Optimization of a Trapezoidal Fuzzy and Time-based Multi-objective Optimization Problem for yi-out-of-ni Systems
Satyajit De, An Bikash Chowdhury · 2024
Time is an important parameter to analyse the cost and reliability of a reliability redundancy allocation model (RRAM). In this study a trapezoidal fuzzy and time based multi-objective reliability redundancy allocation problem (TFT-MORRAP) is proposed for the m level y i − out − of –n i series parallel system. Maximization of system reliability with minimization of system cost according to time through optimization of level wise redundant components by maintaining some constraints are the main objectives of this study. Maximum entropy constraints with bounded redundant components in each level follows to achieve the objectives. Trapezoidal Fuzzy Numbers (TRFN) come into play to manage the vagueness surrounding both component cost and reliability, effectively addressing their uncertainties. The correlation between component cost and reliability reduction corresponds to the arc length of the inverse logarithmic spiral as it varies with the angle, akin to the concept of time in this context. Over-speed protection system of a gas turbine (OPSGT) is used to analyse our introduced problem. Satisfaction level of peak solutions achieved through the utilization of the well-known algorithms, Multi-objective particle swarm optimization (MOPSO) and Non-dominated sorting genetic algorithm-II (NSGA-II), are compared with a focus on time. Several performance metrics are employed to evaluate diverse algorithmic performances, aiming to identify the optimal solutions for the previously mentioned issue. In the majority of instances, MOPSO exhibits superior performance compared to NSGA-II.