A structured framework for performance optimization using JBLTO, FCOPRAS and FCODAS methodologies

• Published in: Journal of quality in maintenance engineering 2023-03, Vol.29 (1), p.220-248
• Main Authors: Gopal, Nand, Panchal, Dilbagh
• Format: Article
• Language: English
• Subjects: Availability; Cavitation; Dairy industry; Electric motors; Engineering; Failure; Failure analysis; Failure modes; Gearboxes; Genetic algorithms; Impellers; Milk; Optimization; Particle swarm optimization; Performance evaluation; Profitability; Pulp & paper industry; Pulp & paper mills; Qualitative analysis; Reliability analysis; Risk assessment; Robustness (mathematics); Sensitivity analysis; Software quality; Subsystems; Uncertainty
• ISSN: 1355-2511; 1758-7832
• DOI: 10.1108/JQME-11-2021-0087

PurposeThe proposed hybridized framework provides a new performance optimization-based paradigm for analysing the failure behaviour of paneer unit (PU) in the dairy industry.Design/methodology/approachA novel fuzzy Jaya-based Lambda–Tau Optimization (JBLTO) approach-based mathematical modelling was developed for calculating various reliability indices of the considered unit. Failure mode and effect analysis (FMEA) was carried using qualitative information gathered from system's expert opinions. Fuzzy-complex proportional assessment (FCOPRAS) approach was integrated within FMEA to recognize the most critical failure causes associated with various subsystem/components.FindingsThe availability of the unit falls by 0.053% as the uncertainty level increases from ±15 to ±25% and further decreases to 0.323% as the uncertainty level increases from ±25 to ±60%. Failure causes, namely wearing in gears of gearbox (MST4), an impeller's cavitation and/or corrosion (CFP4), winding failure of electric motor (WS9), were recognized as the most critical failure causes with FCOPRAS final performance scores of 100, 100 and 100 and fuzzy combinative distance-based assessment (FCODAS) resultant assessment score of 0.5997, 1.1898 and 1.6135.Originality/valueJBLTO approach-based reliability results were compared with traditional particle swarm optimization-based Lambda–Tau (PSOBLT) and traditional fuzzy Lambda–Tau (FLT) approaches for confirming the downward trend in the system's availability. The ranking results of qualitative analysis are compared with the implementation of FCODAS technique. Sensitivity analysis was executed to evaluate the robustness of the proposed hybridized framework.