傳統在處理公車不確定狀況應變營運管理問題時，大多只以需求面之角度探討應 採取何種措施，並且沒有考慮異動司機的意願，不盡合理。而在不確定狀況下所影響之班次 數方面，只能處理車輛故障、人員缺席、道路擁擠與乘客需求增加之個別單一狀況，但是對 於複合狀況所產生之影響，則略過不予探討。此外，傳統之研究亦僅探討主副線之營運管理 問題，且大部分假設副線必須無條件的支援主線，此作法也不符合實際調度站之運作。 針對上述尚待解決問題，本研究由個別路線班次之供給與需求角度，以模糊專家系統區 分三個模組來分析此問題。首先，在增加班次需求知識庫模組中，計算出不確定狀況所影響 的班次數。其次，在可供給班次模組中，計算出可能會受異動司機之願意加班或拆班之總計 模糊隸屬度函數。最後，在可行方案模組中，綜合來自增加班次需求知識庫模組與可供給班 次模組之資訊，進行可行方案之排序。若可供給班次不足以支應增加班次需求，則計算出最 大可供給班次之次佳方案。 本研究並區分供給大於或等於需求與供給小於需求兩種情況作成範例，實際描述個案在 狀況發生時之處理方式，以顯示本研究所研發的公車不確定狀況應變營運管理模糊專家系 統，確具有良好的適應性與強韌性。

　　　　　In dealing with bus contingency plan management, we usually consider what kinds of actions should be taken under uncertain situations from a demand aspect. And the willingness of dispatched drivers is usually being neglected, this is unreasonable. The previous studies also showed that they can handle single uncertain situation such as the absence of drivers, the breakdown of vehicles, the variation of demand and congested traffic, but cannot deal with their combination. Furthermore, those papers are focused on main line and subordinate line, and hypothesize that the operations of subordinate line should support main line under any situations. We develop a fuzzy expert system including three modules to solve the above- mentioned problems from the viewpoint of transportation demand and supply. First, in knowledge module of extra board (frequency) , we estimate the effected line frequency for uncertain situations. Second, we calculate aggregate membership function of dispatched drivers in supply module. Finally in the feasible alternative's module, we synthesize information of the above two modules to determine the ranking of feasible alternatives. We offer the second best one if supply is lower than demand. To illustrate our system's adaptiveness and robustness, two paradigms are provided for different situations: the one is that demand is higher than or equal to supply; the other is that demand is lower than supply.