A Data-Driven Health Prognostics Approach for Steam Turbines Based on Xgboost and DTW

Abstract

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.

Detecting Coal Pulverizing System Anomaly Using a Gated Recurrent Unit and Clustering

“…Therefore, it is difficult to build mathematical models to describe the operation condition and detect anomalies for the coal pulverizing system. Process monitoring data, which contains the system condition information, can be used to assess the operating condition and detect whether there is an anomaly or not [15]. In order to detect the operating condition of the coal pulverizing system more accurately, it is necessary to obtain enough information from multiple process measure data, such as temperature, current, speed, etc.…”

“…Choi employed a long short-term memory (LSTM) network to detect sensor faults [14]. Que proposed a XGBoost-based framework to detect anomalies in steam turbines [15]. These machine learning approaches mostly belong to supervised learning, which is sensitive to the amount and the quality of training data.…”

Anomaly detection of steam turbine with hierarchical pre‐warning strategy

“…The existence of redundant features may reduce the speed and accuracy of the learning algorithm [7]. This problem has been solved in the work of [8]. Que applies XGBoost model to detect anomalies based on learning history data, but this method cannot rank the importance of features and has poor interpretability.…”

Hard-Division and Multi-Model Based Floating Height Prediction for Air Cushion Furnace With Hybrid Nozzles

“…Hou used the least square support vector regression (LSSVR) to predict the maximum and minimum floating height of the strip in the air cushion furnace with double-slot nozzle and obtained good prediction results [14]. Due to the advantages of efficiency, flexibility and scalability, eXtreme Gradient Boosting (XGBoost) is often used in the prediction of vibration signals in the other fields of industry [17]- [20].…”