Genetic algorithm optimization of supercritical fluid extraction of nimbin from neem seeds

• Published in: Journal of food engineering Vol. 97; no. 2; pp. 127 - 134
• Main Authors: Zahedi, G., Elkamel, A., Lohi, A.
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.03.2010; [New York, NY]: Elsevier Science Pub. Co; Elsevier
• Subjects: algorithms; Azadirachta indica; Biological and medical sciences; carbon dioxide; chemical constituents of plants; diffusion; diffusivity; dynamic programming; flow resistance; Food engineering; Food industries; Fundamental and applied biological sciences. Psychology; General aspects; Genetic algorithm; mass transfer; medicinal plants; Modeling; Neem extraction; neem oil; neem seed extract; Optimization; quantitative analysis; seeds; Supercritical extraction; supercritical fluid extraction; Supercritical extraction; Neem extraction; Genetic algorithm; Modeling; Optimization; Seeds; Supercritical fluid extraction
• ISSN: 0260-8774; 1873-5770
• DOI: 10.1016/j.jfoodeng.2009.10.001

The subject of this study is to optimize supercritical extraction of nimbin from neem seeds using a Genetic Algorithm (GA) technique. In order to investigate the effect of parameters on nimbin extraction yield, a partial differential equation model based on mass conservation was developed and solved numerically. The results were successfully validated and a parameter estimation problem that employs laboratory experimental data was solved. Using this validated model and the optimized set of parameters in the model, another problem was formulated with the aim of optimizing the extraction process. Profit was set as the objective function. Using a GA optimization algorithm, it was found that profit achieves its maximum when T = 305 k, P = 200 bar, carbon dioxide flow rate = 0.967 cm 3/min and d p = 0.1431 cm. The ability of the GA algorithm in optimizing the process was compared with a traditional Gradient Search (GS) optimization technique. THE GA technique proved to be a more efficient technique; especially when considering computational effort in reaching an optimal solution.