隨著都市化發展與車輛持有量增加，都市地區停車空間供需失衡問題日益嚴重。為提升既有停車資源之使用效率，共享停車逐漸成為停車場經營與管理上可考量之新興營運方式。近年來，共享停車平台與停車場業者合作，將場內閒置車位整合至平台，提供使用者事先預約車位之服務，以增加停車選擇並提升車位使用彈性。然而，當停車場同時服務月租、臨時停車與預約共享停車等不同需求時，其車位配置與指派問題較單一營業模式更為複雜，若進一步考量臨時停車與預約共享停車需求之隨機波動，則更需要透過系統化方法進行規劃與評估。 本研究以公共停車場委外經營業者之立場，針對月租、臨時停車與預約共享停車三種混合營業模式，建構隨機需求環境下之停車位配置模式，並以最大化營運收入為目標，運用網路流動與數學規劃方法描述車位服務與接續關係。為反映實際需求變動，本研究以既有調查資料為基礎，透過截斷式常態分配與機率式倍率擾動產生隨機需求情境，並結合 C++ 程式語言、CPLEX 數學規劃軟體與啟發式求解演算法進行求解。同時，本研究採用完全資訊期望價值（EVPI）與隨機解價值（VSS）評估模式績效，並透過敏感度分析探討收費價格及停車需求量對營運結果之影響。研究結果顯示，本模式能在考量需求不確定性下，有效配置三種營業模式之車位資源，並兼顧停車需求滿足與營運收入表現，可作為停車場經營業者進行車位配置、收費策略調整與營運規劃之決策參考。;With the development of urbanization and the increase in vehicle ownership， the imbalance between parking supply and demand in urban areas has become increasingly severe. To improve the utilization efficiency of existing parking resources， shared parking has gradually emerged as a potential operating model for parking lot management. In recent years， shared parking platforms have cooperated with parking lot operators to integrate idle parking spaces into platform-based services， allowing users to reserve parking spaces in advance， thereby increasing parking options and enhancing the flexibility of parking space utilization. However， when a parking lot simultaneously serves different types of demand， including monthly parking， temporary parking， and reservation-based shared parking， the parking space allocation and assignment problem becomes more complex than that under a single operating model. When stochastic fluctuations in temporary parking and reservation-based shared parking demand are further considered， a systematic planning and evaluation method is required. From the perspective of an outsourced public parking lot operator， this study develops a parking space allocation model under stochastic demand for a hybrid operating model that includes monthly parking， temporary parking， and reservation-based shared parking. With the objective of maximizing operating revenue， network flow techniques and mathematical programming are employed to describe the service and connection relationships of parking spaces. To reflect variations in actual demand， this study uses existing survey data as the basis for generating stochastic demand scenarios through a truncated normal distribution and probabilistic multiplier perturbation. The model is solved using the C++ programming language in combination with CPLEX optimization software and a heuristic solution algorithm. In addition， the Expected Value of Perfect Information (EVPI) and the Value of the Stochastic Solution (VSS) are adopted to evaluate model performance， and sensitivity analyses are conducted to examine the effects of parking fees and parking demand levels on operational results. The results show that， under demand uncertainty， the proposed model can effectively allocate parking space resources among the three operating modes while accounting for both parking demand satisfaction and operating revenue performance. Overall， the proposed model can serve as a decision-support reference for parking lot operators in parking space allocation， pricing strategy adjustment， and operational planning.
