A fault early warning method for auxiliary equipment based on multivariate state estimation technique and sliding window similarity

The induced draft (ID) fan is an important piece of auxiliary equipment in coal-fired power plants. Early fault detection of the ID fan can provide predictive maintenance and reduce unscheduled shutdowns, thus improving the reliability of the power generation. In this study, an adaptive model was developed to achieve the early fault detection of ID fans. First, a non-parametric monitoring model was constructed to describe the normal operating characteristics with the multivariate state estimation technique (MSET). A similarity index representing operation status was defined according to the prediction deviations to produce warnings of early faults. To deal with the model accuracy degradation because of variant condition operation of the ID fan, an adaptive strategy was proposed by using the samples with a high data quality index (DQI) to manage the memory matrix and update the MSET model, thereby improving the fault detection results. The proposed method was applied to a 300 MW coal-fired power plant to achieve the early fault detection of an ID fan. In addition, fault detection by using the model without an update was also compared. Results show that the update strategy can greatly improve the MSET model accuracy when predicting normal operations of the ID fan; accordingly, the fault can be detected more than 4 h earlier by using the strategy with the adaptive update when compared to the model without an update.

Download Full-text

An Adaptive Early Fault Detection Model of Induced Draft Fans Based on Multivariate State Estimation Technique

10.20944/preprints202104.0499.v1 ◽

2021 ◽

Author(s):

Ruijun Guo ◽

Guobin Zhang ◽

Qian Zhang ◽

Lei Zhou ◽

Haicun Yu ◽

...

Keyword(s):

Power Plant ◽

Fault Detection ◽

State Estimation ◽

Thermal Power ◽

Adaptive Strategy ◽

Normal Operation ◽

Efficient Production ◽

Estimation Technique ◽

Auxiliary Equipment ◽

Early Fault Detection

The induced draft (ID) fan is important auxiliary equipment in the thermal power plant. It is of great significance to monitor the operation of the ID fan for safe and efficient production. In this paper, an adaptive warning model is proposed to detect early faults of ID fans. First, a non-parametric monitoring model is constructed to describe the normal operation states with the multivariate state estimation technique (MSET). Then, an early warning approach is presented to identify abnormal behaviors based on the results of the MSET model. As the performance of the MSET model is heavily influenced by the normal operation data in the historic memory matrix, an adaptive strategy is proposed by using the samples with a high data quality index (DQI) to manage the memory matrix and update the model. The proposed method is applied to a 300 MW coal-fired power plant for early fault detection, and it is compared with the model without an update. Results show that the proposed method can detect the fault earlier and more accurately.

Download Full-text

Development of A Pharmaceutical Early Warning Scoring Model for Cardiac Arrest Using Deep Learning: Retrospective Cohort Study (Preprint)

10.2196/preprints.21720 ◽

2020 ◽

Author(s):

Hsiao-Ko Chang ◽

Hui-Chih Wang ◽

Chih-Fen Huang ◽

Feipei Lai

Keyword(s):

Time Series ◽

Cardiac Arrest ◽

Early Warning ◽

Time Series Data ◽

Predictive Accuracy ◽

Vital Signs ◽

Sliding Window ◽

Early Warning Score ◽

Series Data ◽

Dynamic Time

BACKGROUND In most of Taiwan’s medical institutions, congestion is a serious problem for emergency departments. Due to a lack of beds, patients spend more time in emergency retention zones, which make it difficult to detect cardiac arrest (CA). OBJECTIVE We seek to develop a pharmaceutical early warning model to predict cardiac arrest in emergency departments via drug classification and medical expert suggestion. METHODS We propose a new early warning score model for detecting cardiac arrest via pharmaceutical classification and by using a sliding window; we apply learning-based algorithms to time-series data for a Pharmaceutical Early Warning Scoring Model (PEWSM). By treating pharmaceutical features as a dynamic time-series factor for cardiopulmonary resuscitation (CPR) patients, we increase sensitivity, reduce false alarm rates and mortality, and increase the model’s accuracy. To evaluate the proposed model we use the area under the receiver operating characteristic curve (AUROC). RESULTS Four important findings are as follows: (1) We identify the most important drug predictors: bits, and replenishers and regulators of water and electrolytes. The best AUROC of bits is 85%; that of replenishers and regulators of water and electrolytes is 86%. These two features are the most influential of the drug features in the task. (2) We verify feature selection, in which accounting for drugs improve the accuracy: In Task 1, the best AUROC of vital signs is 77%, and that of all features is 86%. In Task 2, the best AUROC of all features is 85%, which demonstrates that thus accounting for the drugs significantly affects prediction. (3) We use a better model: For traditional machine learning, this study adds a new AI technology: the long short-term memory (LSTM) model with the best time-series accuracy, comparable to the traditional random forest (RF) model; the two AUROC measures are 85%. (4) We determine whether the event can be predicted beforehand: The best classifier is still an RF model, in which the observational starting time is 4 hours before the CPR event. Although the accuracy is impaired, the predictive accuracy still reaches 70%. Therefore, we believe that CPR events can be predicted four hours before the event. CONCLUSIONS This paper uses a sliding window to account for dynamic time-series data consisting of the patient’s vital signs and drug injections. In a comparison with NEWS, we improve predictive accuracy via feature selection, which includes drugs as features. In addition, LSTM yields better performance with time-series data. The proposed PEWSM, which offers 4-hour predictions, is better than the National Early Warning Score (NEWS) in the literature. This also confirms that the doctor’s heuristic rules are consistent with the results found by machine learning algorithms.

Download Full-text

A Functional Observer Based Dynamic State Estimation Technique for Grid Connected Solid Oxide Fuel Cells

IEEE Transactions on Energy Conversion ◽

10.1109/tec.2017.2739153 ◽

2018 ◽

Vol 33 (1) ◽

pp. 96-105 ◽

Cited By ~ 9

Author(s):

Kianoush Emami ◽

Hadi Ariakia ◽

Tyrone Fernando

Keyword(s):

Fuel Cells ◽

Solid Oxide Fuel Cells ◽

State Estimation ◽

Solid Oxide ◽

Dynamic State ◽

Oxide Fuel ◽

Estimation Technique ◽

Functional Observer ◽

Dynamic State Estimation

Download Full-text

Multistage identification method for real-time abnormal events of streaming data

International Journal of Distributed Sensor Networks ◽

10.1177/1550147719894544 ◽

2019 ◽

Vol 15 (12) ◽

pp. 155014771989454

Author(s):

Hao Luo ◽

Kexin Sun ◽

Junlu Wang ◽

Chengfeng Liu ◽

Linlin Ding ◽

...

Keyword(s):

Real Time ◽

Coal Mine ◽

Early Warning ◽

Sliding Window ◽

Geological Structure ◽

Streaming Data ◽

Event Monitoring ◽

Monitoring Method ◽

Identification Method ◽

Abnormal Events

With the development of streaming data processing technology, real-time event monitoring and querying has become a hot issue in this field. In this article, an investigation based on coal mine disaster events is carried out, and a new anti-aliasing model for abnormal events is proposed, as well as a multistage identification method. Coal mine micro-seismic signal is of great importance in the investigation of vibration characteristic, attenuation law, and disaster assessment of coal mine disasters. However, as affected by factors like geological structure and energy losses, the micro-seismic signals of the same kind of disasters may produce data drift in the time domain transmission, such as weak or enhanced signals, which affects the accuracy of the identification of abnormal events (“the coal mine disaster events”). The current mine disaster event monitoring method is a lagged identification, which is based on monitoring a series of sensors with a 10-s-long data waveform as the monitoring unit. The identification method proposed in this article first takes advantages of the dynamic time warping algorithm, which is widely applied in the field of audio recognition, to build an anti-aliasing model and identifies whether the perceived data are disaster signal based on the similarity fitting between them and the template waveform of historical disaster data, and second, since the real-time monitoring data are continuous streaming data, it is necessary to identify the start point of the disaster waveform before the identification of the disaster signal. Therefore, this article proposes a strategy based on a variable sliding window to align two waveforms, locating the start point of perceptual disaster wave and template wave by gradually sliding the perceptual window, which can guarantee the accuracy of the matching. Finally, this article proposes a multistage identification mechanism based on the sliding window matching strategy and the characteristics of the waveforms of coal mine disasters, adjusting the early warning level according to the identification extent of the disaster signal, which increases the early warning level gradually with the successful result of the matching of 1/ N size of the template, and the piecewise aggregate approximation method is used to optimize the calculation process. Experimental results show that the method proposed in this article is more accurate and be used in real time.

Download Full-text