Iet Science Measurement & Technology最新文献

Heroin use disorder alters brain function, leading to significant changes in EEG signals. This study proposes a novel approach to identify distinguishing EEG features in heroin addicts and healthy individuals by integrating frequency and non-frequency domain features. The methodology consists of four main stages: (1) Preprocessing, where EEG signals undergo a 50 Hz notch filter and a Butterworth low-pass filter (0.4–45 Hz) to remove noise and artefacts; (2) Feature Extraction, in which both frequency-domain features (power spectral density in alpha, beta, theta, and delta bands) and non-frequency-domain features (approximate entropy, permutation entropy, wavelet entropy, and fractal dimensions of Katz and Petrosian) are computed; (3) Feature Selection, where the Davies–Bouldin index is employed to determine the most discriminative features, independent of the number of clusters; and (4) Analysis and Interpretation, which highlights that approximate entropy in the Cz channel and power spectral density in the upper alpha band in the O1 channel provide the highest discriminative power. Compared to previous studies that primarily rely on frequency-domain features, this approach captures both linear and nonlinear dynamics of EEG signals, leading to improved differentiation between addicted and healthy individuals. While the method demonstrates high accuracy, its sensitivity to EEG preprocessing and the need for larger datasets remain key considerations. The findings suggest that this framework can contribute to more effective addiction diagnosis and monitoring, with potential integration into machine learning-based EEG classification models.

Operation voltage is disconnected before measuring surface charges on gas-insulated transmission line (GIL) insulators (i.e., offline measurement). However, the spontaneous dissipation of surface charges reduces the accuracy. In this paper, the input aperture, the diameter, and the height of sensitive electrodes are optimised by the measurement model of electrostatic probes. The electric field distribution, induced potential, and spatial resolution of probes are analysed. The results indicate that the radius of the input aperture and the height of sensitive electrodes should exceed 3 and 11 mm, respectively, to avoid the shielding effect of the grounded shell. Furthermore, the maximum electric field on the sensitive electrode surface reduces with increased diameters. However, there is a positive correlation with heights. Therefore, the radius of the input aperture is 4 mm. The height and the diameter of sensitive electrodes are 12 mm and 1.6 mm, respectively. For ±320 kV GIL basin insulators, the induced potential distribution measured by probes matches the surface charge, and the spatial resolution is 1091mm². This shows that capacitive electrostatic probes are suitable for the online measurement of surface charges on basin insulators, which provides a reference for the prevention of flashovers resulting from surface charge accumulation on DC GIL insulators.

This study presents a low-cost Near Infrared (NIR)-based optical microfluidic POCT device for in vitro glucose detection on a paper-based microfluidic platform. The device is made using a UV-curable resin that has been widely used in the publishing industry, which is proposed as a low-cost alternative to the expensive SU-8 photoresist normally used in photolithography. A 940 nm NIR LED–photodiode system was incorporated to non-invasively measure glucose levels. Glucose concentrations ranging from 62.5 to 1000 mg/dL were tested using only 3 µL of sample. To evaluate the effect of light intensity, three different LED currents (0.5, 0.75 and 1 mA) were applied. Output voltage ranged from approximately 570 to 700 mV and decreased with increasing glucose concentration, following Beer-Lambert's law. Regression models yielded R² values above 97%, confirming the device's accuracy and its potential for real-time, disposable, paper-based microfluidic glucose monitoring in point-of-care settings.

An improved Nicolson-Ross-Weir (NRW) transmission/reflection method with the algorithm of adapting propagation constant is proposed in this paper for the permittivity measurement of that material with arbitrary thickness. The basic principle of the NRW method is reviewed and verified first, and the proposed improved method and adaptive algorithm are introduced and simulated for verification using the coaxial line measurement system. Both 2.92 and 7.0 mm measurement systems are fabricated to verify our method by measuring four different materials. By comparing the measured results in existing art and those using the proposed method, as well as the uncertainty and error analyses, one can see that our proposed method has the capability of measuring the permittivity of those materials with arbitrary thickness with high accuracy and efficiency.

Traditional high impedance fault (HIF) detection methods face significant technical challenges, including difficulties in feature extraction and limited flexibility in threshold selection, which lead to misjudgment in extreme fault scenarios. Therefore, an HIF detection method for the distribution network is proposed. Firstly, the time-frequency distribution differences of transient signals between the HIF and normal disturbance condition are analysed by Shannon entropy quantization of wavelet packet. On this basis, the transient signal time-frequency waveform block with the lowest similarity is selected as the input sample, and the Dropout in the traditional Vision Transformer (ViT) is replaced by a new regularization method, DropKey, so as to construct a DropKey-Vision Transformer (DVit) classification model, which is suitable for the small-sample scenario of HIF detection for the distribution network. Finally, simulation and experimental test results demonstrate that the proposed method achieves an average accuracy exceeding 99.5% for detecting 10 kΩ HIFs. This represents an improvement of at least 1.5% compared to previous methods and an enhancement of approximately 2% to 7% relative to other techniques. Additionally, the method is applicable to arc grounding, grassland grounding, and pond grounding fault detection, exhibiting high robustness. Results from Grad-CAM visualization further validate the effectiveness and superiority of the proposed method.

To achieve the detectability and isolability of faults in high-voltage circuit breakers, an optimised sensor configuration method based on structural analysis is proposed. First, the main circuit, control circuit, and mechanical operating mechanism of the circuit breaker are analysed to construct the dynamic model of each part, determine the common fault modes of each part, and define fault variables for each fault mode. Then, the fault variables are introduced into the dynamic model of the circuit breaker to form a structured model containing fault information. Next, integrate the residual analysis method based on analytical redundancy relations (ARRs) with considerations for the detectability and isolability of circuit breaker faults. By applying Dulmage–Mendelsohn (DM) decomposition to solve the structural model containing fault information, determine the optimal sensor configuration scheme for circuit breakers. Finally, the effectiveness of the optimised solution is demonstrated both theoretically and experimentally using redundancy analysis methods and experimental data. Since the structural analysis method incorporates fault detectability and isolability (FDI) for circuit breakers, it introduces a novel direction for sensor configuration design and offers more practical physical significance compared to data-driven approaches.

In order to achieve the online analysis of the status of the current carrying structure by using the limited number of external sensors in three-phase integrated gas insulated transmission line (GIL), this paper proposes a data-driven fast calculation method for the temperature distribution with deep-learning reduced-order model, to address the efficiency issue of finite element and other numerical methods in real-time applications. This method combines a proper orthogonal decomposition (POD) with the BP neural network (BPNN) and the deep convolutional neural network (DCNN) based on U-net structure, respectively, so that the accuracy and efficiency of temperature calculation in the solid and fluid domains can be well balanced. A lower-dimensional approximate system of the temperature in solid domains is constructed by POD so that the computational scale can be reduced. BPNN is introduced to map the external sensors data of the GIL to the feature coefficient obtained by POD nonlinearly. The DCNN based on U-net structure is developed to estimate the temperature of the fluid domains by learning the feature of the solid domains, so as to obtain the overall temperature distribution. The results show that the proposed framework can rapidly and accurately predict the thermal state of sliding contact section in three-phase integrated GIL with limited external data, where the maximum relative error is less than 1.0%. The proposed method achieves an acceleration factor of 5.6 × 10³ compared with the numerical simulation software, providing an available option for the real-time visualization and digital twin diagnosis of GIL temperature distribution.

The field of wearable sensors, particularly motion sensors, has experienced noteworthy advancements in recent years. Motion sensors have significantly assisted doctors by gathering real-time data and transmitting information remotely, proving highly beneficial in their practices. The sensor produced in this research is designed to withstand different kinds of mechanical forces such as tension, pressure, bending, twisting and contact. This sensor consists of a combination of silicone rubber, carbon nanotubes and carbon black, an extremely flexible composite material, and its electrodes are arranged in a spiral to provide sufficient strength under varying and strong forces. In order to capture more forces through the sensor while reducing the size of the sensor and lowering production costs, a buffer layer was created on the sensor. After data collection, the forces were separated using machine learning. The sensor was subjected to various tests and showed good characteristics (21-225 sensitivity, 1000 cycles repeatability, 8% FSO non-linearity, 13% FSO hysteresis, etc.). Finally, we attached the manufactured sensor to various parts of the body and were thus able to detect body movements.

Among the most expensive assets in power grids, power transformers are essential for the reliability of the power supply chain and the overall stability of the grid. Due to their permanent connection to the network, this equipment is exposed to all kinds of faults and phenomena, including short-circuit faults and overvoltages caused by lightning and switching. Hence, ongoing monitoring of the transformer's condition is essential to prevent breakdowns and damage to the transformer. Among the different condition monitoring methods, the frequency response analysis (FRA) method is sensitive to the smallest functional changes of the transformer, as it is completely related to the physics and geometry of the transformer. This method stands out as one of the most effective and efficient approaches to transformer monitoring, especially for detecting mechanical faults. However, the FRA method faces an important challenge of interpretation: the correlation between the type of fault that occurred and the way the transformer's function changes is still not well-known, and studies in this field are ongoing. One of the most widely used methods of interpreting frequency response results is the use of numerical indices, coil modelling, transformer function estimation, and artificial intelligence algorithms. This paper introduces these methods, and their advantages and disadvantages are discussed. Then, the most widely used artificial intelligence algorithms in transformer condition monitoring are presented and compared. Finally, future research directions are anticipated.

Regarding the surface discharge phenomenon of the converter valve, this paper proposes a localisation method based on variational modal decomposition (VMD) and multiple signal classification (MUSIC) to achieve accurate positioning of discharge in the current converter valve. First, a solution for determining the number of modes K in VMD and a new threshold for IMF selection are proposed to enhance the noise removal capability in the valve hall. Second, the improved variational modal decomposition (IVMD) and MUSIC algorithms are combined to accurately identify the discharge phenomenon of the converter valve. This approach addresses the issue of inaccurate localization and failure in conventional methods due to strong noise interference in the converter valve hall. Simulation results indicate that when the signal-to-noise ratio (SNR) is higher than -20 dB, the root mean square error (RMSE) is less than 1.6. Localisation experiments were conducted on three types of discharges—needle-plate, cone-plate and sphere-plate discharges—showing an average angular error of less than 1.8°. Both simulation and experimental results demonstrate that the proposed algorithm exhibits higher localisation accuracy and strong applicability when the SNR is lower than -5 dB. Finally, a partial discharge localisation device was developed based on the proposed algorithm. This device utilises an eight-element cross sensor array and applies the IVMD-MUSIC algorithm for localisation, meeting operational and maintenance requirements and providing convenience for inspection personnel.