Soft sensors with white- and black-box approaches for a wastewater treatment process

Abstract

The increasing degradation of water resources makes it necessary to monitor and control process variables that may disturb the environment, but which may be very difficult to measure directly, either because there are no physical sensors available, or because these are too expensive. In this work, two soft sensors are proposed for monitoring concentrations of nitrate (NO) and ammonium (NH) ions, and of carbonaceous matter (CM) during nitrification of wastewater. One of them is based on reintegration of a process model to estimate NO and NH and on a feedforward neural network to estimate CM. The other estimator is based on Stacked Neural Networks (SNN), an approach that provides the predictor with robustness. After simulation, both soft sensors were implemented in an experimental unit using FIX MMI (Intellution, Inc) automation software as an interface between the process and MATLAB 5.1 (The Mathworks Inc.) software

Clustering-Based Hybrid Soft Sensor for an Industrial Polypropylene Process with Grade Changeover Operation

“…11,[13][14][15] ANNs are widely used for modeling of the nonlinear behaviors of a process because they allow a great deal of flexibility in deciding the model structures. 13,[16][17][18] NN-PLS algorithms 19,20 are hybrid models that combine linear PLS with ANNs. In NN-PLS, the ANNs capture the nonlinearity, while PLS allows the PLS projection to be maintained to obtain a robust generalization property.…”

“…Therefore, the performance of the proposed soft sensor in terms of its estimation accuracy was compared with that of the conventional black-box models, including PLS, ANNs, and neural network partial least squares (NN-PLS). PLS has been widely used as a powerful modeling method for constructing linear models from laboratory and field measurement data. ,− ANNs are widely used for modeling of the nonlinear behaviors of a process because they allow a great deal of flexibility in deciding the model structures. ,− NN-PLS algorithms , are hybrid models that combine linear PLS with ANNs. In NN-PLS, the ANNs capture the nonlinearity, while PLS allows the PLS projection to be maintained to obtain a robust generalization property.…”