Optimization of the Dependability of a Hammer Mill Under Budgetary Constraints Using Evolutionary Algorithms, Stochastic Methods, and Statistical Performance Assessment

ABSTRACT This study proposes a hybrid evolutionary–stochastic framework for optimizing the dependability of an industrial hammer mill under explicit budgetary constraints. The approach integrates stochastic degradation modeling (Weibull for hammer wear, discrete‐time Markov chains for rotor deterioration, and Non‐Homogeneous Poisson Processes for failure intensity), constrained genetic algorithm optimization, and Monte Carlo–based objective evaluation within a unified decision‐support architecture. A composite dependability index combining reliability, availability, maintainability, and safety (RAMS) is maximized subject to a predefined maintenance budget reflecting realistic industrial planning conditions. To ensure scientific rigor, the optimization performance is evaluated through independent‐run statistical assessment, convergence stability analysis, budget trade‐off investigation, and sensitivity testing with respect to failure‐law parameters. Results demonstrate statistically stable convergence, economically rational trade‐off behavior with an identifiable knee region in the budget–performance curve, and limited sensitivity to moderate parameter uncertainty. Comparative benchmarking against representative state‐of‐the‐art reliability optimization approaches highlights the contribution of the proposed framework in terms of methodological integration, statistical validation, and industrial interpretability. The findings confirm that the hybrid constrained formulation provides a robust and practically deployable solution for maintenance planning of wear‐driven industrial systems operating under financial and operational uncertainty.