良好的運務員作業排程規劃可幫助國際快遞物流業者確保資源充分應用與維持快遞服務品質。目前實務上運務員作業排程係由國際快遞物流業者成立之指揮中心依人工經驗進行指派且多以平均旅行時間為依據進行規劃。然而，在實際營運中旅行時間往往受隨機因素造成擾動且新的變動服務需求經常發生，使得業者需經常調整運務員作業排程，以處理隨機旅行時間造成之擾動及新的變動服務需求。此作法不具系統性分析並依賴相關作業人員的主觀判斷，可能導致不佳的規劃決策。有鑑於此，本研究考量隨機旅行時間，針對規劃階段與營運階段分別構建二運務員作業排程規劃模式與二運務員作業排程調整模式，以幫助國際快遞物流業者有效地規劃或調整運務員作業的排程，進而提昇整體的營運績效。 本研究利用時空網路流動技巧分別構建隨機旅行時間下之與運務員排程規劃與調整模式，並進一步修正此等隨機模式之旅行時間為平均旅行時間求，建立二確定性模式。最後，本研究以國際之一大型快遞物流業者的營運資料為例，進行實例測試與分析，模式測試結果良好，比較實際營運績效，最後綜合測試結果提出結論與建議。 A good courier schedule can help international express carriers to effectively allocate resource and maintain service quality. However, not only new demands often occur in actual operations making the carrier frequently adjust its schedule to meet the new demands, but the schedule is often disturbed arising from variations in vehicle travel times. Currently, courier schedules are planned by the decision center staffs with experience. Without a systematic optimization analysis, these decisions highly depend on the staffs’ subjective judgments and could result in inferior feasible solutions. Therefore, in this research, considering stochastic vehicle traveling times we attempt to develop two courier scheduling models and two courier schedule adjustment models, respectively, for the planning and operational stages. These models are expected to help carriers effectively and efficiently plan and adjust their courier schedules, and thus to improve their operational performance. In this research, we will employ time-space network techniques and mathematical programming methods to construct the courier scheduling models and schedule adjustment models under stochastic traveling times. Then, we will modify the stochastic travel times/demands as an average travel time to develop two deterministic scheduling models. These models are expectedly formulated as special integer multiple commodity network flow problems, which are characterized as NP-hard. Finally, to evaluate the models and the solution algorithms in practice we performed a case study based on the operating data from a major international express carrier. The test results show that the model is good for this kind of problem. Conclusions and suggestions were given finally.
