Designing resilient lifeline networks a primal-dual optimization approach

Chen, Zhijie

Permalink

https://hdl.handle.net/2142/120161 Copy

Description

Title

Designing resilient lifeline networks a primal-dual optimization approach

Author(s)

Chen, Zhijie

Issue Date

2023-05-02

Director of Research (if dissertation) or Advisor (if thesis)

Banerjee, Arindam

Department of Study

Computer Science

Discipline

Computer Science

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

M.S.

Degree Level

Thesis

Keyword(s)

• non-smooth optimization
• network design
• resilience

Abstract

The design of lifeline infrastructures ranging from transportation to supply chain networks has traditionally emphasized operational efficiency. However, growing natural, made-made, and technological threats motivate embedding resilience principles for the prevention of and recovery from disruptions. In this paper, we propose new optimization based formulations for assessing and enhancing resilience in lifeline networks. Resilience is measured as the imbalance of supply and demand across network nodes after post-disruption restoration of the network under a budget. We present a novel primal-dual optimization algorithm for the formulation which fully leverages the non-smooth problem structure. The algorithm adopts a single-loop update and avoids routines to solve subproblems or projecting onto intersec tions of (convex) constraint sets, and can be further improved by randomized smoothing. Theoretical analysis is provided for the convergence of the proposed algorithm, including its smoothed variant. Empirical results show the efficiency of the proposed optimization as well as the effectiveness of our formulation for resilience assessment and resilient network design.

Graduation Semester

2023-05

Type of Resource

Thesis

Copyright and License Information

Copyright 2023 Zhijie Chen

Owning Collections