Prediction of nitrate and chlorine in soil using ion selective electrode

The prediction of Soil nitrate and chlorine concentration is an important to in soil testing for formulated fertilization. The determination is performed employing the nitrate and chlorine selective electrode. In this paper, the Cl^- and NO₃^- content was predicted using an artificial neural network based on the Bayesian regularization, applied to the determination of nitrate in the presence of chloride interference. Through comparison and analysis, the verification error of BP neural network model which overcame the localization of multi- variable linear regression model, and it was on that the neural network has a better nonlinear reflection ability, and Call describe the complex relationship between the independent variable and the dependent variable with better precision, and has well feasibility The artificial neural network employed was a connected back propagation model with a 2-2-2 structure. The lower detection limits of nitrate and chlorine ion selective electrode are 7×10^-6 mol/L and 5×10^-5. Results obtained with this approach are compared with the direct determination of nitrate using its ion-selective electrode, it showed that the Artificial Neural Network provided more satisfactory prediction.