Increasing Biodiesel Production Using Machine Learning Algorithms

Microalgae are an excellent source of biomass for next generation biodiesel production. In this study, both theoretical and experimental tests of variables such as lipase, methanol, algal oil concentration, incubation time and temperature used in biodiesel production were performed using RSM, ANN and IRS modeling. According to results, the highest experimental and theoretical biodiesel efficiencies were 87%, 87.5% at 1 % lipase, 2 ml of methanol, 0.2 g of algal oil, 6 h of incubation time and 45 °C of temperature, respectively. Also, we modelled the variables via interactive response surface (IRS). Maximum biodiesel efficiencies of full quadratic and interaction models were 80.1504% and 75.4416% at 1.7917% of lipase, 1.1722 mL of methanol, 0.575 g of algal oil, 4.0833 h of incubation time and 40.2778 °C of temperature, respectively. In conclusion, Biodiesel yield can be increased by changing parameters such as lipase, methanol, algal oil, incubation time and temperature.