Design and implementation of a water quality monitoring device for shrimp ponds and prediction of dissolved oxygen using neural networks

Abstract

newlineIn the first approach a predictive model based on Fuzzy C means -Radial Basis Function Neural Network [F-RBFNN] is developed. Twenty four neurons were chosen as the number of hidden neurons by experimentation techniques, the file esteems are discovered to be bigger and least exact. Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Correlation Coefficient (R) and Willmott Index of Agreement (WIA) were used as the evaluation parameters. In the second approach, the Dolphin Glow Worm Optimization technique (DGO) is used to select the neuron numbers of Radial Basis Function Neural Network (RBFNN). The proposed technique DGO-RBFNN precisely predicts the DO in the shrimp pond as compared to the F-RBFNN with minimum number of neurons and MAE but however suffers from high computational time.