Hybrid decision support system using PLSR-fuzzy model for GSM-based site-specific irrigation notification and control in precision agriculture

Abstract

In this paper, a partial least square regression PLSR and fuzzy-logic based smart decision support system DSS for crop-specific irrigation notification and control in precision agriculture is proposed, and this can be implemented in farm land, green-house and poly-house. The proposed DSS model continuously acquires real-time soil and environmental data using Zigbee wireless sensor network WSN with WiFi gateway developed during this work. The collected data are used to predict hourly soil moisture content MC variation and required evapotranspiration to control farm irrigation by utilising fuzzy logic-based weather model. A comparative analysis of soil MC variations is also performed by calculating root mean square error RMSE, R-squared error RSE, mean squared error MSE, ratio of performance to deviation RPD and algorithm running time. Cropwise evapotranspiration is also calculated using Blaney-Criddle method which is attached to the proposed DSS model by taking weather, soil, water and crop data into considerations.