Neural Network Pattern Classification and Weather Dependent Fuzzy Logic Model for Irrigation Control in WSN Based Precision Agriculture

AbstractWatering system in agricultural lands plays a major activity in water and soil conservations. The future expectation of soil moisture content (MC) utilizing online soil and ecological parameters may give an effective stage to agriculture land watering system prerequisites. This article focuses on two optimization strategies, for example, Scaled Conjugate Gradient and BFGS Quasi-Newton based neural network algorithms utilized to predict hourly requirement of soil MC. The prediction performance of these two optimization techniques are also studied by calculating MSE (Mean Square Error), RMSE (Root Mean Square Error), and R-squared error. The calculations are tried for the forecast of soil MC in every one hour advance by considering eleven distinctive soil and environmental parameters. The best technique is used for the final prediction, and the predicted soil MC is utilized for generating appropriate notifications using fuzzy logic based weather model. The proposed system is hybrid system utilized to solve a single problem that is the generation of best irrigation suggestions for the farmers