Research on Optimization of Regional Agricultural Planting Strategy Based on Sequential Mixed-Integer Linear Programming

Abstract

In the context of global food security and rural revitalization strategies, developing efficient and sustainable modern agriculture has become a core issue in the agricultural sector. This paper addresses the optimization of agricultural production under specific geographical and market conditions by constructing a multi-year, multi-constraint Sequential Mixed Integer Linear Programming (SMILP) model. The model explores optimal arable land resource allocation schemes that balance economic benefits and ecological sustainability. With long-term economic returns maximization as the objective function, it systematically integrates crop sales revenue, planting costs, and implicit management costs arising from dispersed planting. At the same time, it rigorously incorporates constraints such as plot-crop suitability matching, annual crop rotation systems, and legume crop soil fertility maintenance. To quantify the impact of market risks on planting decisions, the model innovatively sets up two scenarios: "complete unsaleability" and "discounted sales," simulating uncertainties in the agricultural product market. Experimental results demonstrate that the model can dynamically generate annual planting plans for the 2024-2030 planning period, effectively balancing economic benefits and ecological constraints. Comparative analysis shows that the "discounted sales" strategy, with flexible sales channels, can significantly enhance overall profitability and reduce operational risks, providing a scientific and practical intelligent decision-making tool for agricultural production decisions. The findings of this study hold significant theoretical value and practical guidance for the intelligent transformation and sustainable development of modern agricultural management.