Detecting and Diagnosing Faults in Wastewater Sensors
Author Information
Author(s): Luca Alexandra-Veronica, Simon-Várhelyi Melinda, Mihály Norbert-Botond, Cristea Vasile-Mircea
Primary Institution: Department of Chemical Engineering, Faculty of Chemistry and Chemical Engineering, Babes-Bolyai University of Cluj-Napoca
Hypothesis
Can principal component analysis and Fisher discriminant analysis effectively detect and diagnose faults in wastewater nitrate and nitrite sensors?
Conclusion
The study demonstrates that combining PCA and FDA can effectively detect and identify faults in wastewater sensors, leading to improved operational efficiency and reduced environmental impact.
Supporting Evidence
- The PCA model achieved a detection accuracy of 98.2%.
- The FDA model demonstrated an overall diagnosis accuracy of 97.7%.
- Faults in sensors led to increased energy consumption by up to 10.5%.
- Environmental assessments showed that faulty sensors increased greenhouse gas emissions.
- The study evaluated five types of sensor faults and their impacts on wastewater treatment efficiency.
Takeaway
This study shows how scientists can find problems with sensors that measure pollution in water, helping to keep our water clean and save energy.
Methodology
The study used principal component analysis (PCA) for fault detection and Fisher discriminant analysis (FDA) for fault identification based on data from a municipal wastewater treatment plant.
Limitations
The methodologies may have limitations when applied to insufficient or imprecise data sets.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website