航空發動機維修排程作業之研究

Abstract

飛航安全是目前國際上相當重視的課題，航空器不僅影響旅客生命安全，同時也關係著國家形象；而航空器中，發動機扮演相當重要的角色，發動機維修作業是否依照標準進行，將影響航空器的正常運作。因此國際上對於發動機維修訂有嚴格的標準，業者必須依固定時間、固定里程進行發動機保養維修。 維修作業之排程工作均以人工方式進行，由排程人員判斷各項作業之剩餘維修容量，但由於維修作業程序相當繁瑣，包含發動機的拆解與組裝、各項零件更換與送修等，總計可達400多項主要作業項目，排程人員不易了解現場狀況與維修容量運用情形，一般均依據經驗安排各項作業，經常有維修容量分配不均的情形發生。若能提供一套分析方法，判斷現有容量與最佳的容量配置方式，以有效運用維修容量、縮短發動機維修時程，對於須兼顧飛航安全與企業利潤的業者來說，是相當有幫助的。 本研究利用發動機維修流程與排程方法發展一套模式，透過現況資料的輸入，求解各情境下維修容量最佳配置方式。由於求取最佳解花費時間較長，因此透過敏感度分析方法，找出影響維修時程之主要因素，以及非關鍵性作業可調動範圍，可做為臨時調動參考之用，減少重新求解的次數。 本研究以數學規劃方法建構模式，並撰寫一現況資料輸入程式，將現況參數設定完成後，利用CPLEX軟體來產生最佳的排程方案。同時，為加速最佳排程方案的求解速度，本研究採用邏輯式分枝切面法(Logic-based Branch-and-Cut Method)產生切面不等式(Cuts)，確保求解時間在可接受的範圍之內。 為測試系統的正確性、合理性，本研究產生各種與現況相同的排程情境，作實例測試與修正，並與現況進行比較，分析績效以及分枝切面法的加速效率，以使排程方案及求解速度能符合業者實際需求。

Aviation safety is an important issue internationally. The engines are the source of power in aircraft and therefore play a significant role in aviation safety. An engine functions well if standard maintenance procedures are followed. Therefore, some international organizations set up a serial of strict, compulsive and complicate standard procedures to assure engine maintenance is well done. However, the current way of making up the schedule in company A (The target of our case study in this thesis) is manually. The schedulers usually have difficulties to know statuses of work fields and capacity of machines. A good schedule is needed to achieve effective usage of manpower and machine capacity and also shorten maintenance duration efficiently. Identically, it could give consideration to both aviation safety and company profits. The objective of this research is to propose a model to generate activities’ schedule for a real world maintenance company. This model is based on the maintenance procedures and scheduling rules, and takes the current operation status into account. We also propose an algorithm for solving this model to its optimality. In addition, because it usually takes significant time to obtain the optimum, this research use the logic-based branch-and-cut method to speed up the solution processes. In order to evaluate the accuracy and rationality of our model, we generate some situations for testing and modifying our model and compare with actual data to analyze the performance. We also analyze the efficiency of the speeding up technique to ensure that our model could satisfy actual works.