，藉此規劃系統相應的保護策略，包含發電機組重新調度、增加柴油機組數目、或是執行最佳化的低頻卸載策略。
本論文的主要目的在於改善目前離島系統的供電穩定性，由模擬結果顯示，本文提出之類神經網路可準確地估測頻率最低點、頻率變動率、以及在維持系統穩定運轉前提下須要卸除的最小負載量。本論文的研究成果可提供調度人員作預防性調度及最佳卸載的規劃。

關鍵字

低頻卸載 ； 類神經網路 ； 頻率響應 ； 智慧型防禦機制

並列摘要

This thesis develops an intelligent defense mechanism and system protection planning, as well as establishes an alarm strategy for isolated island systems. This thesis takes Kinmen and Penghu systems for example to identify the feasibility of the proposed defense system and the forecasting accuracy. First, this thesis uses the actual operational data measured in Kinmen and Penghu systems; then creates an online database by implementing various simulation analyses under different operational patterns. The software, namely PSS/E, was utilized to perform the simulation, and we design a number of diesel generator tripping accidents to acquire the results of frequency responses by simulation. Next, this thesis designs the intelligent forecasting system by using back propagation artificial neuron network (BPANN) that can effectively predict system frequency response when a diesel generator tripping occurs. The forecasting results can assist in planning corresponding system protection strategies, including generator rescheduling, increase of the number of diesel generator units, or performing optimized low-frequency load shedding strategy.
The main target of this thesis is to improve the system stability on the remote island systems. The simulation results show that the proposed BPANN can accurately estimate frequency nadir, rate of frequency change, and the minimum load shedding amount under stable operation. The results of the thesis can provide the system operator with a reference relative to the preventive dispatch and the plan of load shedding.

並列關鍵字

Intelligent defense mechanism ； Under-frequency Load Shedding ； Artificial Neural Networks (ANN) ； Frequency response

參考文獻

[19] 張銘仁，應用類神經網路於高占比風能電力系統中發電機跳機之最低頻率預測，碩士論文，臺北科技大學電機工程系，臺北，2013。

[1] Q. Kejun, et al, “A hybrid power system using wind and diesel generator: A case study at Masirah Island in Oman,” the 20th International Conference and Exhibition on Electricity Distribution-Part 1, pp. 1-4, Jun. 2009.

[2] J. Kabouris and F. D. Kanellos, “Impacts of large-scale wind penetration on designing and operation of electric power systems,” IEEE Transactions on Sustainable Energy, vol. 1, Iss. 2, pp. 107-114, Jul. 2010.

[3] G. Lalor, A. Mullane, and O'Malley M., “Frequency control and wind turbine technologies,” IEEE Transactions on Power Systems, vol.20, Iss.4, pp. 1905-1913, Nov. 2005.

[4] P. B. Eriksen, et. al., “System Operation with High Wind Penetration”, IEEE Power and Energy Magazine, Vol.3, Iss. 6, pp. 65-74, Nov. 2005.