AI-Based Multi-Objective Optimization Algorithm for Intelligent School Lunch Menu Planning System

This study developed an AI-based intelligent menu planning system that integrates big data analytics and nutritional expertise to automate the generation of school lunch menus that meet nutritional standards. The system's algorithm consists of four key processes: (1) data preprocessing, (2) menu reshuffling, (3) menu extraction, and (4) menu optimization. Historical menus are reorganized and selected through multi-criteria decision analysis (MCDA), and further optimized using an optimization-based recommendation model (OMRM) and multi-objective optimization to ensure appropriate portions of the six major food groups. A field study involving 10 schools demonstrated that the system significantly aids nutritionists and lunch administrators in generating balanced menus while reducing planning time by an average of 20 minutes. Additionally, it was confirmed that the menus generated by the system were more nutritionally balanced compared to those created manually by experienced dietitians, particularly in achieving the correct portion sizes for the six major food groups.