Integrating Multi-Criteria Decision-Making with Multi-Objective Optimization for Sustainable Diet Design

Abstract

We present an approach to optimize diet sustainability by combining multi-criteria decision-making (MCDM) with multi-objective optimization (MOO). A sustainable diet must balance cultural acceptability, nutritional adequacy, and environmental sustainability. However, a single food group may perform well in one indicator but poorly in others, necessitating the inclusion of multiple indicators to achieve a truly sustainable diet. This, in turn, increases the complexity of the optimization process and the interpretation of its results. To address this challenge, we applied the SURE method as an MCDM tool before MOO to reduce the number of objectives. The SURE score can integrate multiple environmental indicators, capturing their conflicting characteristics and simplifying the optimization problem. The proposed method was applied to optimize the Estonian diet. Estonian food consumption was categorized into 14 groups, and footprint data with uncertainty ranges were collected for analysis. A bi-objective optimization problem was formulated to simultaneously minimize five aggregated environmental footprints and deviations from the reference diet while satisfying nutritional constraints. For comparison, a classical multi-objective optimization approach was also implemented. The results demonstrated that both approaches successfully reduced all environmental impacts. However, the bi-objective optimization offered a more straightforward decision-making process, allowing for the visual representation of results and easier adjustments to objective weights based on decision-maker preferences. This method facilitates the design of sustainable diets by streamlining complex trade-offs and providing a clear framework for informed decision-making.