台灣地區汽車路線貨運業長途運輸班次規劃之關鍵因素研究--以H公司為例

Abstract

隨著國內整體工、商業環境的快速變遷、電子商務產業的興起，以及都市人口的高密度發展，使得不論是企業或個人，對於貨物商品的運輸需求與日俱增，而貨物運輸的方式亦朝向便捷快速與提昇效率的方向發展，現行國內較具規模的路線貨運業者，每日投入作業的人、車、場站之資源極其龐大，其中所牽涉的長途班次規劃、車輛派遣以及臨時加開之班次車輛調度，常因每日各項變數的改變，即產生不同的結果；在沒有完整而符合業者需求的班次決策系統之情形下，業者只能依賴傳統之人員訓練方式，培養具靈活思考能力的專家，來處理各種變化。 然而專家的培養，常遭遇許多問題，如人員的適任性及異調問題、經驗傳承及知識庫建立的困難、決策品質難以評估…等等，所以如何將專家的知識和經驗建構於電腦上，使具有推論能力的電腦化系統，以類似專家解決問題的方式對運輸班次規劃領域的問題提供建議或答案，並能解釋推論結果，仍是業界極欲找尋的方式。 路線貨運經營問題及班次排班問題在國內外學術文獻中常被提及，然而相關班次排班演算方法多從數學規劃的角度推演，對於實務上多元而需以管理方式處理的變數，均難以有規律的解決方式；隨著資訊科技的逐漸發達，以現代化電腦之人工智慧來取代一些繁瑣而需學習及記錄的工作，確實可以解決以往不能解決的繁雜問題；因此如能建立一套機制或演算方式，可有效解決實務上複雜的班次排班問題，同時可做績效目標的評核，將可有效地使台灣地區汽車路線貨運業者之營運績效提昇。

With the overall domestic business environment rapidly changing, the rise of e-commerce industry, and urban population of high density development, making, whether business or personal, for the increasing demand for cargo transportation of goods, Transport of goods towards the fast and convenient ways to enhance the efficiency, More domestic industry, the daily work of people into the car, the station has the resources to very large, vehicle planning, vehicle dispatch, and the temporary additional vehicle trips Scheduling, often due to daily changes in the variables that produce different results; in the absence of complete and meet the needs of industry decision-making system of of cases, businesses who rely on traditional training methods, training of experts with the ability to think flexibly To deal with change. However, training of experts, often encounter many problems, such as personnel and suitability for different adjustment issues, experience sharing and knowledge is hard to establish, difficult to assess the quality of decision making ... and so on, so how to build the knowledge and experience of experts in the computer, With the inference that the computer system in a similar way to solve the problem of experts on transport planning field trips to provide advice or answer questions, and can explain the inference result, the industry is still looking for ways. The operational and Vehicle Routing Planning problems on Route-permitted Transportation industry are often mentioned in academic literature, however, the frequency shifts calculation methods related to multi-point deduction from mathematical programming, diverse and need for practical ways to manage the processing variables, are difficult to have Regular solution; With the gradual development of information technology to modernize the computer artificial intelligence to replace the cumbersome and need to learn and record the work can really solve the problem of the past can not solve the complex; so as to establish a mechanism or algorithm Approach can effectively solve the practical problem .