On-line Detection Research Articles

MRMEBP: a unified framework for online detection of atrial fibrillation utilizing deep learning

mRMEBP: a unified framework for online detection of atrial fibrillation utilizing deep learning

An innovative MRAS-based technique for online detection of short-circuit faults in three-phase induction motor windings

An innovative MRAS-based technique for online detection of short-circuit faults in three-phase induction motor windings

Online and Offline Fault Detection and Diagnostics in a Nuclear Power Plant Condenser

Nuclear power plants (NPPs) face significant financial pressures due to operational and maintenance costs. This research investigates Fault Detection and Diagnostics (FDD) techniques to optimize maintenance schedules and reduce expenses. The NPP condenser plays a critical role in converting turbine exhaust steam back into water for reuse. Condenser tube fouling, a prevalent fault mode, impedes heat transfer efficiency and can lead to decreased plant efficiency and safety risks. This study proposes an FDD framework that leverages raw signal analyses from temperature and pressure monitoring to detect and diagnose condenser tube fouling in both online and offline settings. The online approach facilitates close-to-real-time predictions, enabling proactive maintenance strategies. Additionally, the framework explores incorporating a condenser’s maintenance history for enhanced diagnostics. We employ a dataset obtained from a simulated nuclear power plant condenser using the Asherah Nuclear Power Plant Simulator (ANS). ANS replicates the operational dynamics of a pressurized water reactor (PWR) type NPP. The proposed methodology utilizes an encoder-decoder (E-D) structured 1DCNN model to analyze the raw signals. The research demonstrates consistent and accurate fault detection and diagnostics for condenser tube fouling in both online and offline scenarios. A high potential for generalization to unseen conditions was observed. However, online detection using small data windows necessitates caution due to potential false alarms around the transition points. Our findings pave the way for further exploration of robust diagnostics by accommodating a wider spectrum of fouling rates within degradation classes using ANS. This combined online and offline FDD approach offers a promising solution for promoting operational safety, efficiency, and cost-effectiveness in NPP condensers.

A method for enhancing near-mirror object detection by integrating AWCS and CBAM into ResNet18

ABSTRACT Defects, such as scratches, frequently arise during the manufacturing of near-mirror workpieces, adversely affecting both their longevity and aesthetic appeal. The identification of these near-mirror defects through machine vision is fraught with challenges, including the scarcity of comprehensive datasets, variability in surface curvature, and the diminutive size of the defects. In response to these challenges, this study integrates Visual Parser and deep learning methodologies for the visual detection of near-mirror defects. A dataset conducive to deep learning was developed utilising a sliding window approach. The proposed model incorporates an Adaptive Weighted Convolution Structure (AWCS) and Convolutional Block Attention Module (CBAM) within the ResNet18 architecture. Additionally, we introduce an efficient and precise technique termed Visual Parser, which facilitates the detection and localisation of defect areas in near-mirror workpieces. When combined with a classification model, the Visual Parser allows for the online inference and detection of extensive near-mirror images. Experimental findings indicate that the performance metrics of the proposed ResNet18-AWCS-CBAM (ResNet18-ASCM) model significantly exceed those of established networks. The defect detection and localisation method based on Visual Parser is capable of fulfilling the online inspection requirements pertinent to the production of near-mirror workpieces.

Development of an Online Cloud Fog Monitor: Design, Laboratory, and Field Deployment at an Unoccupied Coastal Site in Eastern China

Online detection of cloud water chemistry is a pressing issue in atmospheric outfield observation, with online detection modules representing a significant development direction for cloud water observation. Addressing the common problem of time-delayed errors in manual detection, particularly in the context of cloud water acidity, has remained challenging, with limited understanding and effective solutions available. We developed an Online Cloud Fog Monitor (OCFM) featuring automatic pH and electrical conductivity (EC) detection capabilities, and conducted comprehensive laboratory and field tests. The OCFM utilizes a peristaltic pump, water pipe, and diversion chamber to direct cloud samples to distinct detection chambers, enabling real-time analysis. The diversion chamber is equipped with dual liquid level sensors to segregate and preserve samples once the volume exceeds a predetermined threshold. Calibration results indicate that the instrument's background metal elements do not affect cloud water analysis, and detection occurs within the designed response time. Field tests demonstrate that the OCFM can collect over 50 ml of cloud water, with a response accuracy exceeding 63.6%, though influenced by meteorological conditions. The time-delay error for pH was notably larger than for EC. Comparative analysis with the Caltech Active Strand Cloudwater Collector (CASCC) revealed that the OCFM’s sampling process does not introduce errors, and the online detection accuracy of pH and EC is comparable to manual methods. Additionally, water-soluble ions in samples collected by the OCFM showed no significant differences compared to those collected by CASCC. Overall, the OCFM effectively replaces manual testing, mitigating time-delay errors in chemical property testing. The introduction of this cloud water detector promises to significantly reduce labor costs and economic consumption associated with cloud water observation, thereby facilitating long-term, multi-site observation of cloud water chemistry.

Continuous solid-phase extraction spectroscopy and its quantification method for trace analysis

This study designed and developed an innovative online detection device based on Continuous Solid-Phase Extraction Spectroscopy (CSPES) for rapid quantitative analysis of environmental water pollutants. The device is highly automated, eliminating environmental interference. Leveraging CSPES technology and adsorption kinetics theory, an online quantitative analysis model between the spectrum and component concentrations was established, along with a concentration calculation method based on the least squares method. The quantitative analysis method was validated using single-component and binary-component sample systems containing Fluoranthene, Benzo[k]Fluoranthene, and Rhodamine 6G. The model exhibited excellent predictive performance, with overall prediction concentration relative errors (RE) ranging from 0.45 % to 8.75 % and relative standard deviations (RSD) of less than 3 %. In real sample applications, recovery rates ranged from 86.8 % to 124.4 %, with RSDs between 0.33 % and 2.22 %. This method provides a robust tool for water quality monitoring and environmental analysis, holding significant potential for application across various fields.

Ultratrace CO/SF6 Detection System Based on Differential Fourier Transform Infrared Spectroscopy Combined with the Weighted Sine Spectral Reconstruction Convolutional Neural Network (WSSR-CNN) Model.

Carbon monoxide (CO), as an indicator gas for fault diagnosis of gas-insulated switchgear, has the strongest absorption coefficient in the mid-infrared region, making Fourier transform infrared (FTIR) spectroscopy an ideal method for its concentration detection. However, there is extremely abundant absorption of SF6 in this region, resulting in a great challenge for CO/SF6 detection. To address this challenge, we report a high sensitivity online detection system for ultratrace levels of CO, which combines differential Fourier transform infrared (DFTIR) spectroscopy with a weighted sine spectral reconstruction convolutional neural network (WSSR-CNN). The differential technique is first introduced into FTIR for baseline correction of CO absorption signals. The novel WSSR method achieves feature extraction and denoising of weak spectral signals in strong interference by discretizing interfering signals and enhancing target signals. On this basis, the detection of CO concentration is realized using the CNN model. Finally, WSSR-CNN is compared with four other methods. Results showed that WSSR-CNN outperforms all four models with evaluation index R2 reaching 0.99982 and 0.99999 in the range of low concentration (0.096-0.986 ppm) and high concentration (1.984-52.193 ppm) and mean absolute percentage error of 0.97% and 0.22%, respectively. The lowest detection limit of the system at 1σ is 13 ppb, which is the best result reported so far for detecting CO/SF6 concentration based on FTIR.

Real-time fake news detection in online social networks: FANDC Cloud-based system

Social networks have become a common way for people to communicate with each other and share ideas, thanks to their fast information-sharing features. But fake news spread on social networks can cause many negative consequences by affecting people’s daily lives. However, the literature lacks online and real-time fake news detection systems. This study aims to fill this gap in the literature and to handle the fake news detection problem with a system called FANDC, based on cloud computing, to cope with fake news in seven different categories, and to solve the real-time fake news detection problems. The system was developed using the CRISP-DM methodology with a hybrid approach. BERT algorithm was used in the system running on the cloud to avoid possible cyber threats with the dataset created with approximately 99 million big data from COVID-19-TweetIDs GitHub repository. It was trained in two periods with 100% accuracy during the modeling phase in terms of training accuracy. Experimental results of the FANDC system performed the real-time detection of fake news at 99% accuracy. However, previous studies experimental level success rate in the literature, were around 90%. We hope that the developed system will greatly assist social network users in detecting fake news in real-time.

Fluorescent and colorimetric dual-mode recognition of Al3+ in tablets based on chrome azurol S and pH-switchable boron, nitrogen co-doped carbon dots.

The synthesis and characterization ofpH-sensitive boron and nitrogen co-doped CDs (B, N-CDs) is reported. The fluorescence of B, N-CDs exhibited pH-responsive behavior within the range pH3.0-6.0, attributed to their controlled aggregation and disaggregation. Interestingly, the fluorescence of B, N-CDs was quenched by the inner filter effect from chrome azurol S (CAS). With the subsequent addition of Al3+, the fluorescence of B, N-CDs was restored due to the formation of blue Al3+-CAS and B, N-CDs-Al3+ complexes. Consequently, a fluorescent and colorimetric dual-mode sensor for Al3+ detection was developed. The proposed sensor enabled Al3+ detection in the ranges and limits of detection of 3.75-22.5μM and 2.67μM for the fluorescent mode and 5-20μM and 4.8μM for the colorimetric mode, respectively. Furthermore, the efficacy of the sensor was validated through successful detection of Al3+ in tablet samples. The obtained B, N-CDs promised the application in acid condition and the proposed sensor isexpected tobe applied toin situ on-line detection of Al3+.

Research on Impact Prediction Model for Corn Ears by Integrating Motion Features Using Machine Learning Algorithms

The mechanical damage of corn kernels during harvest leads to mildew in the kernel storage process, seriously affecting food safety and quality. Impact force is the primary source of mechanical damage in the corn threshing process, and its accurate detection is of great significance for corn threshing with low damage. A method for the impact force detection of corn ears was proposed in this manuscript. The momentum theorem determined the main factors influencing impact force (weight, falling height, and space attitude). Corn ear weight, falling height, and space attitude were used as experimental factors. The bench test was carried out with the impact force of corn ear as the output variable. During the experiment, piezoelectric sensors were used to collect the impact force of corn ears under different motion states. Then, the impact force detection model was constructed using four machine learning algorithms: multiple linear regression, ridge regression, random forest, and support vector regression. The results showed that the RF algorithm was more suitable for constructing the prediction model of average and maximum impact force when corn ears fall, SD, RMSE, and r were, respectively: 0.9526, 1.2685, 0.9855; 3.8389, 3.6071, and 0.8510. Secondly, the weight characteristics had the most significant influence on the impact force detection of the ear. Therefore, this method can be used as an accurate, objective, and efficient online detection method for impact force.

Solid-to-gas phase transition kinetics of diverse potassium occurrence forms during biomass pellet combustion: Time-resolved detection and multi-step modeling

The solid-to-gas phase transition of potassium during biomass combustion significantly impacts ash-related issues in bioenergy systems, affecting operational efficiency and equipment longevity. However, the specific mechanisms and kinetics of this transition process remain inadequately understood. This work investigates the time-resolved transition of solid-phase potassium to the gas phase during the combustion of rice husk and wheat straw pellets, combining experimental measurements with theoretical modeling. Tunable diode laser absorption spectroscopy (TDLAS) was employed to measure atomic potassium concentrations 15 mm above burning pellets tray, where gas-phase equilibrium is approached. Key combustion characteristics including thermogravimetric profiles, spectral radiation, and temperature were simultaneously monitored. A novel multi-step model was developed to describe the transition of different forms of solid-phase potassium (organic, exchangeable, and inorganic) to the gas phase. This model integrates TDLAS measurements, observed combustion characteristics, and biomass physicochemical properties. Thermodynamic equilibrium calculations were used to estimate the atomic potassium fraction from total gaseous potassium. The results showed that the solid-to-gas phase transition of organic potassium synchronizes with volatiles release. In contrast, the maximum emission rates of inorganic and exchangeable potassium occurred at the onset of char combustion. The developed model agrees well with the online detection experiments and were further validated by offline ICP analysis of residual ash. While not directly simulating gas-solid interface reactions near the particle surface, this work lays groundwork for future multi-scale modeling of particle-laden flows and reactor-scale phenomena in biomass combustion systems.

Chemical Fingerprinting of PM2.5 via Sequential Speciation Analysis Using Electrochemical Mass Spectrometry.

Chemical fingerprinting to characterize the occurrence state and abundance of organic and inorganic constituents within fine particulate matter (PM2.5) is useful in evaluating the associated health risks and tracing pollution sources. Herein, an analytical strategy for the rapid analysis of metal and organic constituents in PM2.5 was developed employing a combination of sequential chemical extraction coupled with mass spectrometry detection. H2O, CH3OH, EDTA-2Na, electrochemical oxidation, and electrochemical reduction were sequentially utilized to extract the chemical constituents in PM2.5 samples on a homemade device employing simultaneous online detection using two linear trap quadrupole mass spectrometers (LTQ-MS) with electrospray ionization (ESI) in positive and negative modes. After a single analytical procedure, dozens of metals (e.g., Pb, Cr, and Cu), organic compounds (e.g., amines, polycyclic aromatic hydrocarbons, and aliphatic acids), and negative ions (e.g., NO3-, NO2-, and Cl-) were comprehensively detected in the water-soluble, liposoluble, insoluble, oxidizable, and reducible fractions of PM2.5 samples, and their physical and chemical relationships were established.

Unsupervised bearing raceway surface defect detection based on improved f-AnoGAN

Abstract For addressing bearing raceway surface defects, supervised deep learning methods face challenges such as insufficient defect samples and imbalance between defect and non-defect samples. To overcome these issues, we propose an enhanced fast anomaly detection with generative adversarial networks (f-AnoGAN) unsupervised detection algorithm. Firstly, SE-Block modules based on SE attention mechanism and residual structure are integrated into the encoder, aiding the network in focusing on channel information while alleviating gradient vanishing problems. Secondly, transfer learning is introduced to effectively enhance the algorithm’s detection performance and generalization ability. Finally, utilizing bearing images collected from industrial sites, a self-built bearing raceway surface defect dataset is constructed, and extensive experiments are conducted. Experimental results demonstrate that the improved algorithm achieves an area under curve score of 99.96% on the self-built bearing raceway surface defect dataset, representing a 7.07% improvement over the f-AnoGAN algorithm, meeting the requirements for online detection in bearing industry applications.

Research on faults diagnosis of dynamic balance in scroll compressor based on CWT-MViTV2

Abstract The dynamic balancing of rotor systems is a critical factor affecting the vibration of scroll compressors. Traditionally, the detection and correction of rotor dynamic balancing systems have relied on time-domain and frequency-domain analysis of vibration signals, as well as validation using rotor dynamic balancing test platforms. This paper proposes a novel fault diagnosis method for the dynamic balancing of scroll compressor rotor systems. This method acquires the time-domain vibration signals of the scroll compressor through online detection experiments. These signals are then transformed into time-frequency maps using continuous wavelet transform(CWT). By integrating the MViTV2 neural network, this approach enables effective identification of dynamic balancing fault types without requiring disassembly or shutdown of the machine. The fault types, corresponding to the offset of the balancing weight’s centroid, will provide direct data support for subsequent validation work. This study compares various neural network models combined with time-frequency maps and demonstrates that the proposed model achieves the highest accuracy of 99.749% compared to other models, and the method's generalizability is validated in the public dataset. Furthermore, the proposed model maintains a high accuracy of 94.872% in high-noise environments. After improvement, the accuracy of the model has been increased to 95.641%, the training time and the diagnosis time has been reduced to 0.240 iter/s and 0.0277s.

Raman spectroscopic quantification of polyethylene particles in water using polydimethylsiloxane-coated nickel foam as a particle-capturing platform

Nickel foam (NF) was evaluated as a medium for the capture of polyethylene (PE) particles in water. NF is a hydrophobic and porous material with a large surface area, making it a promising candidate for attracting PE particles. However, the particle-capturing efficiency using bare NF was only 69.5%. To increase capturing efficiency, a circular polydimethylsiloxane (PDMS)-coated NF (PDMS@NF, diameter: 6 mm) was employed to enhance the hydrophobicity. The capturing efficiency using the PDMS@NF was substantially increased to 97.6 % owing to the increase in hydrophobicity. To quantify the captured PE particles on/in the PDMS@NF using Raman spectroscopy, a wide area illumination (WAI) scheme providing 6 mm-diameter laser illumination was adopted to fully cover the PDMS@NF for representative spectroscopic sampling and accurate quantification. The intensity ratios of PE to PDMS peaks in the collected spectra clearly increased with the quantity of dispersed PE particles (0.1 ∼ 4.0 mg range, R2: 0.992) in the water samples, and the limit of detection was 0.08 mg. Moreover, the capturing efficiencies for polypropylene (PP), polystyrene (PS), and polyethylene terephthalate (PET) particles (1 mg of each) using the PDMS@NF were also superior, ranging from 96.4 to 98.2 %. Therefore, the proposed scheme incorporating the PDMS@NF as a particle-capturing and Raman measurement platform has potential as a method for on-line detection of microplastics in water.

Real-time monitoring and protection strategies for dense granular flow spallation target in Accelerator-Driven System

An innovative concept of a high-power, gravity-driven, dense granular spallation target has been integrated into the prototype design of the China initiative Accelerator-Driven System (CiADS), which is envisioned as a promising nuclear demonstration facility aimed at converting long-lived fission products and minor actinides into short-lived isotopes. This spallation target system, bridging the proton beam accelerator and the subcritical reactor, plays a crucial role in ensuring safety control and facilitating dynamic operational strategies. In this study, we have developed a system design for an online detector array, incorporating several neutron fission chambers and thermocouples installed in the gap between the spallation target and the subcritical reactor to monitor operations. Furthermore, we have devised real-time monitoring and protection methods to tackle the challenges of measuring beam position and the blockage condition associated with granular flow, considering the distribution of neutron flux and heat deposition under abnormal irradiation conditions. Monte Carlo simulations have demonstrated the applicability and feasibility of the control system for the dense granular spallation target in the accelerator-driven system. The numerical results confirm that the proposed control system meets the requirements for stable measurement with acceptable accuracy.

Development of an online prediction system for soil organic matter and soil moisture content based on multi-modal fusion

Realizing accurate determination of in situ soil organic matter (SOM) content and soil moisture (SM) content in the field is of great importance for improving agricultural production efficiency. However, the feature information of a single sensor is limited, and there is still a certain gap in modeling accuracy compared to traditional laboratory methods. The variations of SM in the field also interfere with data collection, limiting the application of single sensor. In order to realize the efficient detection of in-situ SOM and SM content, this study developed an online detection system for SOM and SM content based on the fusion of characteristic wavelengths, visible images and thermal imaging image features. Firstly, based on a vehicle-mounted platform, a visible-thermal imaging camera and a characteristic wavelength integration device were integrated in a deep pine plow to realize the simultaneous acquisition of in-situ soil multi-sensor data. Then, a lightweight multimodal network was constructed to obtain thermal visible light image features and characteristic wavelength features through branch networks, achieving deep fusion of different modal data and predicting SOM and SM content. Finally, output the forecasting results of SOM and SM content, and transmit the predictive information to the cloud platform for storage. It was verified that the Multi-modal proposed in this study worked best in the laboratory environment, with a predicted R2 of 0.91 and RMSE of 2.9 g/kg for SOM, and a predicted R2 of 0.92 and RMSE of 0.77 % for SM. Compared with single image or characteristic spectral data, the fusion of visible image and characteristic wavelength effectively improved the prediction accuracy of SOM. The real-time prediction of SM was realized by fusing thermal imaging data, and the elimination of the moisture effect of visible images and characteristic wavelengths was also realized with the help of deep fusion of Multi-modal network. After field validation, the R2 of the Multi-modal system was 0.84 and the RMSE was 5.0 g/kg for SOM, and the R2 of the SM was 0.88 and the RMSE was 1.03 %. Despite the differences in soil types between the validation field and the sampling field, the system still demonstrated strong generalization and achieved high accuracy prediction of in-situ SOM and SM in the field, which effectively improves the efficiency and applicability of field. The efficiency and applicability of soil testing effectively improve the efficiency and provide technical guidance for precision management in the field.

KPTr: Key point transformer for LiDAR-based 3D object detection

Accurate 3D environmental perception is crucial for ensuring the safety of autonomous vehicles. However, many existing point-based 3D object detection methods rely only on independent point features to extract spatial information and ignore the dependencies between points. Meanwhile, these methods usually use max-pooling operations to aggregate neighboring point features without considering the stability of local geometric information. To address these issues, this paper proposes a novel two-stage LiDAR-based key point transformer that consists of a region proposal network module and a region-based transformer module to enhance 3D detection performance. Firstly, unprocessed raw point clouds are taken as the input of the model to preserve detailed spatial structural information. Secondly, in the region proposal network an enhanced PointNet++ is used to segment foreground points and generate high-quality first-stage proposals. Then, the representation of proposals is improved by the region-based transformer module, which constructs long-range key point cloud relationship features within the regions of interest and diffuses spatial information into the feature channel layers. Additionally experiments on the KITTI dataset show that detection performance of the proposed method surpasses most mainstream 3D object detection methods and achieves a detection speed of 14.28 Hz, striking a good balance between detection accuracy and speed. Finally, the proposed method is deployed on real vehicle hardware platforms for online detection, thereby proving its industrial value.

In-capillary magnetic separation and enrichment coupled with laser-induced fluorescence for rapid determination of biomolecules

Biomolecules are indispensable in the life activities and metabolism of various organisms, and their highly sensitive detection is of great significance for disease diagnosis and food safety. Due to the complexity of the matrix and the low concentration of the target, separation and enrichment of biomolecules prior to detection are necessary and time-consuming. Herein, a novel approach combining in-capillary magnetic separation and enrichment with laser-induced fluorescence detection method was proposed for qualitative and quantitative analysis of biomolecules. Two different sequence aptamers modified with magnetic nanoparticles and labeled with fluorescence were used as capture probes and signal probes to form sandwich structures with target molecules for detection. Under the combined influence of magnetic field force and fluid drag force, the integration of separation, enrichment and on-line detection were finished within 5min. Under optimal conditions, the developed method provided a low limit of detection as 2.7 fmol/L, 0.57 nmol/L, 3 CFU/mL for DNA, thrombin and Staphylococcus aureus, respectively. This proposed method offers a promising platform for the sensitive and rapid determination of biomolecules. It holds significant potential for future applications in fields of environmental and biochemical analysis.

A smartphone-assisted chromogenic and fluorogenic dual-channel sensor for specific determination of toxic Pb2+ ions with multifaceted applications

Heavy metal pollutants, such as Pb2+ ions, pose a substantial hazard to human health and ecosystems, and it was of great significance to develop cost-effective and portable techniques for online and visual detection of Pb2+. Herein, we fabricated a novel chromogenic and fluorogenic dual-channel sensor NQA through a Schiff base reaction involving a dicyanoisophorone derivative and isoniazid, enabling effective detection of Pb2+ ions. Upon complexation with Pb2+ ions at a 1:1 M ratio, the fluorescence intensity of NQA exhibited a remarkable enhancement. The naked-eye colorimetric detection utilizing NQA demonstrated a distinct color change to specific Pb2+ ions concentration from yellow to light pink. Based on this principle, a smartphone-assisted sensing platform was developed for visual and high-quality identification of Pb2+ in real samples. The UV binding constant for Pb2+ ions was determined to be 1.84 × 105 M−1, with a detection limit of 0.38 μM. Furthermore, the versatility of NQA extended to applications in naked-eye test paper experiments, the detection of Pb2+ ions in real water samples and living cells, showcasing its potential for environmental and biological monitoring.