Multi-objective Linear Approach for the Reliability-Redundancy Allocation Problem

This paper goes through the reliability-redundancy allocation problem in its multi-objective form, maximizing the reliability, and minimizing cost, restricted by design limits. The motivation is to adapt a previously proposed problem formulated as a complex, continuous, and nonlinear model, to a simpler linear and discrete model based on some usual market assumptions. Unlike previous works that have used an unlimited set of suppliers, a limited set of components were chosen, each one with a specific value of reliability. All the study is done in a scenario of one system with ten subsystems from a pharmaceutical plant. After that, the Linear Programming technique CPLEX solver is applied, together with an epsilon-restricted multi-objective approach to solve the presented formulation. The purpose is to solve the reliability-redundancy allocation problem inspired in previous works but adapting it to such specific conditions with a limited number of suppliers and components. The paper brings a final solution selected from the Pareto Front, with specific reliability and cost, and goes deeper, describing how many components are needed for each subsystem and from which supplier.