Research on multi-algorithm fusion and comparison of intelligent farmland planting strategies based on BRBMO-BiLSTM-KAN model prediction data

Intelligent farmland planting strategy is an important topic of intelligent planting. In order to improve the adaptability and accuracy of the strategy in the complex market environment, this paper generated simulated data based on the macro data of the National Bureau of Statistics of China, used BRBMO-BiLSTM-KAN to predict the expected sales volume and carried out cluster analysis, and then adopted the simulated annealing algorithm to obtain the crop planting strategy. Finally, a comprehensive evaluation model based on Entropy Weight-Grey Relation-TOPSIS was used to compare the crop planting strategies obtained by the LPO algorithm under dynamic environment.The results show that the model successfully improves the adaptability and accuracy of intelligent farmland planting strategy in complex market environment, and provides strong technical support for intelligent planting.