|
In many practical situations, we need to reduce the cost of the system and improve its reliability simultaneously. At the same time, all the design data involved in the system design are not precisely known. Incompleteness and unreliability of input information are typical for many practical problems in the multi-objective optimization of system design. In this work, we have analyzed fuzzy multi-objective optimization problem of main characteristics of system design such as reliability and cost simultaneously based on non-dominated sorting genetic algorithm-II (NSGA-II). NSGA-II is one of the multi-objective evolutionary algorithms (MOEAs), provides the decision-maker with a complete picture of the Pareto-optimal solutions space. It finds increasing applications in solving the multi-objective optimization problem (MOOP) because of low computational requirements, elitism, and parameter-less sharing approach. A brief description of NSGA-II and its use for MOOP is given. We have obtained multiple solutions (Pareto-optimal solutions) in a single simulation run. A numerical example of a series system is given to illustrate the proposed approach.
|
