RNN-Autoencoder Approach for Anomaly Detection in Power Plant Predictive Maintenance Systems

Abstract

Induced Draft Fans (IDF) and Primary Air Fans (PAF) are critical equipment in steam power plants. Anomaly detection based on machine learning models is an approach that is currently being developed for optimization and increasing the effectiveness of predictive maintenance (PDM) as well as increasing the reliability of thermal power plants. The research aims to develop a data-driven model for diagnostic and prognostic equipment, produce accurate predictions using many sensor data taken in real-time from the SCADA system, and design a PDM management framework using an anomaly detection system. This research proposes to use a combined recurrent neural network (RNN)-autoencoder approach as a "normal" behavior model (NBM) with the Mahalanobis Distance (MD) statistical method for the detection of anomalies in power plant equipment. Based on time-series input sensor data, the RNNautoencoder is utilized to predict the behavior of the equipment in health circumstances. In contrast, the MD is used to determine the distance between the actual parameters of the equipment and its "normal" behavior prediction to determine the anomaly condition. This study examined the performance of the long short-term memory (LSTM) and gated recurrent unit (GRU) models in modeling normal behavior with hyperparameter optimization. The LSTM with the best hyperparameters had a validation loss of 5,690 x 10-4 and a validation accuracy of 93.36 percent, whereas the GRU with the best hyperparameters had a validation loss of 4.484 x 10-4 and a validation accuracy of 93.47 percent. GRU can outperform the LSTM model. The proposed framework can detect anomaly conditions in various cases of IDF and PAF equipment disturbances, both in early warning of equipment failure and during downtime conditions.

Advancements in Gas Turbine Fault Detection: A Machine Learning Approach Based on the Temporal Convolutional Network–Autoencoder Model

“…RNNs (i.e., LSTM and GRU) [30] Suitable for sequential data. They can capture dependencies over time.…”

“…Hybrid models [30,31] They combine features of multiple models. Versatility in feature extraction and modeling.…”

Enhancing Predictive Maintenance in Water Treatment Plants through Sparse Autoencoder Based Anomaly Detection

“…The [9][10] have developed a Dynamic Predictive Maintenance Scheduling (DPMS) method that uses deep auto-encoders and deep forest for failure prognosis, which showcases the method's effectiveness in maintenance and decision-making based on the system degradation FE from raw sensor data. This method is effectively validated using NASA's aircraft engine datasets and has possibly outperformed several other state-of-the-art methods, which underscored the potential of Deep Learning (DL) in predictive maintenance for reducing costs and facilitating precise maintenance decisions.…”