The evaluation and selection of state of transformers based on their mineral oil data are quite difficult. In this study, analytical hierarchy process (AHP) and technique for order preference by similarity to ideal solutions (TOPSISs) are used as the multi-criteria decision support methodologies to provide a simple approach to rank transformers based on their mineral oil data and to help decision makers to identify the transformer, which is in the most critical state. AHP is exercised for both priority weight calculation and ranking of transformers. The AHP method is used for weight calculation, and then transformers are ranked using TOPSIS. Here, data of 69 in-service and 12 failed transformers in Tamil Nadu state in India is collected for assessment. From the results, transformers with good insulation condition are ranked at top, while the transformers are given lower ranks if they are found to have bad health. It is proved that decisions yielded by both methods are in agreement with reality.

Optimised quantity of nano-aluminium (Al) filler mixed with epoxy resin using percolation threshold criteria was adapted to fabricate nanocomposite with desired space-charge and charge-trap properties. Surface potential variation due to charge deposition on to the insulating material under transient voltages was investigated, and it showed the improved decay rate with reduced trap depth for the nanocomposites with nanofillers addition in the range between 0.5 and 5 wt%. Threshold electric field calculated based on space-charge density variation by pulsed electroacoustic measurement was found to be between 7–9 kV/mm and maximum electric field calculated through polarity reversal test is about 9 kV/mm, for 5 wt% (for the optimised value of nanofiller) epoxy nano-Al composites.

A microelectromechanical systems piezoelectric transducer capable of measuring static accelerations and acoustic vibrations has been designed and proposed. The transducer is composed of a circular plate on a pillar, which is fixed at the centre and free at the rim so that it resembles a flat cap mushroom. An annular piezoelectric layer has been employed to convert the vibration-induced stress of the plate to a potential difference. The proposed structure has been compared with diaphragm piezoelectric transducers and hydrophones and its superior performance has been verified. Analytical models for both static and dynamic accelerations have been developed, discussed and the output voltage has been formulated, which is in a very good agreement with the finite element analysis. Results show that the maximum sensitivity is achieved when there is an annular piezoelectric layer on the plate around the pillar perimeter. The effect of different geometrical parameters on the transducer performance has been studied. The proposed flat cap mushroom shape piezoelectric transducer could achieve −186.5 dB sensitivity and a very wide bandwidth. Another important advantage of the proposed structure is that by controlling the pillar radius and without changing the plate size, the sensitivity can be enhanced.

Quantitative depth estimation, along with enhanced defect detectability, is of utmost importance for subsurface analysis in thermal wave imaging for a variety of applications. However, the size and the depth of the subsurface anomalies influence this quantitative analysis due to the non-consideration of back reflection from the defect boundary in addition to three-dimensional scattering effects. This study explores an experimental validation of an analytical model for quantitative depth analysis of subsurface anomalies in thermal wave detection and ranging using quadratic frequency-modulated stimulation with pulse compression based signal processing approach and presents the depth resolution feature by considering the back reflection at the defect boundary. It also presents a study on the influence of the size of the anomaly and bandwidth of the stimulation on quantitative depth prediction using the experimentation carried over a carbon fibre reinforced plastic and mild steel specimen with artificial flat-bottom holes.

Transient voltages from (dis)connecting compensation coils near high voltage transformers are studied. Non-invasive sensors are needed in order to exclude any influence by the measurement equipment on the overvoltage magnitude. This study describes results from open-air capacitive sensors as part of a differentiating/integrating measurement system. Its bandwidth (up to 10 MHz) allows to record large detail in the switching events. However, the coupling strengths and the cross-couplings inherent to the use of air capacitive sensors need to be determined. Calibration based on symmetric phase voltages during normal operation supplies part of the information. Three methods are compared to supplement this knowledge for establishing the complete coupling matrix between phase conductors and sensors. One method employs estimates from the system configuration, in particular, its symmetry. The other techniques extract information from fast transient events within the responses for both switching on and off. The decoupled waveforms from different methodologies align within their confidence bounds and predict the maximum overvoltage magnitudes within an accuracy of 5% when employing information from the fast transients. Based on measurements performed on two compensation coils, both tested three times, larger overvoltages occurred during energisation compared to de-energisation with maximum amplitude exceeding three per unit.

This study discusses stability analysis of the brain emotional learning-based intelligent controller (BELBIC) for model-free tracking control of a class of uncertain non-linear systems. Due to the unique adaptation laws for the controller gains, stability analysis of these controllers is a challenging task. This issue constructs the motivation of this study to undertake this task. Similar to direct adaptive fuzzy control, the BELBIC plays the role of a direct intelligent controller. Thus, the stability of the BELBIC can be guaranteed based on the stability analysis of direct adaptive fuzzy control. The main objective of this study is that the stability of the BELBIC in tracking control of a class of uncertain non-linear systems is guaranteed. Based on the proposed stability analysis and Barbalat's lemma, BELBIC can yield in the asymptotic convergence of the tracking error to zero. In comparison with the existing works in the literature, the proposed approach has fewer tuning parameters. Therefore, the developed approach has a simple structure. The simulation results on an articulated manipulator actuated by permanent magnet DC motors and comparison with other model-free approaches are presented to show the superiority of the proposed approach.

Compared to bushings with lower voltage levels, the ±1100 kV converter valve-side bushing has thicker radial insulation and longer axial insulation resulting in a more uneven temperature distribution and more severe electric field distortion, especially under high-ambient-temperature and high-current conditions. Improving the distribution of the electro-thermal coupled field of the bushing is critical for the safe operation of converter transformers. This study discusses the distribution of electro-thermal coupled field of a typical bushing and proposes a new method to evaluate the conductivity of multiple capacitor layers under inhomogeneous axial temperatures. The effects of four bushing designs are investigated as per the adjustment of the capacitor screen structure, the replacement of the current-carrying structure, and the modification of resin-impregnated paper material. The results show that merely adjusting the capacitor screens is not feasible due to size limitations. By decreasing heat accumulation, adopting a single-tube structure and increasing the thermal conductivity of resin-impregnated paper material can reduce electric field distortion to a certain extent. Unlike the other three methods, hierarchically controlling the material resistance effectively balances the internal and external electric fields. Simulation results demonstrate the feasibility of these methods to improve the distribution of the electro-thermal coupled field in the bushing.

In this study, a particle swarm optimisation (PSO) algorithm coupled with finite-element method (FEM) is implemented to improve the performance of a cap-and-pin insulator by reducing the value of maximum electric field strength. Two goals are set in this work: (i) optimisation of the tangential electric field distribution and magnitude (which is taken as the objectif function to be minimised) to reduce the risk of flashover due to surface pollution, and (ii) reduction of the creepage distance (by adopting adequate constraints) leading to a reduction of the surface area and consequently a decrease in the insulator weight. The investigation is divided into two parts; first, one-unit insulator is optimised with which a chain is constructed, and its performance compared with the reference insulator string. The second part considers a whole chain of four-unit insulator whose performance is optimised and compared to that of a reference insulator string. The main finding of this work indicates that, if an insulator string is constructed using an optimised cap-and-pin unit, the performance of the so-formed string is also optimised. The optimisation of a complete string leads to practically the same performance as that of a string obtained by assembling optimised insulator units.

A p-norm extreme learning machine (ELM) based on sparsity constraint is presented in this study for tracking of fundamental frequency, harmonic and dc in current power signals which finds application in phasor measurement units for wide area power network in smart grid environment. Real-time power applications typically are furnished with on-board controller and hence have constraints to stock a complex architecture. Moreover, the data from online practices are polluted by noises of diverse statistical features obtained on a sample-by-sample basis. Hence, approaches with improved learning paradigm and close model dealing with noises of varied statistical characteristics are essential. The proposed approach formulates a cost function with recursive p-norm error criterion and sparsity penalty that updates the output weights in succession besides adjusting some coefficients of the output weights to zeros that promotes quicker convergence and higher accuracy results. Exhaustive computer simulations have been carried out with synthetic signals and real-time signals to track the dynamic changes in the power signal amplitude, phase and frequency that demonstrate the accuracy, efficiency and robustness of the proposed p-norm ELM. Additionally, the new ELM network also is validated on a field programmable gate array (FPGA) hardware to prove its practicability towards current developments on phasor measurement units.

An efficient diagnosis method dedicated to embedded wiring network based on reflectometry technique is developed in this study. The proposed methodology is based on the two complementary steps. In the first step, the time-domain reflectometry (TDR) method is simulated, by RLCG (R: resistance, L: inductance, C: capacitance and G: conductance) circuit model and the numerical finite-difference time-domain method, and at the same time the datasets are created. In the second step, the support vector machine (SVM) algorithm is combined with a principal component analysis to identify the faults on wiring network from the TDR response. Two types of SVM models have been used in the diagnosis procedure: SVM classifiers and SVM regression models. In order to illustrate the performances and the feasibility of the proposed approach, numerical and experimental results are presented.

The study generalises the calculation method described by Lucca, devoted to the evaluation of induced voltage and current on a pipeline due to a nearby overhead power line in a fault condition, from the case of homogeneous soil to the case of a two-layer soil. By means of some examples of calculation, it is put in evidence that the homogeneous soil hypothesis can lead, in some cases, to excessively precautionary results and, in other cases, to underestimated results. Therefore, a calculation method based on the two-layer soil model seems a better approach.

A routine using multiple methods is proposed for detecting and locating hot spots in power transformers based on thermographic measurements and advanced computational and analytical techniques. The proposed method identifies oxide and hydrocarbon gases dissolved in the insulation oil of power transformers and use the well-established key gas method to provide information about the cause of the hot spots, such as partial discharge, electric arc, and overload heating. The method then uses infrared mappings and temperature measurements to find the approximate location of the hot spots and, finally, the finite element method and an analytical formulation are employed to refine the location of the hotspots and the estimation of their temperature. The proposed method was developed during the design and manufacturing of a commercial 120 MVA generator step-up transformer, produced by a well-renowned manufacturer. The authors conclude that a method is a resourceful tool for identifying hot spots and could be used by manufacturers to identify and predict design and manufacturing failures.