Research on Crop Planting Scheme Based on PSO-BP Neural Network

Abstract

This study explores agricultural optimization through two main questions. First, a PSO-BP neural network model was used to derive an optimal planting strategy for profit maximization and cost minimization. The model incorporated yield gap, planting area, and planting practices constraints, resulting in stable, low errors, with a maximum profit of 3,561,194.9 yuan in 2030. For the second question, an optimization model was developed considering expected sales, crop yield, planting cost, sales price, and non-negativity constraints. The model was solved using the simulated annealing algorithm, achieving stability after 35 iterations. The maximum profit reached 4,985,602.86 yuan in 2030, with a minimum of 3,913,304.33 yuan in 2023. This study provides insights into optimizing agricultural production for improved profitability.