چكيده لاتين
With the significant increase in the consumption of cosmetics and health products, particularly sunscreens, the design and optimization of the supply chain for these products have become major concerns for related industries. A sustainable supply chain, especially in the case of products with direct impacts on health and the environment, such as sunscreens, requires comprehensive and scientific solutions that balance cost reduction, product quality improvement, and environmental impact mitigation. In this research, product demand has been predicted based on historical data analysis using machine learning techniques. Subsequently, a mathematical model has been designed to manage the supply, production, and distribution of sunscreens. In the proposed model, supplier selection has been examined based on quality levels, the location of integrated centers has been determined, and the optimal allocation of resources in the production and distribution processes has been analyzed. After extracting the set of Pareto solutions, two clustering methods, K-means and hierarchical clustering, have been employed to reduce and classify Pareto points, facilitating decision-making. Furthermore, to account for uncertainty in product demand, a robust stochastic optimization approach has been applied. The model has been validated using data from a case study in Iran within the GAMS software, and the results have been analyzed over multiple time periods. The results obtained through the reinforced epsilon constraint method indicate that optimization of the cost minimization objective function has led to a 54.96% reduction in the quality-dependent value maximization objective function and a 0.48% reduction in the emission minimization objective function. In contrast, optimization of the quality-dependent value maximization objective function has resulted in a 0.52% increase in the cost minimization objective function while maintaining stability in the emission minimization objective function. Lastly, optimization of the emission minimization objective function has led to a 0.52% increase in the cost minimization objective function and a 14.97% decrease in the quality-dependent value maximization objective function. The findings demonstrate that the combined approach employed enables cost reduction, product quality enhancement, and environmental emissions mitigation, offering a robust decision-making tool for different managerial levels.
