Sensor Signal Research Articles

Classification of surface microdefects based on cluster analysis of diffuse light reflection sensor signal

Early detection of operational microdefects of protective coatings and corrosion damage is important for reliable and long-term operation of materials and structures. Corrosion damage in the early stages is almost impossible to detect by visual testing methods, given the microscopic size of such damage. Laboratory microscopes are unsuitable for field conditions. Flash meters provide an estimate of surface roughness, but do not provide a surface concentration of microdefects, especially when these defects are heterogeneous or polydisperse. A compact prismatic sensor of the angular characteristic of diffuse light reflection is proposed for the detection of microdefects and express diagnostics of surface quality. To increase the informativity of the sensor signal during measurements, the useful signal was calculated as the difference between the base values of the diffuse reflected signal for a surface without defects and the measured values for a surface with defects. The dependence of the signal distribution on the photoline on the geometry of the sensor was taken into account. The article considers the problem of classification by the average size of surface microdefects collected in submillimeter clusters, randomly placed on the surface of materials. The use of numerical methods for solving inverse problems requires large computing power and time, which is not always acceptable for field measurements. The dependence of solution stability on signal measurement errors and random locations of clusters of microdefects on the illuminated area of the investigated surface is important. An effective set of informative signal features is selected. The study explored the possibility of using alternative sets of informative parameters derived from the base set. It was proposed to determine the size of the defects based on the smoothness criteria of the contour signal of the diffuse light reflection sensor. To evaluate the characteristics of the spatial configuration of defects, the positions of the peaks of the signal distribution along the photoline, determined with subpixel accuracy, were used. The classification is made for the selected information features of the signals of the diffuse light reflection sensor. In particular, the number of local signal extrema together with the subpixel position of the global signal maximum produces well-separated clusters, even for randomly located defect aggregations.

Enhancing Recovery of Structural Health Monitoring Data Using CNN Combined with GRU

Structural health monitoring (SHM) plays a crucial role in ensuring the safety of infrastructure in general, especially critical infrastructure such as bridges. SHM systems allow the real-time monitoring of structural conditions and early detection of abnormalities. This enables managers to make accurate decisions during the operation of the infrastructure. However, for various reasons, data from SHM systems may be interrupted or faulty, leading to serious consequences. This study proposes using a Convolutional Neural Network (CNN) combined with Gated Recurrent Units (GRUs) to recover lost data from accelerometer sensors in SHM systems. CNNs are adept at capturing spatial patterns in data, making them highly effective for recognizing localized features in sensor signals. At the same time, GRUs are designed to model sequential dependencies over time, making the combined architecture particularly suited for time-series data. A dataset collected from a real bridge structure will be used to validate the proposed method. Different cases of data loss are considered to demonstrate the feasibility and potential of the CNN-GRU approach. The results show that the CNN-GRU hybrid network effectively recovers data in both single-channel and multi-channel data loss scenarios.

Rubber-based piezoresistive sensing: a new approach based on hydrodynamic flow of material deformation describing nonlinear signals in stretchable sensors

AbstractThe nonlinearity of piezoresistive response is critical in developing strain sensors, various self-monitoring applications and wearable electronics based on filled rubbers. This parameter could change dramatically when scaling up from small-size prototypes to full-scale production. The present work focuses on the nonlinear signals in stretchable rubber-based sensors, their origin and dependence on size of samples. Thus, a set of rectangular, piezoresistive samples differing in width was prepared from natural rubber reinforced with carbon black filler. Their electric resistance was tested under planar strain/recovery conditions at 25 and 50% strain amplitudes. It was found that piezoresistance and the related nonlinear phenomena significantly depended on the size of the samples. For the first time, hydrodynamic flow of deformed material was used to explain the nonlinearities of the piezoresistive signal. The trajectory, velocity, and magnitude of this flow were accounted for by a newly developed empirical equation describing the evolution of local resistivity under the strain/recovery process.

Research on intelligent cutting control technology of transverse moving machine for large cross-section roadheader

With the continuous development of the coal mine industry in our country, the cross-section of the roadway is getting larger and larger. The widely used EBZ series roadheader cannot complete the roadway cutting operation with a large cross-section at one time, so it usually needs to move the equipment left and right to complete all its cutting tasks. This paper studies the intelligent control technology of large cross-section roadway cutting, which mainly solves the positioning and attitude determination technology of roadheader in the large cross-section, and offers an accurate shape-cutting control technology and a path planning technology using a lateral moving machine. This paper proposes to combine the signals of laser sensor and fiber-optic inertial navigation system to form a combined positioning and attitude determination system, so as to make use of the advantages and avoid the disadvantages of their respective systems, and improve the accuracy of roadheader positioning and attitude determination. A swing kinematics model of the cutting arm is established, based on which the precise motion control of the first roadway section cutting and the second roadway section cutting can be realized. A key parameter calculation model of roadheader lateral moving machine is put forward to calculate the operation parameters for roadheader lateral moving machine path planning. Finally, the experimental research on intelligent cutting control technology of large cross-section roadheader is carried out in 12307 intelligent heading face of Wangjialing Coal Mine. The results show that the positioning error of the combined positioning and attitude determination system proposed in this paper is between 50-90 mm in the X-axis direction and between 25-45 mm in the Y-axis direction. The average attitude measurement error of heading angle is between 0.5 and 1.5°. This makes the technology appropriate for engineering use. The studied cutting control system achieves a cutting error of 25-50 mm for the first section, and a cutting error of 40-90 mm for the complete section. The complete section cutting error meets the ±100 mm cutting error requirement required in engineering. The intelligent cutting control technology of the lateral moving machine of the large-section tunnel boring machine studied in this paper can realize accurate cutting of the large-section tunnel of 5.6 meters. The research results of this paper provide a reference for the intelligent construction of heading faces.

Ultra Enhancement of the Resolution of the Divergence Beam-Based SPR Sensor Using a Wavelet Filter

Surface Plasmon Resonance (SPR) sensors are highly sensitive to refractive index variations, making them ideal for biosensing and chemical detection applications. However, their performance can be constrained by noise, resolution limits, and signal distortions, particularly in divergence beam-based configurations. This research presents an ultra-enhancement of the performance of divergence beam-based SPR sensors by employing advanced wavelet filtering techniques. Wavelets, with their multi-resolution analysis capability, are applied to denoise and refine the sensor signal, significantly improving key performance metrics, including resolution, mean square error (MSE), root mean square error (RMSE), signal-to-noise ratio (SNR), sensitivity matrix (SM), and combined sensitivity factor (CSF).The wavelet filter effectively decomposes the SPR signal into distinct frequency bands, separating noise while retaining high-frequency components required for accurate sensing. This results in a significant reduction in MSE and RMSE of 30 and 92.26%, respectively, while simultaneously improving SNR by 54.08%, maintaining signal quality. The wavelet filter significantly improved the resolution of the SPR sensor by 99.95% (1.3772×10-7 RIU), allowing for more precise detection of refractive index changes and boosting diverge beam-based SPR sensor performance. In addition, a sensitivity matrix (SM) and combined sensitivity factor (CSF) were improved by 42 and 41.94%, respectively, allowing for a more thorough evaluation of the sensor's performance across multiple operational parameters. Simulation and experimentation show that wavelet filtering surpasses standard filtering approaches in terms of noise suppression and signal clarity. This technology has considerable potential for improving the accuracy and reliability of SPR sensors in high-sensitivity applications

Fluorescence sensor based on Methionine-Modified silver nanoparticles located on Fe-BTC metal–organic framework (Meth-AgNPs@Fe-BTC) for trace detection of fenitrothion pesticide in aqueous samples

This research introduces a new “turn-on mode” fluorescence sensor for the detection of fenitrothion (FNT) pesticide in various samples. The sensor is constructed using a porous metal–organic framework (Fe-BTC) as a template to locate silver nanoparticles (AgNPs) and methionine amino acid (Meth). Methionine acts as a bridge, facilitating the interaction between FNT and AgNPs, which subsequently results in the release of AgNPs from the composite structure. The physicochemical properties of the synthesized Meth-AgNPs@Fe-BTC composite were analyzed by Fourier-transform infrared spectroscopy (FT-IR), Brunauer-Emmett-Teller (BET), X-ray diffraction (XRD), Field emission scanning electron microscopy (FESEM), Energy-dispersive X-ray spectroscopy (EDAX), Transmission electron microscopy (TEM), and elemental mapping (MAP) analysis. The sensing system is based on tracking the fluorescence of the synthetic composite in such a way that the intensity of the fluorescence of the composite increases in the presence of different concentrations of fenitrothion (FNT). The effective parameters on the sensor signal, including composite dosage, pH, sonication and reaction time were investigated and optimized. The calibration graph, under optimal conditions, exhibited linearity in the concentration range of 2–95 nM for FNT, with a limit of detection of 1.9 nM. The suggested sensor was successfully validated by analyzing FNT in several real water samples and fruit juices. This research presents a significant technical achievement in the development of a fluorescence sensor for the detection of FNT, offering a sensitive and reliable method for environmental monitoring and public health preservation.

A Spinning Magnetometer Based Approach to Solving the Roll Attitude of Aerial Vehicle

Abstract The accurate attitude estimation is the foundation of autonomous flight control for unmanned aerial vehicle (UAV). However, the existing inertia/magnetic field-based roll attitude estimation presents challenges, including the significant computational burden and high cumulative error. To address these issues, this paper proposes a novel roll attitude estimation approach for UAV based on spinning single-axis magnetometer. Firstly, a distinctive rotating measuring mechanism is structured by incorporating the encoder and geomagnetic sensor. Then, the rotating angle of the sensor in relation to the UAV is ascertained with the specific output signal of the geomagnetic sensor with the rotating mechanism. Based on the above, a roll attitude estimation model to be established with the relative angle relationship. Finally, comparative simulations are conducted to demonstrate the feasibility of the presented method. The suggested design method provides a rapid and cost-effective way for roll attitude estimation.

An ASIC-based system-in-package MEMS gas sensor with impedance spectroscopy readout and AI-enabled identification capabilities

The main advantages of micromachined metal oxide semiconductor (MOX) gas sensors, in particular their low cost, high sensitivity, miniaturization and low power consumption, are compromised by lack of selectivity and long-term sensor signal drifts, which often hinder their application in real-world sensing applications. To increase the sensor data dimensionality, temperature cycled operation (TCO) and impedance spectroscopy (IS) readout was proposed, enabling the acquisition of large datasets from a single sensor, as suitable to be exploited by machine learning techniques. Furthermore, recent studies have shown how the IS readout minimizes the issues represented by the sensor long-term signal drift.The aim of this work is the development of a smart system-in-package, integrating a state-of-the-art micromachined MOX gas sensor based on a wafer-level nanostructured sensing layer with a novel versatile sensor interface chip, capable of high performance IS readout. We propose a miniaturized sensing device, which is a small, lightweight low power consumption chemical sensing micro-system with fast and accurate classification capabilities, suitable to be deployed in long multi-node sensing cables or on unmanned aerial vehicles. A software suite comprising a sensor training wizard was also developed, enabling users with no technical background to train the sensing system.

Sensor-Based Multifaceted Feature Extraction and Ensemble Elastic Net Approach for Assessing Fall Risk in Community-Dwelling Older Adults.

Accurate identification of community-dwelling older adults at high fall risk can facilitate timely intervention and significantly reduce fall incidents. Analyzing gait and balance capabilities via feature extraction and modeling through sensor-based motion data has emerged as a viable approach for fall risk assessment. However, the existing approaches for extracting key features related to fall risk lack inclusiveness, with limited consideration of the non-linear characteristics of sensor signals, such as signal complexity, self-similarity, and local stability. In this study, we developed a multifaceted feature extraction scheme employing diverse feature types, including demographic, descriptive statistical, non-linear, spatiotemporal and spectral features, derived from three-axis accelerometers and gyroscope data. This study is the first attempt to investigate non-linear features related to fall risk in multi-task scenarios from a dynamic system perspective. Based on the extracted multifaceted features, we propose an ensemble elastic net (E-E-N) approach for handling imbalanced data and offering high model interpretability. The E-E-N utilizes bootstrap sampling to construct base classifiers and employs a weighting mechanism to aggregate the base classifiers. We conducted a set of validation experiments using real-world data for comprehensive comparative analysis. The results demonstrate that the E-E-N approach exhibits superior predictive performance on fall risk classification. Our proposed approach offers a cost-effective tool for accurately assessing fall risk and alleviating the burden of continuous health monitoring in the long term.

Fabric‐Reinforced Functional Insoles with Superior Durability and Antifracture Properties for Energy Harvesting and AI‐Empowered Motion Monitoring

AbstractFunctional triboelectric insoles hold promise for advancing self‐powered wearable technologies. However, their durability is compromised by continuous compressive forces and friction, leading to surface abrasion and material fracturing. To address these challenges, an innovative fabric‐reinforced structure combined with a dual‐L backrest design is developed that enhances anti‐fracture capabilities and electric outputs while enabling AI‐empowered motion monitoring. Polydimethylsiloxane (PDMS) is used as the negative triboelectric material with a dual‐L backrest design, while insulated copper wire (icuW) serves as the positive triboelectric material with an annular structure design. These components are intricately nested to enable a multilayered friction pairing. The fabric‐reinforced structure demonstrates excellent compressive rebound resilience, withstanding forces of at least 1000 N. The functional insole, featuring a fabric‐reinforced dual‐L backrest structure (FRdL‐insole), efficiently harvests biomechanical energy with a peak power of 8214 µW and maintains highly consistent performance after 10 washing cycles and 60 000 durability tests. It can power portable electronic devices such as digital watches, calculators, hygrometers, and LEDs. Enhanced with machine learning algorithms, the FRdL‐insole processes sensor signals to monitor human movements, accurately identifying seven distinct motions. This positions the insole as a smart, real‐time, self‐powered tool for activity recognition, showcasing its potential in intelligent wearable technology.

AE-IRMamba: Low-Complexity Inverted Residual Mamba for Identification of Piezoelectric Ceramic and Optical Fiber Acoustic Emission Sensors Signals

AE-IRMamba: Low-Complexity Inverted Residual Mamba for Identification of Piezoelectric Ceramic and Optical Fiber Acoustic Emission Sensors Signals

A New Iterative Algorithm for Magnetic Motion Tracking.

Motion analysis is of great interest to a variety of applications, such as virtual and augmented reality and medical diagnostics. Hand movement tracking systems, in particular, are used as a human-machine interface. In most cases, these systems are based on optical or acceleration/angular speed sensors. These technologies are already well researched and used in commercial systems. In special applications, it can be advantageous to use magnetic sensors to supplement an existing system or even replace the existing sensors. The core of a motion tracking system is a localization unit. The relatively complex localization algorithms present a problem in magnetic systems, leading to a relatively large computational complexity. In this paper, a new approach for pose estimation of a kinematic chain is presented. The new algorithm is based on spatially rotating magnetic dipole sources. A spatial feature is extracted from the sensor signal, the dipole direction in which the maximum magnitude value is detected at the sensor. This is introduced as the "maximum vector". A relationship between this feature, the location vector (pointing from the magnetic source to the sensor position) and the sensor orientation is derived and subsequently exploited. By modelling the hand as a kinematic chain, the posture of the chain can be described in two ways: the knowledge about the magnetic correlations and the structure of the kinematic chain. Both are bundled in an iterative algorithm with very low complexity. The algorithm was implemented in a real-time framework and evaluated in a simulation and first laboratory tests. In tests without movement, it could be shown that there was no significant deviation between the simulated and estimated poses. In tests with periodic movements, an error in the range of 1° was found. Of particular interest here is the required computing power. This was evaluated in terms of the required computing operations and the required computing time. Initial analyses have shown that a computing time of 3 μs per joint is required on a personal computer. Lastly, the first laboratory tests basically prove the functionality of the proposed methodology.

Continuous monitoring of heart failure biomarker using a sensitive graphene-based electrical sensor

Abstract Heart Failure (HF) is a complex condition that imposes significant burdens on both patients and healthcare systems. The economic impact of HF therapy and management is substantial, underscoring the urgency for proactive measures to mitigate its development. Regular assessments play a pivotal role in identifying individuals at risk and initiating timely interventions to prevent or delay HF progression [1, 2]. NT-proBNP, an amine-terminated form of brain natriuretic peptide, emerges as a promising biomarker in the realm of chronic HF diagnosis, prognosis, and risk assessment [3]. Its utility lies in providing valuable insights into disease progression and prognosis. In recent years, biosensors based on Field Effect Transistors with Graphene (Gr) gate (GFET) have garnered attention for their rapid response, sensitivity, and on-chip integration [4]. Here, we present a sensitive miniaturized GFET sensor for continuous NT-proBNP measurement. The sensor is an array of GFET devices with sensing gates of a monolayer Gr decorated with aptamers as catch probes. Averaging over similar devices in an array approach ensures measurement consistency throughout the experiment. Moreover, the significant challenge posed by the Debye length in FET biosensors is effectively addressed through the functionalization with short folding aptamers [4]. The sensor is built based on a chip comprising an array of 28 individual micro GFETs (Fig. 1a). For the functionalization of Gr gates, we utilized 1-pyrene butyric acid N-hydroxysuccinimide ester as a linker between the aptamer probes and the Gr surface. Subsequently, the Gr surface was incubated with an amine-terminated aptamer solution (Fig. 1c). A schematic diagram of the sensory chip having a droplet of the sample is shown in Fig. 1b. To quantify NT-proBNP levels an 8µL droplet of the buffer solution containing a specific NT-proBNP concentration was deposited onto the sensor. The drain current (IDS) served as the sensor signal for each target concentration (Figs. 2a and 2b). The sensor signal decreased with increasing target concentration, while negligible response was observed when injecting blank buffers. This phenomenon can be attributed to the folding of aptamers upon binding to their target. Also, continuous measurement was conducted at a fixed gate voltage of 0.9V, recording the drain current while the target concentration gradually increased in the sample droplet. This process revealed a decrease in sensor current as we expected (Fig. 2c). The sensor exhibits remarkably low detection limits of 50 pg/mL and a linear range from 100 pg/mL to 10 ng/mL, which are highly conducive for early HF diagnosis and aligning well with reported NT-proBNP levels in patient biofluid [5]. The proposed sensor holds promise as a wearable device for continuous monitoring of high-risk patients. Additionally, this technology shows potential as a detection tool for acute heart failure in hospitalized patients.Sensor viewSensor responses

Development of an Algorithm for Determining the Class of an Obstacle in the Process of its Interaction with the Manipulator

The article discusses the problem of determining the nature of an obstacle using a force sensor for terrain mapping. Three classes of objects have been introduced: the first class is "static" obstacles, the second class is "moving" obstacles, the third class is "unstable" obstacles. To conduct experiments, a simulation system was built in the Gazebo modeling environment, including a manipulator with the ability to horizontally move a probe with a force sensor and a set of obstacles corresponding to three classes of obstacles. All experiments were carried out using the ROS framework, the Gazebo simulator and the Movelt manipulation software. Processing of signals from the force sensor showed that the sig­nals for the three classes of obstacles are different, which makes it possible to identify the class based on sensor readings. Based on the analysis of data received from the sensor, rules were formed that make it possible to determine the class of an obstacle during its interaction with the manipulator in real time. To determine the obstacle class, the values of the force sensor signals are considered, smoothed by the moving average method. According to the rules discussed in the article, an algorithm was developed to determine the class of obstacle, which was implemented in the Gazebo modeling environment. Testing of the algorithm showed the stability of its operation under the given experimental conditions. Successful determination of the obstacle class will allow one not only to construct a map of the area, but also to prevent damage to the device when exposed to a "static" object with excessive force. The presented approach can be expanded to narrower classes of obstacles that characterize objects that can be encountered by a mobile device when mapping a forest area.

Research for an enhanced fault-tolerant solution against the current sensor fault types in induction motor drives

Introduction. Recently, three-phase induction motor drives have been widely used in industrial applications; however, the feedback signal failures of current sensors can seriously degrade the operation performance of the entire drive system. Therefore, the motor drives require a proper solution to prevent current sensor faults and improve the reliability of the motor drive systems. The novelty of the proposed research includes integrating the current sensor fault-tolerant control (FTC) function according to enhanced technique into the field-oriented control loop for speed control of the motor drive system. Purpose. This research proposes a hybrid method involving a third difference operator and signal comparison algorithm to diagnose various types of current sensor faults as a positive solution to enhance the stability of the induction motor drive system. Methods. A hybrid method involving a third difference operator for the measured speed signals and a comparison algorithm between measured and estimated current signals are proposed to diagnose the current sensors’ health status in the fault-tolerant process. After determining the faulty sensor, the estimated current signals based on the Luenberger observer are used immediately to replace the defective sensor signal. Results. The current sensor is simulated with various failure types, from standard to rare failures, to evaluate the performance of the FTC method implemented in the MATLAB/Simulink environment. Simultaneously, a fault flag corresponding to a defective sensor should be presented as an indicator to execute the repair process for faulty sensors at the proper time. Practical value. Positive results have proven the feasibility and effectiveness of the proposed FTC integrated into the speed controller to improve reliability and ensure the stable operation of the induction motor drive system even under current sensor fault conditions. References 29, tables 3, figures 10.

On the Origin of Signal and Bandwidth of Converse Magnetoelectric Magnetic Field Sensors

AbstractConverse magnetoelectric sensors enable the detection of low‐frequency and low‐amplitude magnetic fields over a bandwidth of several kilohertz by combining the electrical excitation of a magnetoelectric resonator via a piezoelectric layer with an inductive readout. Here, a comprehensive sensor model is presented to further foster the development of this promising sensor concept. The model relates the output signal to the device characteristics, taking into account the magnetoelastic and electromechanical properties, the resonator geometry, and operating conditions. The sensor system is thoroughly experimentally analyzed to validate the model. Based on the analysis, the sensor concept is explained in detail, including the origin of its loss and bandwidth and their connection with the magneto‐mechanical loss in the magnetostrictive layer. Significant advances have been made in the comprehensive understanding of converse magnetoelectric sensors, providing a solid basis for future improvements in magnetoelectric sensor systems.

Time–frequency domain machine learning for detection of epilepsy using wearable EEG sensor signals recorded during physical activities

Epilepsy is a neurological ailment in which there is a disturbance in the nerve cell activity of the brain, causing recurrent seizures. The electroencephalogram (EEG) signal is widely used as a diagnostic modality to detect epilepsy ailment. The automated detection of epilepsy using wearable EEG sensor data recorded during various physical activities is interesting for continuous monitoring of brain health. This paper proposes a time–frequency (TF) domain machine learning (ML) approach for the automated detection of epilepsy using wearable sensor-based EEG signals. The Gaussian window-based Stockwell transform (GWST) is employed to evaluate the TF matrix from the EEG signal. The features such as the L1-norm and the Shannon entropy are extracted from the TF matrix of the EEG signal. The ML and deep learning (DL) models are employed to detect epilepsy using the TF domain features of EEG signals. The publicly available database containing wearable sensor-based EEG signals recorded from the subjects while performing different physical activities is used to evaluate the performance of the proposed approach. The results show that the random forest (RF) classifier coupled with GWST domain features of EEG signals has obtained an overall accuracy value of 90.74% for detecting epilepsy with hold-out validation using the EEG signals from different physical activity cases. For the 10-fold cross-validation (CV) case, the GWST domain features of EEG signal and multi-layer long short-term memory (LSTM) classifier have produced the average accuracy value of 74.44%. For jogging, running, and idle sitting activities, the GWST-based TF domain entropy features coupled with the multi-layer LSTM model have obtained accuracy values of 82.72%, 82.41%, and 87.30%, respectively. The proposed approach has achieved higher classification accuracy than existing methods to detect epilepsy using wearable sensor-based EEG signals using a 10-fold CV strategy. The suggested approach is compared with existing methods to classify seizure and seizure-free classes using resting-state EEG signals.

Potentiometric Electronic Tongue for the Evaluation of Multiple-Unit Pellet Sprinkle Formulations of Rosuvastatin Calcium.

Sprinkle formulations represent an interesting genre of medicinal products. A frequent problem, however, is the need to mask the unpleasant taste of these drug substances. In the present work, we propose the use of a novel sensor array based on solid-state ion-selective electrodes to evaluate the taste-masking efficiency of rosuvastatin (ROS) sprinkle formulations. Eight Multiple Unit Pellet Systems (MUPSs) were analyzed at two different doses (API_50) and (API_10), as well as pure Active Pharmaceutical Ingredient (API) as a bitter standard. Calcium phosphate-based starter pellets were coated with the mixture containing rosuvastatin. Some of them were additionally coated with hydroxypropyl methylcellulose, which was intended to separate the bitter substance and prevent it from coming into contact with the taste buds. The sensor array consisted of 16 prepared sensors with a polymer membrane that had a different selectivity towards rosuvastatin calcium. The main analytical parameters (sensitivity, selectivity, response time, pH dependence of potential, drift of potential, lifetime) of the constructed ion-selective electrodes sensitive for rosuvastatin were determined. The signals from the sensors array recorded during the experiments were processed using Principal Component Analysis (PCA). The results obtained, i.e., the chemical images of the pharmaceutical samples, indicated that the electronic tongue composed of the developed solid-state electrodes provided respective attributes as sensor signals, enabling both of various kinds of ROS pellets to be distinguished and their similarity to ROS bitterness standards to be tested.

EMA fault diagnosis method based on multi-information fusion of GRU and improved attention mechanism

Aiming at the problems of time series feature loss and fault information loss in traditional electromechanical servo system (EMA) fault diagnosis methods based on machine learning and deep learning, a multi-information fusion EMA fault diagnosis method based on gated recurrent unit (GRU) and improved attention mechanism is proposed. Firstly, the collected different sensor signals are divided into different channels, and the time series features of each channel signal are extracted by GRU. Then, the self-attention mechanism is introduced to further distinguish the important relationship between different time points of the signal. The multi-channel attention mechanism is further introduced to adaptively fuse the features of different channels, and finally the fault diagnosis is realized through the classifier. The experimental results based on the test bench data set show that: compared with the single sensor model, the diagnosis rate is improved 10%; compared with the model without the introduction of the attention mechanism, the diagnosis rate is improved 5.2%; compared with the classic machine learning, deep learning and the improved algorithm based on deep learning in the past two years, the diagnosis rate of the diagnosis model designed in this paper is 98.5%above, and the diagnosis effect is the best.

Multi-Source Information-Based Bearing Fault Diagnosis Using Multi-Branch Selective Fusion Deep Residual Network.

Rolling bearing is the core component of industrial machines, but it is difficult for common single signal source-based fault diagnosis methods to ensure reliable results since sensor signals are vulnerable to the pollution of background noises and the attenuation of transmitted information. Recently, multi-source information-based fault diagnosis methods have become popular, but the information redundancy between multiple signals is a tough problem that will negatively impact the representational capacity of deep learning algorithms and the precision of fault diagnosis methods. Besides that, the characteristics of various signals are actually different, but this problem was usually omitted by researchers, and it has potential to further improve the diagnosing performance by adaptively adjusting the feature extraction process for every input signal source. Aimed at solving the above problems, a novel model for bearing fault diagnosis called multi-branch selective fusion deep residual network is proposed in this paper. The model adopts a multi-branch structure design to enable every input signal source to have a unique feature processing channel, avoiding the information of multiple signal sources blindly coupled by convolution kernels. And in each branch, different convolution kernel sizes are assigned according to the characteristics of every input signal, fully digging the precious fault components on respective information sources. Lastly, the dropout technique is used to randomly throw out some activated neurons, alleviating the redundancy and enhancing the quality of the multiscale features extracted from different signals. The proposed method was experimentally compared with other intelligent methods on two authoritative public bearing datasets, and the experimental results prove the feasibility and superiority of the proposed model.