Back to search

Clemson University

Data-driven Stochastic Programming for Disaster Housing Logistics Planning

Abstract

dc:description.abstract

This dissertation develops an integrated modeling and solution framework for direct temporary disaster housing logistics under demand uncertainty. The proposed framework addresses the problem from short-term, long-term, and computational perspectives. First, a two-stage chance-constrained stochastic programming (TSCC) model is developed for short-term housing logistic planning, with demand scenarios generated via a data-driven spatial regression model that captures the relationship between housing demand, hazard exposure, and socioeconomic factors. A case study based on Hurricane Ian demonstrates that the TSCC model outperforms deterministic and traditional scenario-based models in both solution quality and robustness. Second, to address long-term housing logistics planning, a multi-horizon stochastic programming (MHSP) model is proposed. This formulation captures short-term uncertainty through regression-based demand estimation and long-term uncertainty via time series analysis. Numerical results validate the MHSP model's ability to improve planning performance under uncertainty compared to conventional approaches. Finally, the framework is extended to a periodic multi-stage stochastic programming (PMSP) model that leverages the periodic nature of disaster occurrences to improve computational efficiency. Computational experiments using historical hurricane data demonstrate that the PMSP model achieves comparable solution quality to standard multi-stage stochastic programming while significantly reducing computational time.

Degree

thesis:*
Name thesis:degree_name
Doctor of Philosophy (PhD)
Level thesis:degree_level
Dissertation
Discipline thesis:degree_discipline
Industrial Engineering

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Chen, Sheng-Yin
Contributors dc:contributor
  • Yongjia Song
  • Qi Luo
  • Weichiang Pang
  • Thomas Christos Sharkey
  • Emily Tucker

Subjects

dc:subject × 1

Identifiers

dc:identifier.*
Repository record dc:identifier
https://open.clemson.edu/all_dissertations/4029

Chain of custody

source
Harvested from
Clemson University
Base URL
open.clemson.edu/do/oai/
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Chen, Sheng-Yin. Data-driven Stochastic Programming for Disaster Housing Logistics Planning. Dissertation thesis, https://open.clemson.edu/all_dissertations/4029