Recently, climate change is a key issue that the world must face. Renewable energy is considered a potential technology to stop global warming and meet the energy demand. However, renewable sources are unstable and therefore need a conversion method to store electricity generated. Solar liquid fuel is a possible method to store renewable energy in liquid form. This research project concentrates on solar to liquid fuel. The research demonstrates the technology that converts solar energy into formic acid and stores it the ambient conditions. The simulations will compute the electricity generated by the PV system daily and monthly and the mass of formic acid that could be produced by using solar energy. The two-axis solar tracking system is simulated for the PV panels. Two important parameters of the two-axis solar tracker are investigated: the elevation angle and the Azimuth angle. The machine learning algorithms like ANN and SVM are deployed to model the PV power and formic acid production by the causal variables. SVM has outperformed the ANN in terms of better prediction capacity for the two objectives, i.e., PV power and Formic Acid production. The model can be deployed to predict the two objectives under the impact of the causal variables.

Recently, climate change is a key issue that the world must face. Renewable energy is considered a potential technology to stop global warming and meet the energy demand. However, renewable sources are unstable and therefore need a conversion method to store electricity generated. Solar liquid fuel is a possible method to store renewable energy in liquid form. This research project concentrates on solar to liquid fuel. The research demonstrates the technology that converts solar energy into formic acid and stores it the ambient conditions. The simulations will compute the electricity generated by the PV system daily and monthly and the mass of formic acid that could be produced by using solar energy. The two-axis solar tracking system is simulated for the PV panels. Two important parameters of the two-axis solar tracker are investigated: the elevation angle and the Azimuth angle. The machine learning algorithms like ANN and SVM are deployed to model the PV power and formic acid production by the causal variables. SVM has outperformed the ANN in terms of better prediction capacity for the two objectives, i.e., PV power and Formic Acid production. The model can be deployed to predict the two objectives under the impact of the causal variables.