A multi-objective reliability-redundancy allocation problem with active redundancy and interval type-2 fuzzy parameters

TL;DR

This paper addresses a multi-objective reliability-redundancy allocation problem for series-parallel systems, modeling component reliabilities with interval type-2 fuzzy numbers to handle uncertainty, and applying multiple optimization methods to find compromise solutions.

Contribution

It introduces a novel approach using interval type-2 fuzzy parameters for reliability estimation and compares five multi-objective optimization techniques for solving the problem.

Findings

Interval type-2 fuzzy modeling effectively captures uncertainty in reliability data.

Multiple optimization methods provide diverse trade-off solutions.

Numerical example demonstrates practical applicability in a pharmaceutical plant.

Abstract

This paper considers a multi-objective reliability-redundancy allocation problem (MORRAP) of a series-parallel system, where system reliability and system cost are to be optimized simultaneously subject to limits on weight, volume, and redundancy level. Precise computation of component reliability is very difficult as the estimation of a single number for the probabilities and performance levels are not always possible, because it is affected by many factors such as inaccuracy and insufficiency of data, manufacturing process, environment in which the system is running, evaluation done by multiple experts, etc. To cope with impreciseness, we model component reliabilities as interval type-2 fuzzy numbers (IT2 FNs), which is more suitable to represent uncertainties than usual or type-1 fuzzy numbers. To solve the problem with interval type-2 fuzzy parameters, we first apply various type-reduction and defuzzification techniques, and obtain corresponding defuzzified values. As maximization of system reliability and minimization of system cost are conflicting to each other, so to obtain compromise solution of the MORRAP with defuzzified parameters, we apply five different multi-objective optimization methods, and then corresponding solutions are analyzed. The problem is illustrated numerically for a real-world MORRAP on pharmaceutical plant, and solutions are obtained by standard optimization solver LINGO, which is based on gradient-based optimization - Generalized Reduced Gradient (GRG) technique.