Virtual Sensor Approach Research Articles

Advanced Machine Learning Techniques for Corrosion Rate Estimation and Prediction in Industrial Cooling Water Pipelines.

This paper presents the results of a study on data preprocessing and modeling for predicting corrosion in water pipelines of a steel industrial plant. The use case is a cooling circuit consisting of both direct and indirect cooling. In the direct cooling circuit, water comes into direct contact with the product, whereas in the indirect one, it does not. In this study, advanced machine learning techniques, such as extreme gradient boosting and deep neural networks, have been employed for two distinct applications. Firstly, a virtual sensor was created to estimate the corrosion rate based on influencing process variables, such as pH and temperature. Secondly, a predictive tool was designed to foresee the future evolution of the corrosion rate, considering past values of both influencing variables and the corrosion rate. The results show that the most suitable algorithm for the virtual sensor approach is the dense neural network, with MAPE values of (25 ± 4)% and (11 ± 4)% for the direct and indirect circuits, respectively. In contrast, different results are obtained for the two circuits when following the predictive tool approach. For the primary circuit, the convolutional neural network yields the best results, with MAPE = 4% on the testing set, whereas for the secondary circuit, the LSTM recurrent network shows the highest prediction accuracy, with MAPE = 9%. In general, models employing temporal windows have emerged as more suitable for corrosion prediction, with model performance significantly improving with a larger dataset.

Virtual measurements for exterior vibro-acoustic problems using experimental modal models

This paper presents a virtual sensor scheme for vibroacoustic radiation applications based on experimental modal state-space models and sparse measurements. A Kalman filter is used for the estimation combining data from the acoustical and the structural domain together with the numerical modal process model. It is proposed to use Gramians, as an energy-related measure of a system’s observability which enables assessing the observability of the coupled vibroacoustic system in the modal state-space domain and the physical state domain counterpart. This Gramian-based observability metric reveals modes or states of high observability, which ensure good estimation results in the subsequent estimation. The proposed method is validated using an industrial test setup for which, due to its high complexity, modelling with first-principle techniques is not feasible. The estimation results demonstrate improved performance compared to a pure forward use of the identified model, indicating the effectiveness of the virtual sensor approach. Additionally, the newly proposed Gramian-based observability criterion outperforms state-of-the-art metrics for sensor placement in terms of estimation accuracy for particular states of interest, as demonstrated in a sensor placement task. Overall, this paper provides a promising approach for virtual sensing for vibroacoustic applications, enabling accurate and efficient system monitoring and control.

Virtual sensors for wind turbines with machine learning‐based time series models

AbstractModern wind turbines have multiple sensors installed and provide constant data stream outputs; however, there are some important quantities where installing physical sensors is either impractical or the sensor technology is not sufficiently advanced. An example of such a problem is, for example, sensing the shape and location of wake‐induced wind deficits caused by upwind turbines—a feature which would have relevant application in wind farm control; however, it is hard to detect physically due to the need of scanning the airflow in front of the turbine in multiple locations. Another control‐related example is monitoring and predicting the blade tip‐tower clearance. A “virtual sensor” can be created instead, by establishing a mathematical relationship between the quantity of interest and other, measurable quantities such as readings from already available sensors (e.g., SCADA, lidars, and met‐masts). Machine Learning (ML) approaches are suitable for this task as ML algorithms are capable of learning and representing complex relationships. This study details the concept of ML‐based virtual sensors and showcases three specific examples: blade root bending moment prediction, detection of wind turbine wake center location, and forecasting of blade tip‐tower clearance. All examples utilize sequence models (Long Short‐Term Memory, LSTM) and use aeroelastic load simulations to generate wind turbine response time series and test model performance. The data types used in the examples correspond to channels that would be available from high‐frequency SCADA data combined with a blade and tower load measurement system. The resulting model performance demonstrates the feasibility of the ML‐based virtual sensor approach.

A novel principal component-based virtual sensor approach for efficient classification of gases/odors

Abstract High-performance detection and estimation of gases/odors are challenging, especially in real-time gas sensing applications. Recently, efficient electronic noses (e-noses) are being developed using convolutional neural networks (CNNs). Further, CNNs perform better when they operate on a minimal size of vector response. In this paper, dimensions of the operational vectors have been augmented by using virtual sensor responses. These virtual responses are obtained from the principal components of the physical sensor responses. Accordingly, two sets of data are upscaled as a one-dimensional one. Another level of upscaling is further obtained by using the mirror mosaicking technique. Hence, with our proposed novel approach, the final vector size for CNN operations achieves a new dimension. With this upscaled hybrid dataset, consisting of physical and virtual sensor responses, a simpler CNN has achieved 100 percent correct classification in two different experimental settings. To the best of authors information, it is for the first time that an e-nose has been designed using a principal component-based hybrid, upscaled dataset and achieves 100 percent correct classification of the considered gases/odors.

Digital twin-driven virtual sensor approach for safe construction operations of trailing suction hopper dredger

The stable and safe operation of Trailing suction hopper dredger (TSHD) is one of the most crucial considerations for ensuring its high dredging productivity. However, the instability and sudden failure of physical sensors pose challenges to the monitoring of dredging process. To address these issues, we propose a structure of digital twin-driven virtual sensor (DTDVS) for the construction safety of TSHD. Considering the potential internal relations among construction data, we compare the performance of four machine learning algorithms in predicting the torsional vibration in mechanical failure. The results showed that these algorithms provide high prediction accuracy (R2 > 0.9). Then the DBN model with the best performance was selected as a part of the virtual sensors to predict and analyze the status of TSHD. The digital twin technology provides a more stable and environmentally friendly scheme for TSHD construction safety control. On the one hand, the DTDVS assists physical sensors to monitor the construction state, overcoming the limitation of the sensors on detection targets which are difficult or costly to measure directly. On the other hand, by analyzing the residual between the physical sensor and the virtual sensor, the construction behavior can be diagnosed, and the fault situation can be pre-warned accurately. This improves the time utilization of TSHD and provides an important guarantee for the construction safety.

Hardware-Based Intrusion Detection in E-LAN based Distributed DC Microgrid: A Virtual Sensor Approach

Hardware-Based Intrusion Detection in E-LAN based Distributed DC Microgrid: A Virtual Sensor Approach

Innovative Actuator Fault Identification Based on Back Electromotive Force Reconstruction

The ever increasing adoption of electrical power as secondary form of on-board power is leading to an increase in the usage of electromechanical actuators (EMAs). Thus, in order to maintain an acceptable level of safety and reliability, innovative prognostics and diagnostics methodologies are needed to prevent performance degradation and/or faults propagation. Furthermore, the use of effective prognostics methodologies carries several benefits, including improved maintenance schedule capability and relative cost decrease, better knowledge of systems health status and performance estimation. In this work, a novel, real-time approach to EMAs prognostics is proposed. The reconstructed back electromotive force (back-EMF), determined using a virtual sensor approach, is sampled and then used to train an artificial neural network (ANN) in order to evaluate the current system status and to detect possible coils partial shorts and rotor imbalances.

Smart Sensors Applications for a New Paradigm of a Production Line.

Industrial plants are going to face a deep renewing process within the Industry 4.0 scenario. New paradigms of production lines are foreseen in the very near future, characterized by a strict collaboration between humans and robots and by a high degree of flexibility. Such envisaged improvements will require the smart use of proper sensors at very different levels. This paper investigates three different aspects of this industrial renewing process, based on three different ways of exploiting sensors, toward a new paradigm of a production line. The provided contributions, offering various types of innovation and integration, are relative to: (i) a virtual sensor approach for manual guidance, increasing the potentialities of a standard industrial manipulator, (ii) a smart manufacturing solution to assist the operator’s activity in manual assembly stations, through an original exploitation of multiple sensors, and (iii) the development of an advanced robotic architecture for a flexible production line, in which a team of autonomous mobile robots acts as a meta-sensor net supporting traditional automated guided vehicles. Accurate analyses of existing state-of-the-art solutions compared with the proposed ones are offered for the considered issues.

Neural Virtual Sensors for Adaptive Magnetic Localization of Autonomous Dataloggers

The surging advance in micro- and nanotechnologies allied with neural learning systems allows the realization of miniaturized yet extremely powerful multisensor systems and networks for wide application fields, for example, in measurement, instrumentation, automation, and smart environments. Time and location context is particularly relevant to sensor swarms applied for distributed measurement in industrial environment, such as, for example, fermentation tanks. Common RF solutions face limits here, which can be overcome by magnetic systems. Previously, we have developed the electronic system for an integrated data logger swarm with magnetic localization and sensor node timebase synchronization. The focus of this work is on an approach to improving both localization accuracy and flexibility by the application of artificial neural networks applied as virtual sensors and classifiers in a hybrid dedicated learning system. Including also data from an industrial brewery environment, the best investigated neural virtual sensor approach has achieved an advance in localization accuracy of a factor of 4 compared to state-of-the-art numerical methods and, thus, results in the order of less than 5 cm meeting industrial expectations on a feasible solution for the presented integrated localization system solution.

A combined Moving Horizon and Direct Virtual Sensor approach for constrained nonlinear estimation

This paper presents a novel approach to design estimators for nonlinear systems. The approach is based on a combination of linear Moving Horizon Estimation (MHE) and Direct Virtual Sensor (DVS) techniques, and allows the design of estimators with guaranteed stability, which can account for convex constraints on the variables to be estimated. It is also shown that the designed estimators are optimal, in the sense that they give minimal worst-case estimation error, on the basis of the available finite number of noise-corrupted data, with respect to an ideal MHE filter (obtained by assuming exact knowledge of the system dynamics and of the global solution of the related nonlinear program). The approach is tested on a nonlinear mass–spring–damper system.

From Modeling to Implementation of Virtual Sensors in Body Sensor Networks

Body Sensor Networks (BSNs) represent an emerging technology which has received much attention recently due to its enormous potential to enable remote, real-time, continuous and non-invasive monitoring of people in health-care, entertainment, fitness, sport, social interaction. Signal processing for BSNs usually comprises of multiple levels of data abstraction, from raw sensor data to data calculated from processing steps such as feature extraction and classification. This paper presents a multi-layer task model based on the concept of Virtual Sensors to improve architecture modularity and design reusability. Virtual Sensors are abstractions of components of BSN systems that include sensor sampling and processing tasks and provide data upon external requests. The Virtual Sensor model implementation relies on SPINE2, an open source domain-specific framework that is designed to support distributed sensing operations and signal processing for wireless sensor networks and enables code reusability, efficiency, and application interoperability. The proposed model is applied in the context of gait analysis through wearable sensors. A gait analysis system is developed according to a SPINE2-based Virtual Sensor architecture and experimentally evaluated. Obtained results confirm that great effectiveness can be achieved in designing and implementing BSN applications through the Virtual Sensor approach while maintaining high efficiency and accuracy.