系統復電是一項非常緊急、迫切且複雜的工作。由於配電系統開關組合狀態眾多，若不輔以計算機協助，單靠系統調度員的經驗法則，實難以勝任。為使系統復電工作更快速有效地進行，本計畫透過模糊因果網路(Fuzzy Cause-Effect Networks, FCEN)模式，根據系統狀況及網路所儲存的復電知識，快速推論出可行且最適切之復電策略，以改善傳統方法在緊急迫切情況下，完全依靠調度員操作、極易出錯的缺點。計畫中同時考慮復電量、開關次數、設備過載、及負載平衡等四個目標函數，透過明確定義各目標函數之模糊集，並以滿足此四個目標函數之最大模糊歸屬值，做為最適復電策略之依據。計畫內採用之模糊因果網路模式係整合啟發式搜尋法、模糊因果網路架構、模糊集理論及權值評估方法，故可快速、合理且有效地求解配電系統之多目標復電問題。本計畫以台電公司之一典型配電系統為應用實例，以驗證所提出方法的可行性。
The system restoration is an urgent, emergent, and complicated task. Due to the existence of numerous switching combinations available, it is hard to fulfil the restoration solely depending on the experiential rules by the human dispatchers. According to the system status and the restoration knowledge stored in the networks, the project proposes a fuzzy cause-effect networks (FCEN) model to rapidly conclude a feasible and satisfactory restoration plan in order to improve the disadvantages of making false decision under emergent conditions by the dispatchers. The objectives of the restoration scheme obtained concerns: (1) as many loads restored as possible, (2) as few switches operated as possible, (3) no devices overloaded too much, if they must be, and (4) as load balance as possible. The proposed method first defines the fuzzy sets of each of the objective functions concerned. The solution with maximal membership function after combining these four objectives is the final restoration scheme desired. The FCEN model adopted in the project integrates the heuristic searching method, basic fuzzy cause-effect structures, fuzzy logic theory, and weight assessment approach in an attempt to efficiently, reasonably, and effectively solve for the multi-objective restoration problem of distribution system. The proposed FCEN approach has been practical tested on a typical TPC distribution system.