| 摘要: | 近年來，預拌混凝土已成為公共建設或私人建築之主要建材之一，而其生產方式係於預拌混凝土廠經由電腦化作業後，利用預拌混凝土拌合車裝載運送到工地卸料與澆置。值得一提的是，當預拌混凝土廠發生機具臨時故障時，會因為無法出貨導致預拌混凝土廠無法有效供應預拌混凝土需求，對於預拌混凝土廠與工地雙方的成本有一定程度的影響，然而業者目前對於預拌混凝土廠故障時的因應方式，是以人工方式調整原先的拌合車派遣規劃，未以系統最佳化的觀點調整排班，容易導致資源浪費及降低服務水準的情形。緣此，本研究針對拌合廠臨時故障時的拌合車派遣規劃問題進行探討，利用該廠商在該地區的現有資源進行支援供應，基於預拌混凝土廠商的立場，考量拌合廠與工地之整體性營運成本，並將產業營運特性納入考量，系統最佳化的觀點建構二個即時性拌合車派遣規劃模式。本研究首先構建系統中使用單一車種之即時性單車種拌合車派遣規劃模式，再進一步以此模式為基礎，構建使用不同載運量拌合車之即時性多車種拌合車派遣規劃模式，以幫助業者在拌合廠發生臨時故障時，以更有效並且更具彈性的資源使用方式調整拌合車派遣規劃，以期降低損失並維持服務品質。 本研究利用時空網路流動技巧構建即時性單車種與多車種拌合車派遣規劃模式，以定式車輛在時空中的流動情況，此二模式可定式為含額外限制之混合整數規劃網路流動問題，係屬NP-hard問題。在求解方法上，利用C++程式語言配合數學規劃套裝軟體CPLEX求解模式，並且為求有效地求解，本研究發展一以問題分解策略為基礎之啟發解法。最後，為測試模式與啟發解法之實用性，本研究以國內北區一家預拌混凝土廠商之實際營運資料進行實例測試與分析，結果顯示本研究所發展之模式與啟發解法，可有效率地求解並得到良好的預拌混凝土生產與車輛派遣規劃。 In recent years, ready mixed concrete (RMC) has became one of the most primary construction material of buildings and other public infrastructures. It is produced by automatic process and computer controlled process, and then delivered by RMC trucks. However, when RMC mixer is breakdown unexpectedly, RMC batch might not satisfy the demand of construction sites. Consequently, the operation cost of both RMC batch and construction sides would be increased. To deal with this situation, the dispatcher would manually adjust the original production scheduling and truck dispatching by experienced staff. It might lead to a result of wasting resources and decreasing the quality of service. Therefore, in this research, we aimed at dealing with the situation when RMC mixer is breakdown unexpectedly. To support the supplying, RMC is supplied by using all the applicable resources of the company in the specific area. By angle of owners of RMC Companies, we considered the operation costs of both RMC batch and construction sides. Besides, we also considered the operation characteristics of this industry. In this research, we systematically constructed two models. We first constructed the real time RMC truck dispatching with single type of trucks. Further, based on the first model, we constructed the real time RMC truck dispatching with various types of trucks. These two models are expected to be useful planning tools for helping dispatchers adjust truck dispatching more effectively, and thus to minimize loss of system profits and to maintain their quality of service. We applied time-space network techniques with the system optimization perspective to construct two real-time truck dispatching models. These two models are formulated as mixed integer network flow problems with side constraints, which are characterized as NP-hard. To solve realistic problems efficiently, we developed a solution method, which incorporates a mathematical programming solver CPLEX. To solve the problems more efficiently, we used problem decomposition techniques to develop the heuristic to solve the problems. In order to test the models and solution algorithms in actual operations, we performed a case study based on the real operating data from an RMC firm in Taiwan. The test results show that the models and the solution algorithm would be better then actual operations. |
