A Novel Algorithmic Framework for Optimal Reliability and Cost Allocation in Constrained Complex Systems

Abstract

n order to solve the problem of reliability and cost allocation in complex engineering

systems that are subject to operational and financial constraints, new algorithmic

solutions are developed and presented in this research. Specifically, two different

optimization methodologies are developed and compared in the present paper: the first

one is a Genetic Algorithm (GA) that has the ability to efficiently search in large spaces,

while the second one is a Dynamic Programming (DP) solution that has the ability to

solve the problem optimally by exploiting the optimal substructure of the problem.

Furthermore, the methodologies are implemented in Python in a modular and reusable

way, and different tests are performed in order to demonstrate that the Genetic

Algorithm has the ability to solve the problem even in larger cases, while the Dynamic

Programming has the ability to solve the problem optimally in smaller cases, including

series, parallel, series-parallel, and bridge systems. The trade-offs between scalability,

computational cost, and solution quality are validated by numerical findings. To help

practitioners choose the best approach for their particular system settings, the paper

offers theoretical underpinnings, pseudocode, complete Python source code, and

comparative tables.