This study addresses the challenges of limited arable land resources and uncertain planting risks in the context of rural revitalization by exploring optimization pathways for crop planting strategies. The research preprocesses historical planting data to identify core constraints such as crop characteristics, rotation patterns, and market demand, then constructs a linear programming model aimed at maximizing farm operating profits. To address computational ineffi-ciency in large-scale combinatorial optimization, a Genetic Algorithm with Greedy Initialization Strategy is proposed. Experimental results demonstrate that this approach significantly enhances convergence efficiency, achieving stable con vergence within 15 generations, a substantial improvement over traditional random initialization methods. Furthermore, a dynamic stochastic programming model is introduced to account for yield and price fluctuations. Through sensitivity analysis, the study identifies the optimal minimum planting area threshold to be between 0.30 and 0.45 mu, which balances management convenience with economic benefits. Numerical simulations indicate that under deterministic conditions, the optimal profit over the 2024 2030 period reaches 32.05 million yuan and 52.46 million yuan. In dynamic stochastic scenarios, the model mitigates risks effectively, yielding optimal profits of 32.40 million yuan and 43.61 million yuan respectively. This research not only provides an accurate mathematical model for maximizing agricultural production efficiency but also offers robust decision support for the sustainable development of rural economies.

genetic algorithm, stochastic programming, crop strategy optimization