Modeling and simulation of photocatalytic reactors using deep learning methods

Abstract

Deep learning-based artificial intelligence has shown promise in a variety of fields applications and research areas, compared to traditional machine learning algorithms, it can produce extremely high-level data representations making them promising for a wide range of applications. The primary aim of this study is to present a comprehensive methodology for designing deep learning models specifically tailored for addressing complex non-linear regression challenges. More specifically, the focus is on developing two advanced models: a Convolutional Neural Network (CNN) and a Long Short-Term Memory (LSTM) model. These models will be utilized to tackle practical issues, notably simulating a photocatalytic reactor to accurately predict water purification efficacy and estimate its lifespan. To ensure the utmost credibility and applicability, real-world datasets will be employed in this research endeavor. By harnessing the power of cutting-edge deep learning techniques, this study endeavors to advance the understanding and optimize the intricate processes involved in water purification. As part of this endeavor, two deep learning models were devised to tackle the assigned challenge. The training and validation stages of the deep learning model proved fruitful for the LSTM model, thus substantiating its efficacy in addressing non-linear regression problems. In contrast, the CNN model encountered obstacles during the validation phase, implying that the LSTM model is better suited to handle such complexities