Design And Implementation Of Fertilizer Spreader Based On I.MX6UL

In agricultural production,outdated fertilization facilities and extensive management systems have led to high labor costs,wastage of fertilizer resources,and improper fertilizer application,greatly limiting the development of agriculture.This article proposes a fertilizer machine based on the i.MX6 ULL to address these issues,with the following main research content and results:First,to address the limitations of single sensor measurements,a two-stage data fusion algorithm is proposed to fuse multiple sensor data and provide accurate inputs for the fertilizer prediction model,using the fused value as the basis for fertilizer decisionmaking.Second,to address the difficulty of determining hyperparameters for the LSTM neural network,an improved particle swarm optimization algorithm is used to optimize the LSTM neural network,resulting in the IPSO-LSTM fertilizer prediction model.This model is used to simulate and analyze the nitrogen,phosphorus,and potassium fertilizer amounts for multiple corn plots.Then,the main controller,relay,and acquisition end of the fertilizer machine are designed and implemented,with the acquisition end responsible for soil parameter acquisition and first-level data fusion,the relay for data transmission and second-level data fusion,and the main controller using the i.MX6 ULL chip as the processing core for fertilizer application.Finally,the autonomous fertilization function of the fertilizer machine can effectively reduce fertilizer application and reduce waste.In conclusion,the fertilization machine based on i.MX6 ULL designed in this article can solve the problems of soil parameters being unobservable,unstable data transmission,unreasonable fertilization,high labor costs,and the reliance on human experience to determine fertilization amounts in traditional fertilization methods,and has practical significance.