An Evaluation of ANFIS Models for Predicting the Oil Flotation Behavior in a Stable Oil-Water Emulsion

Autor: Ku Esyra Hani Ku Ishak, Mohammed Abdalla Ayoub, Muhammad Irman Khalif Ahmad Aminuddin
Rok vydání: 2023
Zdroj: Advances in Science and Technology.
ISSN: 1662-0356
DOI: 10.4028/p-071yk6
Popis: This work's main objective is to investigate the flotation process's efficiency in removing the oil from a stable oil-water emulsion containing surfactant (MFOMAX) and polymer (GLP-100). A total of 45 flotation tests were carried out with varying factors such as gas bubble flowrate, MFOMAX and GLP-100 concentration, as well as the flotation time. These factors have become the model's input, and the effluent's oil concentration calculated in term of flotation efficiency (%) has been the model’s output. 75% of the total data was used for training and 25% was used for testing. Coefficient of determination (R2) and average absolute percentage error (AAPE) was used as the models' performance indicators. A high R2 value (0.956) was given by the ANFIS model and AAPE of 10.14%, indicating that the predicted data agrees well with the actual data. Potential optimization to the flotation equipment on separating oil-water in the stable emulsion containing MFOMAX and GLP-100 have been discussed.
Databáze: OpenAIRE