Accurately predicting the short- and long-term variations of total nitrogen (TN) is vital for operating the wastewater treatment plants (WWTPs), considering the critical role TN plays in reflecting the eutrophication of wastewater. However, only a few relevant water quality parameters with limited samples can be obtained in WWTPs, which tremendously increases the difficulty in precisely predicting TN concentration. In this study, a multiphase attention-based recurrent neural network (MPA-RNN) is proposed. Benefited from its unique decomposition-summary attention structure, MPA-RNN first learns the temporal correlations and effectively excavates the useful information hidden in the historical data. Then, by designing a two-channel structure to transmit attention information, summary attention can integrate the decomposed information and learn the spatial relationships without information loss. Experimental results demonstrate that MPA-RNN achieves the best performance on both the SML2010 and practical TN datasets with the smallest root-mean-squared error, mean absolute error, and mean absolute percentage error when compared with the other state-of-the-art methods.

Modern industry processes are typically composed of multiple operating units with reaction interaction and energy–mass coupling, which result in a mixed time-varying and spatial–temporal coupling of process variables. It is challenging to develop a comprehensive and precise fault detection model for the multiple interconnected units by simple superposition of the individual unit models. In this study, the fault detection problem is formulated as a spatial–temporal fault detection problem utilizing process data of multiple interconnected unit processes. A spatial–temporal variational graph attention autoencoder (STVGATE) using interactive information is proposed for fault detection, which aims to effectively capture the spatial and temporal features of the interconnected unit processes. First, slow feature analysis (SFA) is implemented to extract temporal information that reveals the dynamic relevance of the process data. Then, an integration method of metric learning and prior knowledge is proposed to construct coupled spatial relationships based on temporal information. In addition, a variational graph attention autoencoder (VGATE) is suggested to extract temporal and spatial information for fault detection, which incorporates the dominances of variational inference and graph attention mechanisms. The proposed method can automatically extract and deeply mine spatial–temporal interactive feature information to boost detection performance. Finally, three industrial process experiments are performed to verify the feasibility and effectiveness of the proposed method. The results demonstrate that the proposed method dramatically increases the fault detection rate (FDR) and reduces the false alarm rate (FAR).

Multivariate regression models are commonly used in industrial process measurements. Locally weighted partial least squares (LWPLS) is a just-in-time learning (JITL) method for nonlinear multivariate regression models. However, it considers only the sample importance. In this article, a novel JITL multivariate regression model termed double LWPLS (D-LWPLS) is proposed to solve complex nonlinear problems and promote the model accuracy. Both the sample importance and variable importance are taken into consideration for the weighted model. Variable contribution is modified according to the variable importance. Sample similarity measurement is a significant basis for the model sample selection and sample weight calculation. The traditional sample similarity measurement methods take only the input information of samples as the index. In the D-LWPLS, a weighted similarity measurement is employed. The information on both input and output is considered by weighting the input variables according to the correlation between them. The effectiveness and the flexibility of the D-LWPLS model are demonstrated by the prediction results of a numerical example, a low-dimensional data set of an industrial process, and a high-dimensional data set of UV-Vis spectra of a polymetallic ion solution, which represent three kinds of nonlinear problems of classical, time-varying, and blind background fluctuation.