Answers are grounded in this paper and cite their sources.
Agricultural planning for short-cycle crops involves integrating planting and harvesting decisions under temporal and spatial variability in production. This study develops an integrated decision-support framework for transient multi-crop planning that combines complementary production forecasting mechanisms with a multi-objective mixed-integer linear programming model. Two forecasting approaches are evaluated: dynamic time warping coupled with linear regression (DTW-regression), which identifies temporal relationships in historical production patterns, and seasonal autoregressive integrated moving average (SARIMA(X)) models, which capture seasonal dynamics. The resulting production estimates are incorporated into the optimization model, which simultaneously maximizes total production and minimizes the number of active harvesting fronts. The multi-objective problem is solved using an ε -constraint approach, enabling the analysis of compromise solutions between production and harvesting requirements. Computational experiments with progressively larger instances, reaching 2.38 million variables and 696,198 constraints, were conducted to assess the behavior of the formulation and compare the forecasting approaches. The results show a consistent trade-off between production and harvesting fronts, with intermediate solutions providing substantial production levels while requiring fewer harvesting fronts than production-maximizing solutions. DTW-regression generally yields higher production levels, whereas SARIMAX produces comparable harvesting-front requirements. The computational results also indicate that the mathematical formulation remains tractable for the evaluated instances, with MIP gaps below 1% in most cases. The proposed framework provides an integrated basis for analyzing forecasting-driven production decisions and operational trade-offs in transient multi-crop agricultural planning.
No takes yet. Share an insight, caveat, or question.
Supo et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: