Fault Detection and Diagnosis in Industrial Processes with Variational Autoencoder: A Comprehensive Study

Anomaly Detection Method for Harmonic Reducers with Only Healthy Data

A harmonic reducer is an important component of industrial robots. In practical applications, it is difficult to obtain enough anomaly data from error cases for the supervised training of models. Whether the information contained in regular features is sensitive to anomaly detection is unknown. In this paper, we propose an anomaly detection frame for a harmonic reducer with only healthy data. We considered an auto-encoder trained using only healthy features, such as feature mapping, in which the difference between the output and the input constitutes a new high-dimensional feature space that retained information relevant only to anomalies. Compared to the original feature space, this space was more sensitive to abnormal data. The mapped features were then fed into the OCSVM to preserve the feature details of the abnormal information. The effectiveness of this method was validated by multiple sets of data collecting from harmonic reducers. Three different residual calculations and four different AE models were used, showing that the method outperforms an AE or an OCSVM alone. It is also verified that the method outperforms other typical anomaly detection methods.

Fault diagnosis based on counterfactual inference for the batch fermentation process

Fault diagnosis plays a pivotal role in identifying the root causes of a fault. Current fault diagnosis methods encounter the shortcomings being unable to assess the fault amplitude or having low efficiency for batch fermentation process. In order to solve the above problems, this paper proposes a fault detection model named convolutional neural network based on variational autoencoder (CNN-VAE) and a fault diagnosis based on counterfactual inference (FDCI). To begin with, quality-related process variables are selected using mutual information (MI). Next, a two-dimensional moving window is used to obtain input sequences from the process data. Then, two statistics from the latent and residual domains of the CNN-VAE model are constructed for fault detection. Additionally, once a fault occurs, FDCI is used to locate the root cause of a fault. Finally, a simulation process and a real-world L. plantarum batch fermentation process are provided to demonstrate the effectiveness of the proposed approache.

Integration of Encoding and Temporal Forecasting: Toward End-to-End NOx Prediction for Industrial Chemical Process

Forecasting NOx concentration in fluid catalytic cracking (FCC) regeneration flue gas can guide the real-time adjustment of treatment devices, and then furtherly prevent the excessive emission of pollutants. The process monitoring variables, which are usually high-dimensional time series, can provide valuable information for prediction. Although process features and cross-series correlations can be captured through feature extraction techniques, they are commonly linear transformation, and conducted or trained separately from forecasting model. This process is inefficient and might not be an optimal solution for the following forecasting modeling. Therefore, we propose a time series encoding temporal convolutional network (TSE-TCN). By parameterizing the hidden representation of the encoding-decoding structure with the temporal convolutional network (TCN), and combining the reconstruction error and the prediction error in the objective function, the encoding-decoding procedure and the temporal predicting procedure can be trained by a single optimizer. The effectiveness of the proposed method is verified through an industrial reaction and regeneration process of an FCC unit. Results demonstrate that TSE-TCN outperforms some state-of-art methods with lower root mean square error (RMSE) by 2.74% and higher R2 score by 3.77%.

Ultra-Lightweight Fast Anomaly Detectors for Industrial Applications

Quality inspection in the pharmaceutical and food industry is crucial to ensure that products are safe for the customers. Among the properties that are controlled in the production process are chemical composition, the content of the active substances, and visual appearance. Although the latter may not influence the product's properties, it lowers customers' confidence in drugs or food and affects brand perception. The visual appearance of the consumer goods is typically inspected during the packaging process using machine vision quality inspection systems. In line with the current trends, the processing of the images is often supported with deep neural networks, which increases the accuracy of detection and classification of faults. Solutions based on AI are best suited to production lines with a limited number of formats or highly repeatable production. In the case where formats differ significantly from each other and are often being changed, a quality inspection system has to enable fast training. In this paper, we present a fast method for image anomaly detection that is used in high-speed production lines. The proposed method meets these requirements: It is easy and fast to train, even on devices with limited computing power. The inference time for each production sample is sufficient for real-time scenarios. Additionally, the ultra-lightweight algorithm can be easily adapted to different products and different market segments. In this work, we present the results of our algorithm on three different real production data gathered from food and pharmaceutical industries.