Predictive Analysis of Soil Parameters for Solar-Powered Smart Irrigation System

Abstract

Agriculture is the backbone of any economy. More than 70% of households depend on farming to provide a living, making advancements in farming technologies are necessary. Irrigation is the most crucial and defining parameter for producing a healthy yield lack thereof can cause draught and low produce. On the flip side, over-irrigation causes deterioration of soil properties hence affecting the yield growth. An IoT-based irrigation system that integrates automation with the traditional irrigation system helps overcome these drawbacks. The paper primarily focuses on developing an intelligent system through Deep Learning Algorithms. Using Recurrent Neural Network (RNN), Gated Recurrent Unit (GRU), and Long Short Term Memory (LSTM) models of deep learning, the paper proposes a strategy to establish a system that can operate automatically through the forecasted data.