Historical Sensor Data Research Articles

Advancing Fault Detection Efficiency in Wireless Power Transmission with Light GBM for Real-Time Detection Enhancement

This study introduces WirelessGridBoost, an innovative framework designed to revolutionize real-time fault detection in wireless electrical grids by harnessing the power of the LightGBM machine learning algorithm. Traditional fault detection systems in electrical grids often face challenges such as latency and scalability due to the intricate nature of grid operations and limitations in communication infrastructure. To overcome these challenges, WirelessGridBoost integrates LightGBM, a highly efficient gradient boosting decision tree algorithm, with wireless technology to facilitate advanced fault detection capabilities. Trained on historical sensor data, the LightGBM model demonstrates exceptional proficiency in discerning complex fault patterns inherent in electrical grid operations. Deployed across strategically positioned wireless nodes within the grid, WirelessGridBoost enables prompt identification of anomalies in real-time. Extensive simulations and experiments conducted on a real-world grid testbed validate the effectiveness of WirelessGridBoost, achieving a fault detection accuracy of 96.80% and reducing latency by 38% compared to conventional methods. This research presents a promising avenue for enhancing fault detection efficiency in wireless electrical grids through the innovative WirelessGridBoost framework.

Noise resistant steam generator water level reconstruction for nuclear power plant based on deep residual shrinkage network

Nuclear power plant (NPP) steam generator (SG) water level has strong nonlinear pattern and lagging characteristic under control system and manual adjustment. SG water level signal accidental missing and frequent abnormity may lead to incorrect operational intervention or even emergency shutdown. Deep learning SG water level signal reconstruction method trained with historical sensor data can provide supportive information for operators’ decision-making to decrease the financial cost of maintenance and improve the safety of NPP assets. However, the sensor data transferred by electronic devices can be mixed up with noise in transients or complex conditions. This study proposed a noise resistant SG water level signal reconstruction method based on deep residual shrinkage network (DRSN). The dataset in this study is collected from a power-down process collected from a NPP in China. Results from the experiments under different levels of noise are conducted to illustrate the efficacy and robustness of the proposed approaches compared with other deep learning signal reconstruction methods.

Lessons learned on the implementation of probabilistic graphical model-based digital twins: A space habitat study

Habitats for future human spaceflights will require more resilient environmental control and life support systems (ECLSS). To that end, it is important to facilitate decision making in case of unexpected failure by quantifying the uncertain and dynamic nature of the physical phenomena involved. Combining probabilistic and deterministic models is a particularly promising approach to address this issue. In particular, Probabilistic Graphical Model (PGM) based digital twins are relevant as they embed random variables evolving overtime. Previous research used this modeling method for several applications such as monitoring structural health or manufacturing processes. We envision that the space exploration sector can also benefit from this approach by using the insight gained on specific sub-systems. In this study, we propose lessons learned on the implementation process of PGM-based Digital Twin to quantify uncertainties for temperature prognosis in ECLSS. These findings are introduced as a step-by-step guideline which result in developing a probabilistic model applicable to space habitats. We focused on directed acyclic graphs as this type of PGM can integrate expert's knowledge with data which has been proven to enhance accuracy. A literature review was conducted to identify the state-of-the-art practices and the proposed lessons learned were derived from the study of a physical infrastructure meant to predict the behavior of a space habitat. A temperature control failure scenario was considered, and the Digital Twin estimated the time available before the temperature would become critical. Experiments were conducted on three office rooms to simulate the behavior of an ECLSS. The model was trained offline using historical sensor data and performed inference online by computing the conditional probability of a multivariate normal density. We found that a successful implementation process requires to iteratively go through four stages: outline, design, calibrate and evaluate. It involves selecting ECLSS relevant functionalities and an associated decision-making problem that relies on habitability criteria. Observable variables must be chosen according to a sensors architecture that is compatible with a typical habitat infrastructure. As real space systems are not easily available for model validation, we suggest evaluating early designs on high fidelity analogs. In future work, we envisage to further assess the impact of the design stage on the model's performance by considering computational cost and inference capability.

Multiannuli Solution Uses Intelligent Pipe for Real-Time Monitoring

_ This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 211546, “Intelligent Pipe: A Multiannuli Monitoring Solution,” by Francois-Xavier Bulard, Vallourec; Emmanuel Tavernier, OpenField Technology; and Lucas Kling, Vallourec. The paper has not been peer reviewed. _ The authors describe a multiannular monitoring solution that measures pressure and temperature for each well annulus. These key parameters are captured by sensors based on microelectromechanical-system technology. A transmission methodology is used that allows power supply and data retrieval even for annuli beyond the B annulus. All electronic elements of the solution will be qualified through an international recommendation protocol [Advanced Well Equipment Standards (AWES)] that consists of a series of mechanical and long-term tests to ensure that the system will survive for 20 years at 125°C. The solution could be used in all storage applications, especially those related to the energy transition. Introduction The initial goal of the intelligent-pipe solution was to add functionality to tubulars by embedding miniaturized sensors into the pipe body. The secondary objective was to monitor the pipe’s environment to properly characterize well loads and compare them to its strength resistance. The main application of the solution was then to monitor annulus conditions in real time during the whole life of the well and to anticipate any leakage or annular-pressure-buildup issue that could not be monitored currently on subsea wells. A tubular design was developed to incorporate pressure and temperature sensors without affecting the integrity of the pipe. The design was verified by a testing protocol performed by an in-house testing laboratory. This protocol consisted, first, of finite-element analysis (FEA) performed on different tubular designs with the objective of optimizing the different shapes of the grooves created to incorporate the sensors. The second steps included physical tests performed on a specimen inside test frames that confirmed that the tubular integrity was not affected after several pressure and axial load cycles, even when applying maximum pipe-resistance loads. This mechanical test qualification demonstrated that the monitoring mandrel was able to withstand the constraints observed in a well. The qualification level achieved was the same as that achieved on the premium pipe connection. Much of the complete paper is dedicated to the development of the permanent solution; this synopsis focuses on the system’s pilot deployment and its global system qualification. First Deployment in a Deepwater Subsea Well The operator’s objective was to monitor several annuli of a subsea well to determine the effect of completion operations. This characterization was completed before the well-production phase. The main outputs of the pilot were annuli trapped pressure and temperature. The pilot was defined initially in two stages. The first was to monitor the B and C annuli conditions (pressure and temperature) immediately after casing deployment. The second was to do so again after running completions and various stimulation operations. A secondary objective was to retrieve historical sensor data, pressure, and temperature records from the time of sensor deployment to data-retrieval operation.

Adaptive attitude determination of bionic polarization integrated navigation system based on reinforcement learning strategy

<p style='text-indent:20px;'>The bionic polarization integrated navigation system includes three-axis gyroscopes, three-axis accelerometers, three-axis magnetometers, and polarization sensors, which provide pitch, roll, and yaw. When the magnetometers are interfered or the polarization sensors are obscured, the accuracy of attitude will be decreased due to abnormal measurement. To improve the accuracy of attitude of the integrated navigation system under these complex environments, an adaptive complementary filter based on DQN (Deep Q-learning Network) is proposed. The complementary filter is first designed to fuse the measurements from the gyroscopes, accelerometers, magnetometers, and polarization sensors. Then, a reward function of the bionic polarization integrated navigation system is defined as the function of the absolute value of the attitude angle error. The action-value function is introduced by a fully-connected network obtained by historical sensor data training. The strategy can be calculated by the deep Q-learning network and the action that optimal action-value function is obtained. Based on the optimized action, three types of integration are switched automatically to adapt to the different environments. Three cases of simulations are conducted to validate the effectiveness of the proposed algorithm. The results show that the adaptive attitude determination of bionic polarization integrated navigation system based on DQN can improve the accuracy of the attitude estimation.</p>

An Ordinal Multi-Dimensional Classification (OMDC) for Predictive Maintenance

Predictive Maintenance is a type of condition-based maintenance that assesses the equipment's states and estimates its failure probability and when maintenance should be performed. Although machine learning techniques have been frequently implemented in this area, the existing studies disregard to the natural order between the target attribute values of the historical sensor data. Thus, these methods cause losing the inherent order of the data that positively affects the prediction performances. To deal with this problem, a novel approach, named Ordinal Multi-dimensional Classification (OMDC), is proposed for estimating the conditions of a hydraulic system's four components by taking into the natural order of class values. To demonstrate the prediction ability of the proposed approach, eleven different multi-dimensional classification algorithms (traditional Binary Relevance (BR), Classifier Chain (CC), Bayesian Classifier Chain (BCC), Monte Carlo Classifier Chain (MCC), Probabilistic Classifier Chain (PCC), Classifier Dependency Network (CDN), Classifier Trellis (CT), Classifier Dependency Trellis (CDT), Label Powerset (LP), Pruned Sets (PS), and Random k-Labelsets (RAKEL)) were implemented using the Ordinal Class Classifier (OCC) algorithm. Besides, seven different classification algorithms (Multilayer Perceptron (MLP), Support Vector Machine (SVM), k-Nearest Neighbour (kNN), Decision Tree (C4.5), Bagging, Random Forest (RF), and Adaptive Boosting (AdaBoost)) were chosen as base learners for the OCC algorithm. The experimental results present that the proposed OMDC approach using binary relevance multi-dimensional classification methods predicts the conditions of a hydraulic system's multiple components with high accuracy. Also, it is clearly seen from the results that the OMDC models that utilize ensemble-based classification algorithms give more reliable prediction performances with an average Hamming score of 0.853 than the others that use traditional algorithms as base learners.

Predicting river water height using deep learning-based features

The paper presents the river height prediction model using real-world historical sensor data such as rainfall, cumulative rainfall, and river water heights. The study evaluates using a Support Vector Regression, a Long Short-Term Memory, and a combination of a Long Short-Term Memory as the feature extraction and a support vector regression. Through experiments, various future predictions are tested, including a few hours or a day. As expected, RNN achieved the lowest error, but it could not capture rapid changes in river height levels. In comparison, the LSTM-SVR can better represent rapid transient changes in the data by using nonlinear kernels.

Real-time coal classification in thermal power plants

Continuous variations in coal quality have a strong influence on operation of thermal power plants. However, in absence of real-time measurements of quality, the operators remain partially blind to the actual coal being consumed, leading to sub-optimal operation. A two-stage solution is proposed for real-time soft sensing of coal type. First, a novel semi-supervised cascaded clustering algorithm (SSCC) is utilized to extract coal classes from the historical sensor data of the plant and create a coal class library. The online stage includes a coal change detection algorithm to detect transition of coal and a SSCC-based coal classifier that enables real-time classification of coal using only the live sensor data. The algorithms are tested and verified with a set of synthetically generated industrial scale coal mill operation data. The real-time classification of coal can facilitate continuous optimum operation of plant vis-à-vis the emissions, efficiency and maintenance.

Sensor-based surveillance for digitising real-time COVID-19 tracking in the USA (DETECT): a multivariable, population-based, modelling study

BackgroundTraditional viral illness surveillance relies on in-person clinical or laboratory data, paper-based data collection, and outdated technology for data transfer and aggregation. We aimed to assess whether continuous sensor data can provide an early warning signal for COVID-19 activity as individual physiological and behavioural changes might precede symptom onset, care seeking, and diagnostic testing.MethodsThis multivariable, population-based, modelling study recruited adult (aged ≥18 years) participants living in the USA who had a smartwatch or fitness tracker on any device that connected to Apple HealthKit or Google Fit and had joined the DETECT study by downloading the MyDataHelps app. In the model development cohort, we included people who had participated in DETECT between April 1, 2020, and Jan 14, 2022. In the validation cohort, we included individuals who had participated between Jan 15 and Feb 15, 2022. When a participant joins DETECT, they fill out an intake survey of demographic information, including their ZIP code (postal code), and surveys on symptoms, symptom onset, and viral illness test dates and results, if they become unwell. When a participant connects their device, historical sensor data are collected, if available. Sensor data continue to be collected unless a participant withdraws from the study. Using sensor data, we collected each participant's daily resting heart rate and step count during the entire study period and identified anomalous sensor days, in which resting heart rate was higher than, and step count was lower than, a specified threshold calculated for each individual by use of their baseline data. The proportion of users with anomalous data each day was used to create a 7-day moving average. For the main cohort, a negative binomial model predicting 7-day moving averages for COVID-19 case counts, as reported by the Centers for Disease Control and Prevention (CDC), in real time, 6 days in the future, and 12 days in the future in the USA and California was fitted with CDC-reported data from 3 days before alone (H0) or in combination with anomalous sensor data (H1). We compared the predictions with Pearson correlation. We then validated the model in the validation cohort.FindingsBetween April 1, 2020, and Jan 14, 2022, 35 842 participants enrolled in DETECT, of whom 4006 in California and 28 527 in the USA were included in our main cohort. The H1 model significantly outperformed the H0 model in predicting the 7-day moving average COVID-19 case counts in California and the USA. For example, Pearson correlation coefficients for predictions 12 days in the future increased by 32·9% in California (from 0·70 [95% CI 0·65–0·73] to 0·93 [0·92–0·94]) and by 12·2% (from 0·82 [0·79–0·84] to 0·92 [0·91–0·93]) in the USA from the H0 model to the H1 model. Our validation model also showed significant correlations for predictions in real time, 6 days in the future, and 12 days in the future.InterpretationOur study showed that passively collected sensor data from consenting participants can provide real-time disease tracking and forecasting. With a growing population of wearable technology users, these sensor data could be integrated into viral surveillance programmes.FundingThe National Center for Advancing Translational Sciences of the US National Institutes of Health, The Rockefeller Foundation, and Amazon Web Services.

ECoalVis: Visual Analysis of Control Strategies in Coal-fired Power Plants.

Improving the efficiency of coal-fired power plants has numerous benefits. The control strategy is one of the major factors affecting such efficiency. However, due to the complex and dynamic environment inside the power plants, it is hard to extract and evaluate control strategies and their cascading impact across massive sensors. Existing manual and data-driven approaches cannot well support the analysis of control strategies because these approaches are time-consuming and do not scale with the complexity of the power plant systems. Three challenges were identified: a) interactive extraction of control strategies from large-scale dynamic sensor data, b) intuitive visual representation of cascading impact among the sensors in a complex power plant system, and c) time-lag-aware analysis of the impact of control strategies on electricity generation efficiency. By collaborating with energy domain experts, we addressed these challenges with ECoalVis, a novel interactive system for experts to visually analyze the control strategies of coal-fired power plants extracted from historical sensor data. The effectiveness of the proposed system is evaluated with two usage scenarios on a real-world historical dataset and received positive feedback from experts.

BIM and IoT Sensors Integration: A Framework for Consumption and Indoor Conditions Data Monitoring of Existing Buildings

The low accessibility to the information regarding buildings current performances causes deep difficulties in planning appropriate interventions. Internet of Things (IoT) sensors make available a high quantity of data on energy consumptions and indoor conditions of an existing building that can drive the choice of energy retrofit interventions. Moreover, the current developments in the topic of the digital twin are leading the diffusion of Building Information Modeling (BIM) methods and tools that can provide valid support to manage all data and information for the retrofit process. This paper shows the aim and the findings of research focused on testing the integrated use of BIM methodology and IoT systems. A common data platform for the visualization of building indoor conditions (e.g., temperature, luminance etc.) and of energy consumption parameters was carried out. This platform, tested on a case study located in Italy, is developed with the integration of low-cost IoT sensors and the Revit model. To obtain a dynamic and automated exchange of data between the sensors and the BIM model, the Revit software was integrated with the Dynamo visual programming platform and with a specific Application Programming Interface (API). It is an easy and straightforward tool that can provide building managers with real-time data and information about the energy consumption and the indoor conditions of buildings, but also allows for viewing of the historical sensor data table and creating graphical historical sensor data. Furthermore, the BIM model allows the management of other useful information about the building, such as dimensional data, functions, characteristics of the components of the building, maintenance status etc., which are essential for a much more conscious, effective and accurate management of the building and for defining the most suitable retrofit scenarios.

Prediction of boiler gas side effective heat transfer coefficients using mixture density networks and historic plant data

Machine learning has received increased recognition for applications in engineering such as the thermal engineering discipline due to its abilities to circumvent thermodynamic simulation approaches and capture complex inter-dependencies. There have been recent headways to couple deep learning models to process simulations, given the deeper insight they can provide. The present study entails the development of a mixture density network (MDN) capable of predicting effective heat transfer coefficients for the various heat exchanger components of a utility scale boiler. Large boilers are susceptible to dead zones and other anomalous phenomena that influence performance and manifest as multimodalities in the measured data, which system-level 1D process models struggle to capture. The overall water-side heat load calculation, as well as mass and energy balances around the components were done to determine the heat transfer coefficients at each stage of the boiler using historic sensor data. The measured data was then used to train a deep learning model capable of outputting predicted heat transfer coefficients and local model uncertainty. The predictive model can be coupled to a water circuit process model which can be used to study aspects such as metal temperatures and operating philosophies at the different operating loads of the plant.

An explicit evolutionary approach for multiobjective energy consumption planning considering user preferences in smart homes

Modern Smart Cities are highly dependent on an efficient energy service since electricity is used in an increasing number of urban activities. In this regard, Time-of-Use prices for electricity is a widely implemented policy that has been successful to balance electricity consumption along the day and, thus, diminish the stress and risk of shortcuts of electric grids in peak hours. Indeed, residential customers may now schedule the use of deferrable electrical appliances in their smart homes in off-peak hours to reduce the electricity bill. In this context, this work aims to develop an automatic planning tool that accounts for minimizing the electricity costs and enhancing user satisfaction, allowing them to make more efficient usage of the energy consumed. The household energy consumption planning problem is addressed with a multiobjective evolutionary algorithm, for which problem-specific operators are devised, and a set of state-of-the-art greedy algorithms aim to optimize different criteria. The proposed resolution algorithms are tested over a set of realistic instances built using real-world energy consumption data, Time-of-Use prices from an electricity company, and user preferences estimated from historical information and sensor data. The results show that the evolutionary algorithm is able to improve upon the greedy algorithms both in terms of the electricity costs and user satisfaction and largely outperforms to a large extent the current strategy without planning implemented by users.

Transfer learning driven sequential forecasting and ventilation control of PM2.5 associated health risk levels in underground public facilities

Particulate matter with aerodynamic diameter less than 2.5 µm (PM2.5) has become a major public concern in closed indoor environments, such as subway stations. Forecasting platform PM2.5 concentrations is significant in developing early warning systems, and regulating ventilation systems to ensure commuter health. However, the performance of existing forecasting approaches relies on a considerable amount of historical sensor data, which is usually not available in practical situations due to hostile monitoring environments or newly installed equipment. Transfer learning (TL) provides a solution to the scant data problem, as it leverages the knowledge learned from well-measured subway stations to facilitate predictions on others. This paper presents a TL-based residual neural network framework for sequential forecast of health risk levels traced by subway platform PM2.5 levels. Experiments are conducted to investigate the potential of the proposed methodology under different data availability scenarios. The TL-framework outperforms the RNN structures with a determination coefficient (R2) improvement of 42.84%, and in comparison, to stand-alone models the prediction errors (RMSE) are reduced up to 40%. Additionally, the forecasted data by TL-framework under limited data scenario allowed the ventilation system to maintain IAQ at healthy levels, and reduced PM2.5 concentrations by 29.21% as compared to stand-alone network.

Discovery and Recognition of Emerging Human Activities Using a Hierarchical Mixture of Directional Statistical Models

Human activity recognition plays a significant role in enabling pervasive applications as it abstracts low-level noisy sensor data into high-level human activities, which applications can respond to. With more and more activity-aware applications deployed in real-world environments, a research challenge emerges-discovering and learning new activities that have not been pre-defined or observed in the training phase. This paper tackles this challenge by proposing a hierarchical mixture of directional statistical models. The model supports incrementally, continuously updating the activity model over time with the reduced annotation effort and without the need for storing historical sensor data. We have validated this solution on four publicly available, third-party smart home datasets, and have demonstrated up to 91.5 percent accuracies of detecting and recognising new activities.

Moodable: On feasibility of instantaneous depression assessment using machine learning on voice samples with retrospectively harvested smartphone and social media data

Abstract Depression is a leading cause of disability and is associated with suicide risk. However, a quarter of patients with major depression remain undiagnosed. Prior work has demonstrated that a smartphone user's depression level can be detected by analyzing data gathered from their smartphone's sensors or from their social media posts over a few weeks after enrollment in a user study. These studies typically utilize a prospective study design, which is burdensome as it requires participants smartphone data to be gathered for prolonged periods before their depression level can be assessed. In contrast, we present a feasibility study of our Mood Assessment Capable Framework (Moodable) that facilitates almost instantaneous mood assessment by analyzing instantaneous voice samples provided by the user as well as historical sensor data harvested (scraped) from their smartphone and recent social media posts. Our retrospective, low-burden approach means that Moodable no longer requires study participants to engage with their phone for weeks before a depression score can be inferred. Moodable has the potential to minimize user data collection burden, increase user compliance, avoid study awareness bias and offer a near instantaneous depression screening. To lay a solid foundation for Moodable, we first surveyed 202 volunteer participants about their willingness to share voice samples and various smartphone and social media data types for mental health assessment. Based on these findings, we then developed the Moodable app. Thereafter, we utilized Moodable to collect short voice samples, and a rich array of retrospectively harvested data from users' smartphones (location, browser history, call logs) and social media accounts (instagram, twitter and facebook), with appropriate permissions, of 335 volunteer participants who also responded to 9 depression related questions of the Patient Health Questionaire (PHQ-9). Moodable then used machine learning to build classification models and classify the user's depression and suicidal ideation, for users which scores where unknown to the models. Results of Moodable's screening capability are promising. In particular, for the depression classification task we achieved F1 scores (the harmonic mean of the precision and recall) of 0.766, sensitivity of 0.750, and specificity of 0.792. For the suicidal ideation task we achieved F1 scores of 0.848, sensitivity of 0.864, and specificity of 0.725. This work could significantly increase depression-screening at the population level and opens numerous avenues for further research into this newly proposed paradigm of instantaneously screening depression and suicide risk levels from voice samples and retrospective smartphone and social media data.

Sensor Data Visualization for Indoor Point Clouds

Abstract. Integration and analysis of real-time and historic sensor data provides important insights into the operational status of buildings. There is a need for the integration of sensor data and digital representations of the built environment for furthering stakeholder engagement within the realms of Real Estate 4.0 and Facility Management (FM), especially in a spatial representation context. In this paper, we propose a general system architecture that integrates point cloud data and sensor data for visualization and analysis. We further present a prototypical web-based implementation of that architecture and demonstrate its application for the integration and visualization of sensor data from a typical office building, with the aim to communicate and analyze occupant comfort. The empirical results obtained from our prototypical implementation demonstrate the feasibility of our approach for the provisioning of light-weight software components for the service-oriented integration of Building Information Modeling (BIM), Building Automation Systems (BASs), Integrated Workplace Management Systems (IWMSs), and future Digital Twin (DT) platforms.

Cloud Detection with Historical Geostationary Satellite Sensors for Climate Applications

Can we build stable Climate Data Records (CDRs) spanning several satellite generations? This study outlines how the ClOud Fractional Cover dataset from METeosat First and Second Generation (COMET) of the EUMETSAT Satellite Application Facility on Climate Monitoring (CM SAF) was created for the 25-year period 1991–2015. Modern multi-spectral cloud detection algorithms cannot be used for historical Geostationary (GEO) sensors due to their limited spectral resolution. We document the innovation needed to create a retrieval algorithm from scratch to provide the required accuracy and stability over several decades. It builds on inter-calibrated radiances now available for historical GEO sensors. It uses spatio-temporal information and a robust clear-sky retrieval. The real strength of GEO observations—the diurnal cycle of reflectance and brightness temperature—is fully exploited instead of just accounting for single “imagery”. The commonly-used naive Bayesian classifier is extended with covariance information of cloud state and variability. The resulting cloud fractional cover CDR has a bias of 1% Mean Bias Error (MBE), a precision of 7% bias-corrected Root-Mean-Squared-Error (bcRMSE) for monthly means, and a decadal stability of 1%. Our experience can serve as motivation for CDR developers to explore novel concepts to exploit historical sensor data.

Online Fault Diagnosis Method Based on Transfer Convolutional Neural Networks

Fault detection and diagnosis (FDD) is crucial for stable, reliable, and safe operation of industrial equipment. In recent years, deep learning models have been widely used in data-driven FDD methods because of their automatic feature learning capability. In general, these models are trained on historical sensor data, and therefore, it is very difficult to meet the real-time requirement of online FDD applications. Since transfer learning can solve different but similar problems in the target domain efficiently and effectively with the knowledge learned from the source domain, this paper proposes an online fault diagnosis method based on a deep transfer convolutional neural network (TCNN) framework. The TCNN framework is made up of an online CNN based on LeNet-5 and several offline CNNs with a shallow structure. First, time-domain signal data are converted into images that contain abundant fault information and are suitable as the input of CNN. Then, the online CNN is constructed to automatically extract representative features from the converted images and classify faults. Finally, in order to improve the real-time performance of the online CNN, several offline CNNs are also constructed and pretrained on related data sets. By directly transferring the shallow layers of the trained offline CNNs to the online CNN, the online CNN can significantly improve the real-time performance and successfully address the issue of achieving the desired diagnostic accuracy within limited training time. The proposed method is validated on two bearing data sets and one pump data set, respectively. The prediction accuracy of the proposed method using three data sets are 99.88%, 99.13%, and 99.98%, respectively. The experimental results also indicate that the improvement of accuracy is 19.21% for the motor bearing case, 29.82% for the rolling mill bearing case, and 33.26% for the pump case during the early stage of learning.

Ontology–based access to temporal data with Ontop: A framework proposal

Abstract Predictive analysis gradually gains importance in industry. For instance, service engineers at Siemens diagnostic centres unveil hidden knowledge in huge amounts of historical sensor data and use it to improve the predictive systems analysing live data. Currently, the analysis is usually done using data-dependent rules that are specific to individual sensors and equipment. This dependence poses significant challenges in rule authoring, reuse, and maintenance by engineers. One solution to this problem is to employ ontology-based data access (OBDA), which provides a conceptual view of data via an ontology. However, classical OBDA systems do not support access to temporal data and reasoning over it. To address this issue, we propose a framework for temporal OBDA. In this framework, we use extended mapping languages to extract information about temporal events in the RDF format, classical ontology and rule languages to reflect static information, as well as a temporal rule language to describe events. We also propose a SPARQL-based query language for retrieving temporal information and, finally, an architecture of system implementation extending the state-of-the-art OBDA platform Ontop.