Teodor Gabriel Crainic (editor) - Network Design with Applications to Transportation and Logistics

This Springer imprint is published by the registered company Springer Nature Switzerland AG

The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Network design problems arise whenever optimal choices have to be made that can be represented conceptually as the selection of a subset of links in a graph. Typically, these optimal choices are the result of complex tradeoffs between various types of costs and constraints. In particular, most network design problems involve fixed costs associated with link selection and variable costs associated with flows (of people, goods, information,). Because of their combinatorial nature and the complexity of their objective functions and constraints, network design problems are inherently difficult (most of them are -hard). For this reason, models and algorithms for network design problems involve approaches from several areas of combinatorial optimization and mathematical programming.

In this book, we study network design problems, as well as models and algorithms to solve them, through techniques derived from network optimization, linear programming (LP), mixed-integer linear programming (MILP), metaheuristics, and large-scale optimization. The book is intended to be useful not only to researchers specialized in these areas, but first and foremost, to graduate students in operations research, management science, analytics, and related fields. A special effort was made by all authors involved in the different chapters to make the exposition didactic. In particular, to facilitate reading, the main body of each chapter contains very few citations. Instead, a section on bibliographical notes is included at the end of each chapter, helping the reader to summarize the content of the chapter, to get an appreciation of the historical development of the particular topic addressed in the chapter, and to look for additional references. We assume that the reader is familiar with the basics of LP and MILP, including an exposition to linear network flow problems. More advanced topics, such as mathematical decomposition (including Lagrangian duality and Benders decomposition) and metaheuristics, are reviewed in Chaps. , although they are only covered in enough details to make the exposition self-contained. The reader who is less familiar with these topics would benefit from reading the articles and books that are referenced in these chapters.

Network design problems are prominent in transportation and logistics, but also in other areas, such as telecommunication and manufacturing. This book is about applications in transportation and logistics, although many problems, models and algorithms can be adapted to other areas. In fact, the first ten chapters of the book are dedicated to the network design methodology, common to most problem settings and applications, and many application-oriented chapters contain material that is highly relevant to other areas than transportation and logistics. Even though we purposely restricted ourselves to transportation and logistics, covering all applications of network design in that area is impossible and we had to make some difficult choices. In particular, the broad areas of facility location and supply chain management yield optimization problems that can be cast as the selection of a subset of nodes in a graph. By duplicating nodes and by introducing links between each node and its copy, such problems can be seen as involving the selection of a subset of links, hence as network design problems. As such, they are covered in some chapters of the book, but not to the same extent as they are in books dedicated to facility location and supply chain management.