ANN dual-fluid PV/T data and code in R programming - Architecture of the Artificial Neural Network

Published: 08-03-2021| Version 1 | DOI: 10.17632/gxxszgy85t.1
Contributors:
Hasila Jarimi,
Ali H.A. Al-Waeli,
Tajul Rosli Razak,
Mohd Nazari Abu Bakar,
Ahmad Fazlizan,
Adnan Ibrahim,
Kamaruzzaman Sopian

Description

This dataset provides the artificial neural network architecture for a dual-fluid photovoltaic thermal (PV/T) collector which was experimentally tested in the outdoor environment of Malaysia. The system was set up and tested in three modes, which are (i) air mode, (ii) water mode and (iii) simultaneous mode. For modes (i), (ii) and (iii) air flows through the cooling channels, water flows through the cooling channels and both air and water flow together. To create this dataset, the following steps were carried out: 1. Select input variables: 5 data inputs were selected, which are Ambient temperature, wind speed, solar irradiance, inlet air temperature and inlet water temperature. 2. Select Algorithm: for training, the Backpropagation neural network (BPNN) was used. 3. Select output variables: 6 data output were selected, which are PV surface temperature, PV temperature, temperature of the back plate, the temperature of the outlet air and outlet water, in addition to the electrical efficiency. Step 1: Import the data Step 2: Normalize the data Step 3: Split the dataset into training and testing data Step 4: Create the NN model in R studio. The package 'Neuralnet' in the R programming language was used. The coding in R studio is provided in the attached file.

Files

Steps to reproduce

1. Select input variables. 2. Select Algorithm. 3. Select output variables. Step 1: Import the data Step 2: Normalize the data Step 3: Split the dataset into training and testing data Step 4: Create the NN model in R studio. The coding in R studio is provided in the attached file.