考量可更動座位配置之偏鄉客貨混載最佳化

Abstract

偏鄉地區的居民在就學、就醫以及日常生活採買等民生需求時，經常受到交通不便的困擾，近年來政府推廣需求反應式運輸服務，提供乘客較為彈性的運輸服務，然而由於乘客數量較少，偏鄉客運仍需依賴補貼才得以營運。為了改善上述問題，客貨混載的概念因而產生，結合乘客和貨物的運輸，貨物的收入不僅可維持業者營運，也能吸引更多業者投入此服務。由於過往客貨混載研究多固定乘客和貨物容量，因此本研究將探討使用不同容量的可更動配置小客車同時運送乘客和貨物，亦考量到乘客乘坐的舒適度，減少乘客途中因車輛更改配置而需換座位的情形。本研究目的為建構可更動配置車輛之偏鄉客貨混載模式，亦開發一確定性退火演算法以提升求解效率，研究結果顯示演算法相較於數學規劃模式可在較短時間內求得良好的解，模式亦驗證使用客貨混載能夠降低營運成本，考量乘客乘坐時舒適度可以有效減少乘客更換座位情形。透過敏感度分析，可得不同車隊大小、乘客數量占總需求數的比例和乘客及貨物分布位置等參數對模式結果之影響，期望能提供未來營運業者評估之參考。

Residents in rural areas often face transportation inconveniences when it comes to accessing education, healthcare, and daily necessities. Recently, the government has promoted demand-responsive transport services to provide passengers with more flexible transportation services. However, rural public transportation still relies on subsidies to operate due to the low number of passengers. To address these issues, the concept of integrated passenger and freight delivery has emerged. The revenue from freight not only supports the operation of service providers but also attracts more participation from businesses. Since previous studies on mixed passenger and freight transportation often fixed the capacities of passengers and goods, this paper discusses the transportation of passengers and goods using small and configurable vehicles with different capacities. The study also takes into consideration passenger comfort to minimize the need for seat changes during the route. The objectives of this research are to construct an integrated passenger and freight delivery with configurable vehicles model for rural areas and to develop a deterministic annealing algorithm to enhance solution efficiency. The results indicate that the proposed algorithm achieves good solutions within a shorter time compared to the mathematical model. The model also verifies that utilizing mixed passenger and freight transportation can reduce operating costs and considering the comfort of passengers can effectively reduce the need for passenger seat changes. Through sensitivity analysis, the impact of parameters such as fleet size, the proportion of passengers to total demand, and the distribution of passengers and goods on the model's results is investigated. These results are expected to provide references for future service providers in evaluating different scenarios.