Title
Artificial intelligence-based approach for improving the aeration efficiency of a perforated plate aeration system

Authors
Subha M. Roy, C. M. Pareek, Taeho Kim, Bimal Chandra Mal

Journal
Multiscale and Multidisciplinary Modeling, Experiments and Design, Volume 7, Pages 869–882, 2024

Date Published
October 10, 2023

Abstract
Aeration of a water body is artificially performed to create a close interface between the air and the water, allowing for the enhancement of dissolved oxygen (DO) content. High levels of DO are essential for maintaining good water quality and supporting aquatic life. The perforated plate aerator (PPA) can be considered as an important device to increase DO concentration in a water body. Its performance was assessed in the laboratory using a tank of size 4 × 4 × 1.5 m3. The standard aeration efficiency (SAE) is a critical variable used to evaluate the effectiveness of an aerator. Various design and operating parameters of an aerator influence the SAE. There is a need to optimize design and operating parameters for maximizing the efficiency of aerator. In this study, an attempt was made to use the integrated artificial neural network (ANN) and particle swarm optimization (PSO) approach to serve the purpose. A 3-5-1 ANN model was developed for modeling the SAE of PPA, and the PSO technique was applied to find out the optimal values of design and operating parameters of PPA. The most suitable optimal design and operating parameters' values of the numbers of plate (N), hole diameter of plate (D), and water flow rate (Q) were found to be 3, 5.38 mm, and 0.0110 m3/s respectively. The applied ANN-PSO technique can forecast the optimal values satisfactorily with a maximum of 3.99% difference with the observed values. Therefore, the suitability of the applied technique was confirmed.

Keywords
Perforated plate aerator, dissolved oxygen, standard aeration efficiency, artificial neural network, particle swarm optimization

DOI
https://doi.org/10.1007/s41939-023-00250-0