MULTIOBJECTIVE EVOLUTIONARY OPTIMIZATION OF OBSERVED BRAZILIAN MEALS ACROSS DIETARY REFERENCE PROFILES
This study examined whether menus assembled from observed Brazilian meals could improve nutritional and environmental outcomes while accounting for changes in food quantities. Four evolutionary algorithms were compared across six objective combinations using 2017–2018 Brazilian Household Budget Survey records for women and men aged 30 and 60 years, with weight-maintenance requirements. Adjusted hypervolume comparisons did not identify differences among the algorithms. However, choosing a menu from the resulting alternatives changed the dietary outcomes. Selection by distance to the ideal point of the searched objectives raised mean attainment of nutrient recommendations from 75.78% in observed diets to 97.98% and lowered carbon footprint, but increased water footprint. A retrospective rule requiring simultaneous improvements in nutrition, carbon, water and departure from observed quantities found an eligible menu in every run when it used the full candidate histories. Those choices averaged 95.67% attainment, with carbon and water 58.43% and 58.98% below observed profile means. Milk-based beverage and mate infusion quantities varied with the objectives and selection rule. Menu analysis showed that mean improvements coexisted with nutrient gaps, distinguishing front performance from the foods and quantities actually selected.
Authors
- Ana Luiza Barros (ORCID: https://orcid.org/0000-0002-5018-3031)
- Thiago Queiroz de Oliveira (ORCID: https://orcid.org/0000-0001-6500-3729)
- Bonfim Amaro
- Gustavo Augusto Lima de Campos (ORCID: https://orcid.org/0000-0001-7175-9071)
- Janaide Nogueira de Sousa Ximenes (ORCID: https://orcid.org/0000-0002-5676-577X)
- Thalyson Gomes Nepomuceno da Silva (ORCID: https://orcid.org/0000-0003-4258-2945)
- Rhyan Ximenes de Brito
Institutions
- Universidade Estadual do Ceará (BR)
- Instituto Federal de Educação, Ciência e Tecnologia do Ceará (BR)
- Instituto Federal de Educação, Ciência e Tecnologia do Pará (BR)
Publication Details
- Journal
- RECIMA21 - Revista Científica Multidisciplinar
- Published
- 2026-11-17
- DOI
- https://doi.org/10.47820/recima21.v7i11.9654
- Primary Topic
- Agriculture Sustainability and Environmental Impact
- Type
- article
- Field-Weighted Citation Impact
- 0.00