An Artificial Neural Network Model of Nutrient Release from Polymer-Coated Fertilizers

In real world, mathematical models are used to explain complicated systems and help understanding the effects of different elements and phenomena. Controlled-release fertilizers (CRFs) are very useful and beneficial in the agricultural field over conventional fertilizers [1]. Polymer-coated fertilizers (PCFs) are the most popular type of controlled-release fertilizers in the market [2]. However, the complexity of the process of release of nutrients from PCFs makes it difficult to optimize the design of PCFs applications by mathematical models [1]. Also, the biodegradation of PCFs into the soil usually takes months, and a numerus sets of experiments have to be conducted in order to estimate the release pattern of PCFs [1, 3]. Therefore, the objective of this project is to develop a generalized regression neural network (GRNN) model to predict the release profiles of PCFs over time, and to study the parameters that affect the release pattern of nutrients from these fertilizers. Matlab simulation is used to show the release pattern of PCFs contents over time. In order to develop a GRNN model, the field of the study involves basic knowledge of artificial neural network theory and rules.The model can quickly predict the release profiles of PCFs and it can become very useful in optimizing the design of polymer-coated applications.