Premium
Performance evaluation of fault detection methods for wastewater treatment processes
Author(s) -
Corominas Lluís,
Villez Kris,
Aguado Daniel,
Rieger Leiv,
Rosén Christian,
Vanrolleghem Peter A.
Publication year - 2011
Publication title -
biotechnology and bioengineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.136
H-Index - 189
eISSN - 1097-0290
pISSN - 0006-3592
DOI - 10.1002/bit.22953
Subject(s) - fault detection and isolation , ewma chart , benchmark (surveying) , engineering , process (computing) , benchmarking , actuator , real time computing , control chart , computer science , artificial intelligence , business , geodesy , marketing , geography , operating system
Several methods to detect faults have been developed in various fields, mainly in chemical and process engineering. However, minimal practical guidelines exist for their selection and application. This work presents an index that allows for evaluating monitoring and diagnosis performance of fault detection methods, which takes into account several characteristics, such as false alarms, false acceptance, and undesirable switching from correct detection to non‐detection during a fault event. The usefulness of the index to process engineering is demonstrated first by application to a simple example. Then, it is used to compare five univariate fault detection methods (Shewhart, EWMA, and residuals of EWMA) applied to the simulated results of the Benchmark Simulation Model No. 1 long‐term (BSM1_LT). The BSM1_LT, provided by the IWA Task Group on Benchmarking of Control Strategies, is a simulation platform that allows for creating sensor and actuator faults and process disturbances in a wastewater treatment plant. The results from the method comparison using BSM1_LT show better performance to detect a sensor measurement shift for adaptive methods (residuals of EWMA) and when monitoring the actuator signals in a control loop (e.g., airflow). Overall, the proposed index is able to screen fault detection methods. Biotechnol. Bioeng. 2011;108: 333–344. © 2010 Wiley Periodicals, Inc.