AUTHOR=Mng’ombe M. H. , Mtonga E. W. , Chunga B. A. , Chidya R. C. G. , Malota M. 

TITLE=Comparative study for the performance of pure artificial intelligence software sensor and self-organizing map assisted software sensor in predicting 5-day biochemical oxygen demand for Kauma Sewage Treatment Plant effluent in Malawi

JOURNAL=Frontiers in Environmental Engineering

VOLUME=Volume 3 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/environmental-engineering/articles/10.3389/fenve.2024.1373881

DOI=10.3389/fenve.2024.1373881

ISSN=2813-5067

ABSTRACT=Establishing influent and effluent wastewater characteristics is crucial for the control of internal processes within the treatment facility. The standard procedure for measuring biochemical oxygen demand (BOD5 is time-consuming, tough and expensive. Accordingly, relying on such procedures result in delays in the execution of crucial decisions and mitigation works, particularly in wastewater treatment plants where pollution is also inevitable. This study tested the reliability of the Adaptive Fuzzy Inference System (ANFIS), a kind of artificial intelligence algorithm that integrates both neural networks and fuzzy logic principles, to predict BOD5. The ANFIS models were developed and validated with historical wastewater quality data for Kauma Sewage Treatment Plant (KSTP), located in Lilongwe City, Malawi. A self-organizing map was applied to extract features of the raw data to enhance the performance of ANFIS. Cost-effective, quicker, and easier-to-measure variables were selected as possible predictors while using their respective correlations with effluent BOD5. Influents; temperature, pH, dissolved oxygen (DO), and Effluent chemical oxygen demand (COD) were the model predictors. The results demonstrated that the SOM-ANFIS model outperformed the ordinary ANFIS model in terms of modeling capabilities with the coefficient of determination (R)=0.96, coefficient of residual mass (CRM)≈0, and mean percent Error (MPE)≈0. A colour-coded graphic user interface (GUI) was developed to improve user interaction and friendliness of the developed model. Future improvement of the developed model will include integrating it with hardware components through the supervisory control and data acquisition (SCADA) system. The model developed in this study will enable timely intervention and cost-effective problem diagnosis of the wastewater treatment plant at KSTP and other similar facilities. The study recommends testing the ANFIS on data obtained from rivers and other wastewater treatment plants in Malawi to widen its application.