A steam turbine is one of the critical components in a power generation system whose failure may result in unexpected consequences, even catastrophic losses. Thus, the reliability of steam turbines needs to be guaranteed all the time, which requires that its health state can be monitored and predicted effectively. Due to various failure modes, it is difficult to build physics-of-failure models used for health prognostics for steam turbines. In this paper, a data-driven integrated framework for health prognostics for steam turbines, which is based on extreme gradient boosting (XGBoost) and dynamic time warping (DTW), is proposed. The proposed framework includes two modules: anomaly detection and remaining useful life (RUL) prediction. The anomalies refer to the overall abnormal operation of steam turbines. In the process of anomaly detection, the temporal variables which can represent the operating conditions of the considered steam turbine are selected first. Appropriately selected temporal variables can reduce the input dimension and will improve real-time performance. Then, XGBoost is used to detect anomalies based on learning historical data. In the process of RUL prediction, a similarity-based algorithm with DTW is used to gain the RUL by contrasting the measured temporal variables with those in the historical cases. The similarity-based algorithm can predict the RUL without establishing a degradation path model, which can avoid the difficulties in parameter estimation for the degradation model and model generalization. The proposed framework is validated by real case studies from an industrial steam turbine. The results show that the proposed approach can detect the anomalies successfully and predict the RUL effectively.INDEX TERMS Dynamic time warping, extreme gradient boosting, remaining useful life prediction, steam turbine.
The coal pulverizing system is an important auxiliary system in thermal power generation systems. The working condition of a coal pulverizing system may directly affect the safety and economy of power generation. Prognostics and health management is an effective approach to ensure the reliability of coal pulverizing systems. As the coal pulverizing system is a typical dynamic and nonlinear high-dimensional system, it is difficult to construct accurate mathematical models used for anomaly detection. In this paper, a novel data-driven integrated framework for anomaly detection of the coal pulverizing system is proposed. A neural network model based on gated recurrent unit (GRU) networks, a type of recurrent neural network (RNN), is constructed to describe the temporal characteristics of high-dimensional data and predict the system condition value. Then, aiming at the prediction error, a novel unsupervised clustering algorithm for anomaly detection is proposed. The proposed framework is validated by a real case study from an industrial coal pulverizing system. The results show that the proposed framework can detect the anomaly successfully.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.