Algal growth rate modeling and prediction optimization using incorporation of MLP and CPSO algorithm

The cause of global warming is the existence of greenhouse gases that trap the emitted infrared wave and cause the increasing of the earth´s temperature. One of the predominant greenhouse gases in the atmosphere is CO₂. Biosequestration by utilizing micro algae is one of the promising method to reduce the concentration of CO₂ in the atmosphere. This research focuses on the modeling of the algal growth which is one of the parameter that defines the amount of CO₂ which can be fixated by algal. From the observation data, the growth behavior is modeled by regression method, Multilayer Perceptron (MLP) algorithm. To optimize the algorithm, MLP is also combined with The Canonical Particle Swarm Optimization (CPSO). The result shows that modeling using MLP-CPSO is more accurate than the original MLP and MLP-PSO respectively by 25% and 15% in RAE. MLP-CPSO also shows the best performance in RMSE with 0.091 and coefficient correlation (r) with 0.92.