Model Optimization Using Artificial Intelligence Algorithms for Biological Food Waste Degradation
- 138 Downloads
Food waste is categorized as the largest degradable component in the waste stream. Degradation of food waste that involved aerobic bacteria is the most suitable approach to dispose of this waste. The main objective of this research is to evaluate the optimum condition of aerobic bacteria growth for food waste degradation by comparing the implementation of response surface method (RSM) and genetic algorithm. Preliminary experiment is conducted to determine the best time for aerobic bacteria growth. Then, evaluation of five factors such as temperature, time, type of nutrient, agitation rate and inoculum size is done by conducting experiments according to the experimental table that is constructed by using design expert software. Growth of aerobic bacteria can be determined by measuring the optical density (OD) of the bacteria. Aerobic bacteria at the best growth condition are mixed with the food waste for degradation process. The ability of aerobic bacteria to degrade food waste is determined by monitoring the pH, moisture content and ratio of volatile solid to total solid (VS/TS) of food waste on the first and twentieth days of degradation. The result analysis using RSM showed that the optimum condition for aerobic bacteria growth is at 37 °C and 200 rpm in commercial nutritional supplement (CNS) medium with 10% (v/v) of inoculum size for 20 h. At this optimum condition, the OD value was 2.264 while optimization using genetic algorithm generated the OD value at 2.643 where this is 14% improvement from the RSM.
KeywordsGenetic algorithm Response surface method Optimization Food waste degradation
The authors wish to acknowledge the Universiti Malaysia Pahang for funding the project under grant RDU1803119 and RDU1703295.
- Dhanarajan, G., Mandal, M., & Sen, R. (2014). A combined artificial neural network modeling–particle swarm optimization strategy for improved production of marine bacterial lipopeptide from food waste. Biochemical Engineering Journal, 84, 59–65. https://doi.org/10.1016/J.BEJ.2014.01.002.CrossRefGoogle Scholar
- Fakharudin, A. S., Sulaiman, M. N., Salihon, J., & Zainol, N. (2013). Implementing artificial neural networks and genetic algorithms to solve modeling and optimization of biogas production. In Proceedings of the 4th International Conference on Computing and Informatics, ICOCI 2013 (pp. 121–126), Sarawak, Malaysia. Universiti Utara Malaysia, August 28–30, 2013.Google Scholar
- Gill, S. S., Jana, A., & Shrivastav, A. (2014). Aerobic bacterial degradation of kitchen waste: A review. Journal of Microbiology, Biotechnology and Food Sciences, 3(6), 477–483.Google Scholar
- Heaton, J. (2018). Encog machine learning framework. Retrieved May 15, 2018, from https://github.com/encog/encog-java-core.
- Jacob, S., & Banerjee, R. (2016). Modeling and optimization of anaerobic codigestion of potato waste and aquatic weed by response surface methodology and artificial neural network coupled genetic algorithm. Bioresource Technology, 214, 386–395. https://doi.org/10.1016/J.BIORTECH.2016.04.068.CrossRefGoogle Scholar
- Komandai, N. (2017). Free Malaysia Today Corporation. Retrieved from Free Malaysia Today Web site: http://www.freemalaysiatoday.com/category/opinion/2017/08/29/food-wastage-management-crucial-for-a-better-environment/.
- Tortora, G., Funke, B., & Case, C. (2016). Microbiology: An Introduction (12th ed.). San Fransisco: Pearson Benjamin Cummings.Google Scholar
- Wilhelmstötter, F. (2018). Jenetics. Retrieved May 15, 2018, from http://jenetics.io/.