Department of Biotechnolgy, Chemical Engineering - Environment, Faculty of Engineering, University of Technology and Education, The University of Danang, Vietnam
Title: Assessment of the current state of water environment in Phu Loc river, Da Nang using artificial intelligence – Important factor: plankton as bioindicators
Name(s): Truc-Xuyen Nguyen Phan1; Thi Dong Phuong Nguyen2*
Plankton are important components of aquatic systems because the diversity and stability of their community structure reflect changes in many environmental factors. Specifically, in river and lake ecosystems, plankton are primary producers and changes in their community composition affect the structure, function and stability of the ecosystem. Eutrophication causes excessive growth of phytoplankton, which in turn affects changes in the quantity and quality of zooplankton predation, leading to lake ecological instability and reduced biodiversity. Existing studies provide guidance on environmental factors associated with cyanobacterial blooms, using correlations between a limited number of environmental factors or a single type of environmental factor with chlorophyll a (chl-a) concentrations, cyanobacterial density or cyanobacterial blooms, which may lead to oversimplified conclusions. These research groups have used laboratory methods to study the environmental factors involved in cyanobacterial blooms. Given the impact of nutrient conditions on microalgae growth and environmental changes on phytoplankton populations, modeling their survival in these environments is important for predicting natural fluctuations. Therefore, mathematical models are capable of predicting the growth of these species under changing physicochemical conditions. Traditionally, artificial neural networks (ANNs) are a set of algorithms that are part of machine learning (ML) and respond by making predictions based on past data. Based on the limitation discussed above and the need to explore advanced ML tools to capture the behavior of plankton as indicators that respond to highly variable environmental factors, we modeled the growth of Chlorella sp. in an environment with variable principal chemical compounds (PCC) including C, N, and P based on real data of measured environmental parameters of Phu Loc River in Da Nang city, Vietnam, which collects large amounts of domestic wastewater from households in a large residential area before discharging into the Cu De River (16°07’17.0”N 107°59’01.0”E) for wastewater treatment plants. This was established to collect baseline data on the relationship between nutritional parameters of water quality and the presence of microalgae using machine learning algorithms. To achieve this, this study aims to accomplish the following objectives: 1. Identify PCC variables and their concentration gradients; 2. Field sampling of water and evaluate the water quality parameters including plankton diversity; 3. Assemble the data sets into RF, ANN, XGBoost models to determine the most accurate results for predicting the fluctuation of plankton under the most impact of chemical environmental conditions. Specifically, this approach allows us to predict which variable of PCC has the greatest influence on plankton. It is also a breakthrough to introduce a new hybrid model combining XGBoost with Fennec Fox Optimization (FFO) in predicting the early interaction of plankton and water quality, an efficiency of the hybrid model compared to traditional algorithms in ANN and FR.
In general, the remainder of the paper is structured as follows. Section 2 outlines the research methodology, including the dataset, experimental variables, and the key concepts behind FFO, XGBoost, and the proposed FFO-XGBoost hybrid model. Section 3 presents the experimental setup, performance comparisons, and discussion, along with a sensitivity analysis-based assessment of feature importance. Finally, Section 4 concludes the study with key findings and potential directions for future research.
Biography of the presenting author
Dr. Thi Dong Phuong Nguyen is a Senior lecturer at Department of Biotechnolgy, Chemical Engineering – Environment, Faculty of Engineering, The University of Danang, University of Technology and Education. She successful obtained the PhD in three years’ time after he graduated her Engineer degree from University of Science and Technology Danang. She currently is a Professional member of Institution of French Association of Engineering process (Société française de génie des procédés) during PhD student time in France. She was memeber in a grand project of Biodiesel production by microalgae – DIESALG project [ANR-12- BIME-000] on biodiesel fuel production from microalgae (https://algosolis.com/algosolis-a-la-une/ )during PhD student time and collaborated with experts in the fields of large-scale microalgae cultivation as well as buildings intergrated with microalgae photobioreactors. She is also a member of Asian Federation of Biotechnology (AFOB). She is a dynamic researcher in biotechnology with more than 15 SIC/SCIE articles.
Presenting author details
Full name: Thi Dong Phuong Nguyen
Contact number: 84 935 878 117
Email: ntdphuonng@ute.udn.vn
Session name/ number: 5
Category: (Oral presentation/ Poster presentation)