Optimizing pigment production from agricultural waste using metaheuristic-based algorithms

Abstract

Due to the uncontrolled industrial applications of synthetic pigments that can cause a serious hazard to human health and the environment, the scientific community skewed towards natural colors. The simplest and efficient method to increase pigment production is by manipulating the medium. Among classical and statistical methods, one factor at a time and response surface methodology (RSM) is the most widely used in medium optimization. However, the main drawback of these methods is tedious wet experiments need to be conducted to predict the output for a new input data and prior to data processing and analytic for decision making. In the past few years, the rapid advances in the field of metaheuristic optimization algorithm have provided a solution in optimization problems. In this study, metaheuristic optimization scheme, together with the mathematical model which is regression analysis have been implemented to minimize time and cost of wet-lab experiments by increasing the pigment productions using the proposed compact experiments. Moreover, the predictive optimization performance and sensitivity analysis of metaheuristic algorithm have been evaluated to validate the results, and the authenticity has been proven by wet laboratory experiments. Analysis of the optimization showed that the percentage improvement for the proposed compact experiment which is particle swarm optimization (PSO) model improved from RSM model by 1.32%, while the percentage improvement for all compact experiments was better than multiple polynomial model (MPR) model with the highest PSO percentage of 2.0507%. Hence, the experimental findings revealed that, the metaheuristic-based approach successfully predicted the optimum fermentation parameters condition and concentration with better achievement on pigment production by using proposed compact experiment.