Prediction of failure rate by artificial neural networks

The performance of treatment processes is generally influenced by many factors such as qualitative and quantitative changes in the wastewater and the inherent variability of the treatment process. Algerian legislation has established quality criteria for treated and discharged water, so it should be possible to assess process performance and plant reliability to ensure compliance.Predicting the behavior of complex systems has been a vast field of application for artificial neural networks. Applications such as economic forecasting, power load/demand forecasting, and forecasting of natural and physical phenomena have been widely studied, hence the numerous papers presented at annual conferences in this field of work. The cognitive capacity of artificial neural networks for mapping complex input-output and non-linear relationships, which would allow for better process forecasting and control, make them particularly attractive.