Iet Science Measurement & Technology最新文献

The bulk conductivity of the epoxy spacer exhibits temperature-dependent behaviour, leading to unstable electric field (E-field) distributions under varying temperature gradients. To collaboratively regulate the steady-state and transient E-field distributions of the spacer under temperature gradients, a ε/σ-multi-dimensional functionally graded materials (MFGM) spacer was proposed in this paper. With the combination of surface conductivity and bulk permittivity gradient design, the ε/σ-MFGM spacer demonstrates superior E-field regulation under steady-state, AC/DC superposition and polarity reversal voltages. Additionally, the ε/σ-MFGM spacer maintains higher E-field stability with temperature variation than the uniform spacer. Based on the gradient spraying and settlement pouring technology, the fabrication of the scaled basin-type ε/σ-MFGM spacer was achieved. Experimental results indicate that the flashover voltage of the ε/σ-MFGM spacer is significantly enhanced compared to the uniform spacer, with an improvement ranging from 2.1% to 21.1% under various temperature gradients and voltage conditions.

The bulk conductivity of the epoxy spacer exhibits temperature-dependent behaviour, leading to unstable electric field (E-field) distributions under varying temperature gradients. To collaboratively regulate the steady-state and transient E-field distributions of the spacer under temperature gradients, a ε/σ-multi-dimensional functionally graded materials (MFGM) spacer was proposed in this paper. With the combination of surface conductivity and bulk permittivity gradient design, the ε/σ-MFGM spacer demonstrates superior E-field regulation under steady-state, AC/DC superposition and polarity reversal voltages. Additionally, the ε/σ-MFGM spacer maintains higher E-field stability with temperature variation than the uniform spacer. Based on the gradient spraying and settlement pouring technology, the fabrication of the scaled basin-type ε/σ-MFGM spacer was achieved. Experimental results indicate that the flashover voltage of the ε/σ-MFGM spacer is significantly enhanced compared to the uniform spacer, with an improvement ranging from 2.1% to 21.1% under various temperature gradients and voltage conditions.

Accurate wet gas flow measurement is essential for production optimisation, custody transfer, and regulatory compliance in the energy and chemical industries. Conventional ultrasonic flow meters often overestimate gas flow rates due to liquid entrainment, while microwave sensors alone struggle with phase discrimination under dynamic conditions. This study introduces a hybrid metering system, Ultrasonic Flow Meters and Microwave Sensing Measurement of Wet Gas (USMMW), that integrates transit-time ultrasonic flow measurement with microwave dielectric sensing to correct over-reading errors. Experimental data were collected from a controlled multiphase flow loop using a 2-inch pipeline equipped with an ultrasonic meter and a 2.7 GHz microwave sensor. A data-driven over-reading correction model (OR) was developed using detected liquid volume fraction (LVF) and eight dimensionless parameters derived via the Buckingham Pi theorem. Multiple regression and machine learning techniques, including multilinear regression (MLR) and random forest regression (RFR), were applied to optimise model performance. Validation results showed that the USMMW system achieved corrected gas flow rates with an average relative absolute error (RAE) of 3.02%, outperforming conventional differential pressure models. The findings demonstrate that USMMW offers a robust, non-intrusive solution for real-time wet gas metering under mist and stratified flow regimes, with potential for scalable industrial deployment.

The utilization of fault current limiters (FCLs) results in a decrease in short-circuit levels. However, as FCLs are integrated within the fault loop circuit, they may exert a significant influence on the transient recovery voltage and rate of rise of recovery voltage (RRRV) across the circuit breaker (CB) and its interruption duty. This paper investigates analytical formulas for the RRRV in the presence of a resistive FCL (RFCL) in which an improved model incorporating two stray capacitors is proposed to better represent the distributed nature of the RFCL. Furthermore, both terminal faults and short-line faults are analysed for various topologies, including downstream and upstream connections to the CB, and considering different limiting factors. Computational analysis is conducted on a 20-kV distribution feeder under a three-phase fault using the EMTP-RV environment to assess the most severe transient conditions impacting the CB. The discussed findings present crucial considerations for the appropriate design and implementation of RFCLs to ensure reliable operation of the CB and successful interruption of fault currents.

The utilization of fault current limiters (FCLs) results in a decrease in short-circuit levels. However, as FCLs are integrated within the fault loop circuit, they may exert a significant influence on the transient recovery voltage and rate of rise of recovery voltage (RRRV) across the circuit breaker (CB) and its interruption duty. This paper investigates analytical formulas for the RRRV in the presence of a resistive FCL (RFCL) in which an improved model incorporating two stray capacitors is proposed to better represent the distributed nature of the RFCL. Furthermore, both terminal faults and short-line faults are analysed for various topologies, including downstream and upstream connections to the CB, and considering different limiting factors. Computational analysis is conducted on a 20-kV distribution feeder under a three-phase fault using the EMTP-RV environment to assess the most severe transient conditions impacting the CB. The discussed findings present crucial considerations for the appropriate design and implementation of RFCLs to ensure reliable operation of the CB and successful interruption of fault currents.

Accurate wet gas flow measurement is essential for production optimisation, custody transfer, and regulatory compliance in the energy and chemical industries. Conventional ultrasonic flow meters often overestimate gas flow rates due to liquid entrainment, while microwave sensors alone struggle with phase discrimination under dynamic conditions. This study introduces a hybrid metering system, Ultrasonic Flow Meters and Microwave Sensing Measurement of Wet Gas (USMMW), that integrates transit-time ultrasonic flow measurement with microwave dielectric sensing to correct over-reading errors. Experimental data were collected from a controlled multiphase flow loop using a 2-inch pipeline equipped with an ultrasonic meter and a 2.7 GHz microwave sensor. A data-driven over-reading correction model (OR) was developed using detected liquid volume fraction (LVF) and eight dimensionless parameters derived via the Buckingham Pi theorem. Multiple regression and machine learning techniques, including multilinear regression (MLR) and random forest regression (RFR), were applied to optimise model performance. Validation results showed that the USMMW system achieved corrected gas flow rates with an average relative absolute error (RAE) of 3.02%, outperforming conventional differential pressure models. The findings demonstrate that USMMW offers a robust, non-intrusive solution for real-time wet gas metering under mist and stratified flow regimes, with potential for scalable industrial deployment.

Characterisation of magnetic core materials using a ring sample may lead to inaccuracies because of field nonuniformity along the radial direction in the sample. This study developed an analytical formulation for accurately identifying the static hysteretic BH curves of the core material from measured current–flux curves using a ring sample. Computational tests demonstrated that the proposed method can reconstruct hysteretic properties more accurately than conventional approximation methods based on averaged fields. The inaccuracies of conventional characterisation methods have also been discussed theoretically.

Machine Learning (ML) approaches are getting increasingly common in numerical modelling, thanks to their promptness once trained and to their capability of self-extracting numerical models from behavioural examples. While classical numerical models have been developed for many years and presently represent a robust and well-known solution to handle complex electromagnetic (EM) models, they still suffer from some drawbacks—the need to create a discrete version of the model from basic laws and the need for massive computational power in more complex cases. Although commercial software tools have been developed that are able to handle semi-automatically these issues, in some applications, such as optimised design and iterative resolution of inverse problems, the cited issue may still represent a relevant limitation. We aim to balance the drawbacks and the advantages of both approaches, by investigating the performance of representative methods in each class on a simple yet relevant electrostatic problem described by Elliptic Partial Differential Equations (E-PDE).

Machine Learning (ML) approaches are getting increasingly common in numerical modelling, thanks to their promptness once trained and to their capability of self-extracting numerical models from behavioural examples. While classical numerical models have been developed for many years and presently represent a robust and well-known solution to handle complex electromagnetic (EM) models, they still suffer from some drawbacks—the need to create a discrete version of the model from basic laws and the need for massive computational power in more complex cases. Although commercial software tools have been developed that are able to handle semi-automatically these issues, in some applications, such as optimised design and iterative resolution of inverse problems, the cited issue may still represent a relevant limitation. We aim to balance the drawbacks and the advantages of both approaches, by investigating the performance of representative methods in each class on a simple yet relevant electrostatic problem described by Elliptic Partial Differential Equations (E-PDE).

Machine Learning (ML) approaches are getting increasingly common in numerical modelling, thanks to their promptness once trained and to their capability of self-extracting numerical models from behavioural examples. While classical numerical models have been developed for many years and presently represent a robust and well-known solution to handle complex electromagnetic (EM) models, they still suffer from some drawbacks—the need to create a discrete version of the model from basic laws and the need for massive computational power in more complex cases. Although commercial software tools have been developed that are able to handle semi-automatically these issues, in some applications, such as optimised design and iterative resolution of inverse problems, the cited issue may still represent a relevant limitation. We aim to balance the drawbacks and the advantages of both approaches, by investigating the performance of representative methods in each class on a simple yet relevant electrostatic problem described by Elliptic Partial Differential Equations (E-PDE).