Multi-objective network optimization: models, methods, and applications
Metadata[+] Show full item record
There can be an array of planning objectives to consider when identifying alternatives for using, modifying, or restoring natural or built environments. In this respect, multi-objective network optimization models can provide decision support to both managers and users of the system. While there can be an infinite number of feasible solutions to any multi-objective optimization problem in large networks (e.g., urban transportation systems), the efficient ones are usually more desirable in the decision-making process. However, identification of efficient solutions can be challenging in practical applications. To address this issue, this dissertation details mathematical formulations and solution algorithms for a range of real-world planning problems in the context of intelligent transportation systems, vehicle routing problem, natural conservation and landscape connectivity. While the combination of objectives being optimized is unique for each application, the underlying phenomena involves modeling movement between origins and destinations of a networked system. To demonstrate the type of insights that can be achieved using these modeling approaches, the location and number of times solutions appear in different realizations of system and given different solution approaches (e.g., exact and approximate methods) are visualized on network using a commercial geographic information system.