Feedback-driven error-corrected single-sensor analytics for real-time condition monitoring

Bhowmik, Basuraj; Panda, Satyam; Hazra, Budhaditya; Pakrashi, Vikram

doi:10.1016/j.ijmecsci.2021.106898

Cited by 18 publications

(13 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Equation ( 7) defines the moisture balance in the coal mill, and Equation ( 8) indicates the moisture evaporated by the pulverizing process. Equations ( 5), (6), and (9) respectively define the three output parameters of the model. Equation (10) presents the computation approach of individual fitness value.…”

Section: Coal Mill Dynamic Modelmentioning

confidence: 99%

“…Tian et al 3 utilized spectral kurtosis and cross correlation to compute the eigenvalues of different faults and combined them into a HI with principal component analysis and k-nearest neighbor. Bhowmik et al [4][5][6] investigated the use of first-order eigen perturbation (FOEP) techniques and recursive canonical correlation analysis for fault detection and pattern recognition of structures, and utilized the fundamental concept of FOEP to propose an online filtering approach based on recursive singular spectrum analysis, which provide the valuable reference for researching on equipment health monitoring and pre-processing of online data. Xu et al 7 constructed a rolling bearing HI based on the unsupervised deep belief network (DBN).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Health indicator construction and application of coal mill based on the dynamic model

Yin

Ding

et al. 2023

Structural Health Monitoring

View full text Add to dashboard Cite

As the vital auxiliary machine of the coal-fired power plant, monitoring the real-time operating status of coal mills is critical to the secure and stable operation of the power plant. In this study, a new method of construction of the coal mill health indicator (HI) is proposed, and the operation condition monitoring approaches of the device are designed based on the HI value. Firstly, an improved coal mill dynamic model considering the joint influence of drying force, ventilation force, and grinding force is established, and a synchronous optimization approach of model structures and parameters based on the genetic algorithm is designed. Then the deviation between the model output and the actual value is computed by the designed distance measuring approach, and the typical fault characteristic factors are designed based on the relation between the dynamic model and the actual operating state. And the HI value is calculated by fusing the deviation with the characteristic factors. Finally, the HI value is applied to the process of operation condition evaluation, fault diagnosis, and trend prediction, and has obtained favorable application effects. The results of this research show that the established HI value can reflect the operating status of the coal mill promptly and accurately, and the monitoring method designed based on the values have satisfactory practicality.

show abstract

Section: Coal Mill Dynamic Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Health indicator construction and application of coal mill based on the dynamic model

Yin

Ding

et al. 2023

Structural Health Monitoring

View full text Add to dashboard Cite

show abstract

“…Therefore, according to Equation ( 6), the loss function of the current value network and the target value network in deep reinforcement learning is shown in Equation (7).…”

Section: New Loss Functionmentioning

confidence: 99%

“…5 SSA 6 is another powerful signal-processing method, which has been well applied in rolling bearing fault detection in recent years. To improve the detection efficiency of SSA, Bhowmik et al 7 combining with eigen-perturbation (EP) theory, proposed an error-corrected First-order eigen perturbation (FOEP)-based formulation of Real-time singular spectrum analysis (RSSA) for adaptive filtering and verified the proposed model. At the same time, Bhowmik et al 8 studied the application of first-order EP techniques FOEP in structural health monitoring and verified it through experiments.…”

mentioning

confidence: 99%

Dual-input anomaly detection method based on deep reinforcement learning

Kang,

Chen,

Wang

et al. 2023

Structural Health Monitoring

View full text Add to dashboard Cite

Aiming at the problem of low accuracy of unsupervised learning anomaly detection algorithm, a dual-input anomaly detection method based on deep reinforcement learning was proposed. The proposed model mainly consists of a feature extractor and anomaly detector. Based on the deep reinforcement learning framework, the feature extractor uses a dual-input deep neural network to form the current value network and the target value network, which are used to extract the low-dimensional feature vectors. Based on the 3 σ principle, the reward function of reinforcement learning is designed to reward and punish the output results of the model during training. The model was trained only with the normal data, and the extracted feature vector of the normal class was used as the input of the anomaly detector to complete the learning of the detector. During the test, the input anomaly detection was realized based on the dual-input convolutional neural network, and the anomaly detector was completed by learning. To illustrate the generality and generalization performance of the proposed method, four sets of image data and two sets of rolling bearing fault data in different fields were verified respectively. At the same time, the proposed method is applied to the fault detection of a real aero-engine rolling bearing.The results show that the proposed model has high anomaly detection accuracy, which is superior to the current optimal method.

show abstract

“…Basuraj et al proposed a single-sensor online filtering method for recursive singular spectrum analysis based on the concept of first-order feature perturbation, which proved its effectiveness in several data sets. [ 6 ]. Ayati et al used KNN for single-sensor fault classification after extracting features using fast Fourier transform and wavelet packet transform [ 7 ].…”

Section: Introductionmentioning

confidence: 99%

Single-Sensor Engine Multi-Type Fault Detection

Tang

Cheng

et al. 2023

Sensors

View full text Add to dashboard Cite

Engine fault detection is conducive to improving equipment reliability and reducing maintenance costs. In practical scenarios, high-quality data is difficult to obtain. Usually, only single-sensor data is available. This paper proposes a fault detection method combining Variational Mode Decomposition (VMD) and Random Forest (RF). At first, the spectral energy distribution is obtained by decomposing and statistic the engine data of multiple working conditions. Based on the spectral energy distribution, the overall optimal mode number was identified, and the quadratic penalty term was optimized using SNR. The improved VMD (IVMD) improves mode aliasing and iterative efficiency and unifies feature dimensions. Decomposition of real signals demonstrates the effectiveness. The paper designs a feature vector composed of seven types of attributes, including unit bandwidth energy, center frequency, maximum singular value and so on. The feature vector is then fed to RF for classification. Features are selected in order of importance to classification to improve the training efficiency. By comparing with various algorithms, the proposed method has higher accuracy and faster training efficiency in single-speed, multi-speed and cross-speed single-sensor data diagnosis. The results show that the method has application prospects with little training data and low hardware requirements.

show abstract

Feedback-driven error-corrected single-sensor analytics for real-time condition monitoring

Cited by 18 publications

References 56 publications

Health indicator construction and application of coal mill based on the dynamic model

Health indicator construction and application of coal mill based on the dynamic model

Dual-input anomaly detection method based on deep reinforcement learning

Single-Sensor Engine Multi-Type Fault Detection

Contact Info

Product

Resources

About