Detecting Anomalous Air Flow-ammonia Load Ratios, Using Gaussian Process Regression

In this paper we propose a method to detect abnormal air flow-ammonia load ratios in active sludge basins.

Summary

In this paper we propose a method to detect abnormal air flow-ammonia load ratios in active sludge basins. The purpose is to detect faulty sensors and process disturbances affecting the air flow-ammonia load ratio.

The method is based on Gaussian process regression, and is evaluated on plant data. Results indicate that drift in an ammonia on-line sensor, over flow during storm events and changed sludge properties can be detected by the proposed method.

Coworkers: Oscar Samuelsson, Anders Björk

Year: 2017.0

Report number: A2280

Authors: Oscar Samuelsson, Anders Björk, Jesús Zambrano, Tatiana Chistiakova ,Bengt Carlsson

Published in: 2015, 9th IWA Symposium on Systems Analysis and Integrated Assessment (WATERMATEX), At Gold Coast, Queensland, Australia