网络收益管理中动态定价问题研究
发布时间：2021-09-12 11:02
收益管理被广泛应用在现实生活中,从最开始的航空领域到不同行业下的商业应用,比如酒店管理、流行商品零售。比起收益管理,网络收益管理更加复杂,不同产品会用到同一种资源,企业会在有限周期下销售产品。定价问题和容量控制问题是收益管理中两类重要的问题。在本文中,我们将研究网络收益管理中的动态定价问题。在该问题中,企业在有限库存容量约束和有限销售周期中决定产品的价格来最大化收益,而顾客需求的发生是一个随机过程,发生强度是产品价格的函数。一种处理动态定价问题的有效方法是使用需求均值近似顾客随机需求,然后建立以收益最大化为目标,以资源数量为约束的确定性模型。然而,动态规划通过追踪资源水平随时间的变化而在构建模型上具备独特的优势。在本文中,我们针对动态定价问题建立了动态规划模型并且使用近似动态规划方法来求解该问题。近似动态规划方法首先把动态规划模型转化为线性规划模型,线性规划模型中的决策变量为动态规划中的价值函数,然后使用近似函数近似决策变量,得到近似线性规划。与容量控制问题不同,因为动态定价问题中的决策空间是连续的,该问题下的近似线性规划是半无限线性规划。基于该规划我们提出了列生成算法。因为规划中数量...
【文章来源】：上海交通大学上海市 211工程院校 985工程院校 教育部直属院校
【文章页数】：136 页
【学位级别】：博士
【文章目录】：
摘要
ABSTRACT
Chapter 1 Introduction
1.1 General Framework for Dynamic Pricing Problems
1.2 Dyanmic Pricing Problems under Independent Demand
1.3 Dynamic Pricing Problems under Dependent Demand
Chapter 2 Literature Review
2.1 Static Problems in Revenue Management
2.1.1 Static Problems without Resource Constraints
2.1.2 Deterministic Problems and Resolving
2.2 Dynamic Problems in Revenue Management
2.2.1 Capacity Control Problems
2.2.2 Dynamic Pricing Problems
2.3 Approximate Linear Programming Approach
2.3.1 Approximate Linear Programs
2.3.2 Approaches of solving ALPs
Chapter 3 General Framework for Dynamic Pricing
3.1 Problem Formulation
3.2 General Framework
3.2.1 Strong Duality
3.2.2 A Column Generation Algorithm for (P)?-(D)?
3.3 Unified Formulation
3.3.1 The Linear Programming Based Approximate Dynamic Programming
3.4 A General Scheme to Establish Compact Reformulations
3.4.1 Constraint Reformulation
3.4.2 Illustrative Example:Affine Approximation under Independent Demand
3.5 Applications of the General Scheme in Capacity Control Problems
3.5.1 An Omnibus Linear Relaxation of ?t(x,u,V)
3.5.2 The Separable Piecewise Linear Approximation under Independent Demand
3.5.3 The Affine Approximation under Discrete Choice Model of Demand
3.5.4 The Separable Piecewise Linear Approximation under the Discrete Choice Model of Demand
Chapter 4 Dynamic Pricing Problems under Independent Demand in Network Revenue Management
4.1 Affine Approximation with Linear Independent Demand
4.1.1 Column Generation Subproblem
4.2 A Compact Nonlinear Programming Formulation
4.2.1 Reformulation and Strong Duality
4.2.2 An Alternative Proof for SOCP Duality
4.2.3 Aggregation and Equivalence
4.2.4 Deterministic Nonlinear Programming Formulation and Its Con-nection to (D2)~(A,R)
4.3 Affine Approximation with Log-Linear Independent Demand
4.4 Discretization as an Alternative Solution Strategy
4.4.1 The Formulation with a Discrete Price Set
4.4.2 A Compact Formulation
4.5 Pricing Policies
4.5.1 The Policy DBD
4.5.2 The Policy SBD
4.6 Numerical Study
4.6.1 Computational Setup
4.6.2 Computational Results for the Continuous Formulation
4.6.3 Computational Comparison between Continuous and Discrete Formulations
4.6.4 Comparison of Pricing Policies
Chapter 5 Dynamic Pricing Problems under Dependent Demand in Network Revenue Management
5.1 Problem Formulation
5.1.1 Approximate Dynamic Programming Approach
5.1.2 Constraint Generation Algorithm
5.2 A New Approach
5.2.1 An Improved Constraint Generation Algorithm
5.2.2 A Reduced ALP
5.2.3 Extension to Separable Piecewise Linear Approximation
5.3 Numerical Study
Chapter 6 Conclusion
Bibliography
Acknowledgements
Publications
Resume
本文编号：3394108
【文章来源】：上海交通大学上海市 211工程院校 985工程院校 教育部直属院校
【文章页数】：136 页
【学位级别】：博士
【文章目录】：
摘要
ABSTRACT
Chapter 1 Introduction
1.1 General Framework for Dynamic Pricing Problems
1.2 Dyanmic Pricing Problems under Independent Demand
1.3 Dynamic Pricing Problems under Dependent Demand
Chapter 2 Literature Review
2.1 Static Problems in Revenue Management
2.1.1 Static Problems without Resource Constraints
2.1.2 Deterministic Problems and Resolving
2.2 Dynamic Problems in Revenue Management
2.2.1 Capacity Control Problems
2.2.2 Dynamic Pricing Problems
2.3 Approximate Linear Programming Approach
2.3.1 Approximate Linear Programs
2.3.2 Approaches of solving ALPs
Chapter 3 General Framework for Dynamic Pricing
3.1 Problem Formulation
3.2 General Framework
3.2.1 Strong Duality
3.2.2 A Column Generation Algorithm for (P)?-(D)?
3.3 Unified Formulation
3.3.1 The Linear Programming Based Approximate Dynamic Programming
3.4 A General Scheme to Establish Compact Reformulations
3.4.1 Constraint Reformulation
3.4.2 Illustrative Example:Affine Approximation under Independent Demand
3.5 Applications of the General Scheme in Capacity Control Problems
3.5.1 An Omnibus Linear Relaxation of ?t(x,u,V)
3.5.2 The Separable Piecewise Linear Approximation under Independent Demand
3.5.3 The Affine Approximation under Discrete Choice Model of Demand
3.5.4 The Separable Piecewise Linear Approximation under the Discrete Choice Model of Demand
Chapter 4 Dynamic Pricing Problems under Independent Demand in Network Revenue Management
4.1 Affine Approximation with Linear Independent Demand
4.1.1 Column Generation Subproblem
4.2 A Compact Nonlinear Programming Formulation
4.2.1 Reformulation and Strong Duality
4.2.2 An Alternative Proof for SOCP Duality
4.2.3 Aggregation and Equivalence
4.2.4 Deterministic Nonlinear Programming Formulation and Its Con-nection to (D2)~(A,R)
4.3 Affine Approximation with Log-Linear Independent Demand
4.4 Discretization as an Alternative Solution Strategy
4.4.1 The Formulation with a Discrete Price Set
4.4.2 A Compact Formulation
4.5 Pricing Policies
4.5.1 The Policy DBD
4.5.2 The Policy SBD
4.6 Numerical Study
4.6.1 Computational Setup
4.6.2 Computational Results for the Continuous Formulation
4.6.3 Computational Comparison between Continuous and Discrete Formulations
4.6.4 Comparison of Pricing Policies
Chapter 5 Dynamic Pricing Problems under Dependent Demand in Network Revenue Management
5.1 Problem Formulation
5.1.1 Approximate Dynamic Programming Approach
5.1.2 Constraint Generation Algorithm
5.2 A New Approach
5.2.1 An Improved Constraint Generation Algorithm
5.2.2 A Reduced ALP
5.2.3 Extension to Separable Piecewise Linear Approximation
5.3 Numerical Study
Chapter 6 Conclusion
Bibliography
Acknowledgements
Publications
Resume
本文编号：3394108
本文链接：https://www.wllwen.com/shoufeilunwen/jjglbs/3394108.html
