不確定因素下整合風能微電網系統之設備規劃與配電策略最佳化

Abstract

在能源發展的議題中，再生能源開發與電力系統轉型是邁向能源永續的兩大關鍵，而為了有效率地將再生能源應用於電力系統中，勢必需要長期且有系統的評估與規劃。 本論文針對整合再生能源的微電網系統規劃問題，提出一套完整的配電與設備規劃流程，此流程不僅包含了風能評估與用電預測，更同時考量了電力系統運行時的配電策略。此流程使用小波轉換與時間序列等數據分析方法建構風速與用電模型，並將所得之模型應用於流程中的配電與設備規劃，以獲得最適合該電力系統的的設備規模以及配電策略。在完成了配電與設備規劃後，本研究使用歷史數據模擬電力系統的實際運作狀況，以驗證最佳化結果的可行性。模擬結果顯示，在整合配電策略的設備規劃問題中，必須考量風速與用電等不確定因素，才能確保電力系統的可行性。 為了量化風速與用電不確定性對配電策略與電力系統所產生的影響，本論文提出了一套以機率理論為基礎的長期配電規劃。此長期配電規劃方法考量了風速與用電的不確定性，可以提供發電廠的可能操作範圍與儲能設備的電能存量變動範圍。發電廠的可能操作範圍使此配電規劃具有即時調整的彈性；而儲能設備的電能存量變動範圍則提供了更充足的資訊，以利決策者決定適當的裝置容量。在儲能設備的裝置容量決定後，文中亦使用歷史數據進行模擬驗證，而其結果顯示，此長期配電規劃方法可以有效地量化不確定因素對於電力系統所產生的影響，因此決策者可以透過此配電規劃的結果決定適當的儲能設備裝置容量。 本論文所提出之設計流程整合了電力系統的配電與設備規劃，針對目前的電力系統提出了一個漸進式的轉型方案。決策者可以藉由反覆的執行此流程，引領目前的電力系統逐步朝向可以獨立運作的微電網系統邁進，以達到能源永續的最終目標。

The global quest for energy sustainability has motivated the development of transforming various natural resources into energy efficiently. Combining these renewable energy sources with existing power systems requires systematic assessments and planning. The present work proposes a design procedure for obtaining the optimal sizes of wind turbines and storage devices considering power dispatch with wind and load forecasting. At first, the wind and load models are constructed by wavelet packet analysis and moving average technique. These models are applied to the design procedure to determine the optimal sizes and optimal dispatch strategy. Then, a real-time operating simulation is used to validate the feasibility of the optimal results in the real world. Results show that the models used in the optimization framework should consider the uncertainties to maintain high system feasibility. To quantify the influence of wind and load uncertainties in the optimal sizing and dispatch problem more practically, a novel probability-based power dispatch strategy is proposed. The new strategy estimates a probable dispatch range for a long-term power dispatch and quantifies the variation of the state of charge of energy storages under wind and load uncertainties. The probable dispatch range provides more real-time flexibility for the long-term power dispatch, and the variation provides more information for determination of storage capacity. After determining a suitable storage capacity, a validation simulation is also used to observe the behavior of the power system. Results show that the probability-based power dispatch strategy could estimate the probable range and variation effectively, and that the capacity of energy storage is well determined. This work integrates equipment sizing and power dispatch problem into the design procedure. The procedure provides a gradual planning of a power system, leads the existing power system toward microgrid system, and eventually reaches energy sustainability.