Flexible DC transmission technology based on voltage source converter (VSC) is an effective solution to solve the reliability of power supply to weak AC systems. In this study, a VSC–high-voltage DC (HVDC) electromechanical transient simulation model is established based on the Chongqing-Hubei flexible DC transmission project which is the highest voltage level and the largest transmission capacity in the world. This model can be used to simulate the behaviour of the limiters and AC fault ride through strategies. A large disturbance measurement method for VSC's non-linear power characteristics is proposed. Stability of flexible DC and weak AC hybrid grid under large disturbance are analysed, behaviour mechanisms and key influence factors of different instability forms are revealed. In this way, a solution is proposed to improve the hybrid grid's ability against large disturbances. By the construction of the VSC–HVDC electromechanical transient simulation model, the correctness of the mechanism analysis and effectiveness of the solution are verified.

This paper proposes a WECS based on a back to back diode-clamped multilevel inverter systems (DCMLI) fired comparatively by sinusoidal pulse width modulation (SPWM) and third harmonic injection pulse width modulation (THIPWM) techniques. DFIG performance is compared using these technologies for different wind speed under normal operation condition. The proposed approach shows that the DCMLI systems generate a nearly sinusoidal voltage with reduced total harmonic distortion (THD) thus upgrading the power quality of that produced by DFIG. As SPWM and THIPWM MLI are most common usage techniques in research, we applied them in the system to discuss a comparison of using those techniques to determine the number of levels that achieve the IEEE 519 criteria of THD which is less than 5% for both techniques (5th level for SPWM and 4th level for THIPWM) in all operation wind speed. Compared with simple PWM, and THIPWM have many advantages as lower THD, minimal number of switching to decrease switching losses and the output fundamental voltage is increased. Lastly, the paper investigates the variation of the frequency of induced rotor voltage and the active power flow due to the wind speed change when the rotor speed changes from super-synchronous to sub-synchronous speeds.

Line routing is a critical issue in distribution network planning and conventional raster solutions for line routing have certain deficiencies. This study presents a new methodology for a distribution network planning problem based on a hexagonal raster in a geographic information system. Differing from the conventional square raster, the hexagonal raster-based model involves less binary variables, but with the same arm length, and guarantees that the planning routes are arm-connected. An iterative refining method is proposed to reduce the computational burden with the guaranteed resolution, in which a mixed-integer linear programming problem is solved in each stage. The proposed model can simultaneously optimise electric line routes and pole positions, in contrast to conventional models where the poles are not considered. Numerical case studies illustrate the effectiveness of the proposed approach for distribution network planning.

To manage and control massive distributed energy resources (DERs) integrated into active distribution networks (ADNs), dispersed DERs are aggregated as virtual power plants (VPPs). This study introduces a distributed optimal dispatch strategy of VPPs considering operational constraints in unbalanced ADN. First, a practical multi-VPP coordination optimal dispatch model considering P–Q coupling characteristics in unbalanced ADN is proposed. Then, the primal–dual interior-point algorithm is employed to transform the original model into a Karush–Kuhn–Tucker equation, where afterwards distributed quasi-Newton method is developed to solve the equation. Under the proposed distributed strategy, each VPP only communicates limited boundary information with utility grid. With the advantage of quasi-second-order information, the proposed method outperforms the popular distributed gradient-based algorithm, alternating direction method of multipliers. The distributed method provides merits such as efficient calculation and privacy protection for multi-VPP operation in ADN. Numerical tests evaluate the efficiency and accuracy of the proposed method.

Distribution feeder reconfiguration (DFR) and demand response (DR) are the common energy consumption management methods to enhance the operation quality of distribution networks (DNs). Moreover, DFR and DR may lead to system improvement through different aspects such as reliability, total cost, and power quality. Nevertheless, the high complexity of the new smart grids has caused a lot of uncertainty in the reconfiguration problem. Given that, the use of an adequate probabilistic framework is necessary to deal with such issues. Hence, the current study investigates the effect of DR on the DFR strategy in a stochastic environment as a single-period and multiple-objective model at peak load. A parametric probabilistic analysis is administrated based on unscented transform (UT) method to capture the load and generation uncertainties. High-precision results and low computational burden are the outstanding features of the UT method. Furthermore, an improved version of Cuckoo search Algorithm (CSA) as an optimizing tool is proposed. The effectiveness of the suggested method is tested on the modified IEEE 32- and 69-bus systems. Moreover, PDF and CDF functions of random output variables are plotted and compared to verify the proposed stochastic method. The results confirm the efficient operation of the proposed algorithms.

This study presents the control of distribution static compensator (DSTATCOM) using band-dependent variable step size (BD-VSS) individual weighting factor with sign error based adaptive filter for power quality improvement in a weak distribution grid. Here, the proposed algorithm is used for estimation of fundamental active weight components from the distorted load currents in order to generate the reference grid currents to mitigate the grid currents power quality issues. This new control algorithm is proposed for fast and accurate estimation of active weight components with low steady-state error, without dynamic oscillation at fast convergence speed. Moreover, the frequency locked loop-double self-tuning second-order generalised integrator based voltage filter is used to extract the positive sequence voltages of the distorted grid voltages for estimating the harmonics, and noise-free voltage unit templates. Therefore, the DSTATCOM with the proposed control algorithm is capable of mitigating the harmonic currents, providing reactive power compensation and operating at unity power factor in a weak distribution grid. Test results demonstrate the viability and robustness of the proposed control algorithm under balanced and unbalanced non-linear load conditions.

In this study, a new method for ultra-fast speed busbar protection is presented. The method is based on processing the incoming/outgoing current signals of lines or transformers connected to a substation busbar by the improved morphological gradient. The proposed algorithm is simple and requires neither fault detector nor fault classifier and it is immune to fault type, fault inception angle, fault resistance, and noise with SNRs down to 5 dB. It is applicable to different substation configurations or bus systems. According to the simulation results, the proposed algorithm has high security and dependability in discrimination of external and internal faults with the ultra-fast operation. Moreover, in the case of the existing two busbars in a substation layout, the proposed indices can discriminate between the faults of the two bus zones.

This study proposes a systematic approach for estimating intervals of distribution locational marginal price (DLMP) and corresponding confidence levels without requiring an exact probability distribution of renewable generation. The DLMP is evaluated based on a variant of the alternating current (AC) power flow model, namely, the branch flow model, in which network losses, bus voltage, and reactive power are taken into account. Based on the exactness of convex relaxation of optimal power flow, the authors developed a linear AC power flow model by applying second-order cone relaxation and global polyhedral approximation. Given a set of renewable power volatility, interval prediction of the DLMP is formulated as a bilevel linear program, which is solved via a mixed-integer linear program (MILP). Considering variances and unimodality of renewable power forecast error, a conservative estimation of the confidence level is expressed by a generalised Gauss inequality, which comes down to semidefinite programming. The proposed method is a natural extension of existing interval and probabilistic forecast methods, leveraging the proposed linear AC power flow model and inexact probability distributions of renewable power output, and could be promising in future distribution power markets. Case studies corroborate the effectiveness of the proposed method.

This study introduces a distributed secondary control method-based adaptive pinning control for inverter-based distributed generations (DGs) in an isolated microgrid (MG) structure with frequency and voltage compensation and power sharing. Without MG centre controller or energy management system, local DGs’ controllers have no global information of the whole communication network, which is the second smallest eigenvalue of the Laplacian matrix. In practice, the large coupling weight gains are mostly selected. However, a sufficient large coupling weight may cause instability. It is difficult to choose the coupling weight in the fully distributed complex network. To solve the above problems, the authors investigate a pinning control for synchronisation using the local adaptive pinning algorithm on coupling weight gains. The proposed method has the advantage that the consensus controller design is independent of the Laplacian matrix associated with the communication network. And the collective dynamics of the communication network can be adjusted to steady states without global information. However, the adaptive pinning control scheme is not suitable for the multi-disturbance situation, which will accumulate the coupling weight and even cause integral saturation. Therefore, an improved strategy is proposed to adapt to the load changing conditions for islanded MGs.

Formation of dry zones around power cables is an important factor affecting the cable loading and may lead to cable failure. This phenomenon is ignored by some standards in the calculation of power cables capacity. Numerous efforts have been done by many researchers for cable monitoring. In this study, monitoring system based on measurements of temperature distribution surrounding the cable route is suggested. When reaching the temperature of the dry zone formation the system will produce a warning alarm. To the best of the authors knowledge, the invention of a device that gives a warning in the case of drying layers formation around the power cables has not received attention so far. The proposed device is characterised by cheap price, high accuracy and its simplicity in the design. The system is tested in the laboratory using different soil types. Field tests are done to verify the reliability of the device.

This study presents a bi-level optimised emergency control strategy, including load and generators shedding following DC blocking. Due to different constraints for networks with different voltages, a bi-level optimised model spanning from the high voltages networks to the lower voltages is put forward. For the higher voltage level network, the overload area is identified and the seriousness of the overload is classified after DC blocking with the help of wide-area measurement systems. Then, the objective function is built with the minimum total amount of load shedding and generator tripping to be solved by improved particle swarm optimisation (IPSO), considering the constraints of frequency and voltage stability, respectively. For the low voltage level networks, the least cost as the objective function with the constraints of load shedding and generator tripping strategy based on the analytic hierarchy process-fuzzy comprehensive evaluation method. The modified 10-machine 39-bus system with HVDC lines is used as the test system. The simulation results demonstrated that the proposed strategy can give good control performance with consideration of different optimisation constraints following the DC line blocking.

Polarisation–depolarisation current (PDC) measurement and its analysis is a popular technique for assessing the condition of transformer insulation. Owing to the low magnitude of PDC, recording noise-free PDC data from in-situ power transformers is a challenge. Once the relaxation current data get affected by noise, it becomes difficult to formulate insulation model (as recorded data loses its characteristic shape). This further makes the data difficult to analyse and predict insulation condition. In this study, two de-noising techniques are discussed (one is based on Wavelet Transform while the other is based on Stockwell Transform) for eliminating low-frequency non-stationary noise from recorded PDC data. Comparison between these two techniques suggests de-noising using Stockwell Transform is advantageous over wavelet analysis. The proposed methodology is first tested on data recorded from the sample prepared in the laboratory and then on data measured from real-life in-service power transformer.

This study proposes a novel scheme for dynamic distribution expansion planning (DEP) in the presence of plug-in electric vehicles (PEVs). The model considers investment, production and maintenance cost and identifies the substations and feeders to be built, reinforced or replaced. Owing to the increasing penetration of PEVs into the distribution network, traditional strategies to expand the network have to be updated to cope with the new uncertainties incurred by the PEVs integration. In this regard, a two-stage scenario-based strategy is presented, in which the uncertainties related to the PEVs are modelled via stochastic optimisation using Monte Carlo simulation. In the first stage, the binary decision variables (here and now decision variables) are determined, whereas in the second stage, the optimal production of substations and optimal charging of PEVs (wait and see decision variables) are identified in a day-ahead electricity market. Moreover, the daily electricity price and load volatilities have also been taken into account. The DEP problem is formulated as a mixed-integer linear programming problem and is solved using the efficient Benders’ decomposition algorithm. The results of the case study based on a 24-node distribution system show the feasibility, tractability and effectiveness of the proposed model.

This study proposes a coil current model and an energy storage motor current (ESMC) model of circuit breakers (CBs) with spring operated mechanism. To make sure the signals generated by the models are identical to the actual ones, this study proposes a stochastic optimisation algorithm to optimise the model parameters. Based on the data produced by the optimised models, two fault diagnosis methods are proposed to assess operational condition and detect faults. The first method is based on fast template matching, which adopts K-means clustering algorithm to cluster the data and form a template library. The second one combines deep belief network and Softmax classifier, which can not only extract high level information of the characteristic signals, but also avoid the negative impact of the large dimension on classification results. In the simulation studies, the two methods are tested on various scenarios and their merits are demonstrated, respectively, where the latter one shows superior performance.

This study proposes a decentralised methodology to deal with the optimal energy flow (OEF) problem of the electricity–natural gas system (EGS). The OEF problem of EGS is formulated as a multi-stage dynamic programming (DP) process, and the approximate DP algorithm is used to decompose it into subproblems. The independent decision of each subproblem can be made by solving Bellman's equation, requiring just a moderate interchange of information among different subproblems. Such that the information privacy and dispatch independency of subsystems (including electricity and natural gas (NG) subsystems) can be ensured. Different from most existing decentralised algorithms, the impact of one subproblem's decision on other subproblems can be estimated in the proposed algorithm, and the parameters of the proposed algorithm are not required for tuning, which makes more sense in real applications. Furthermore, an improved linear cut technique is proposed to handle the non-convexity of the NG network model, so that the tightness of all gas flow constraints can be ensured. Case studies containing small, large and complicated EGSs validate the effectiveness of the proposed approach.

During recent years, allocating the network losses to loads and generation units based on network transactions is an important issue in the power industry. This study presents an appropriate method to allocate the network losses to loads and generation units according to their network transactions. This method is appropriate and faster than other loss allocation techniques and does not consider unreasonable assumptions in the loss allocation process. This method explicitly assigns the network losses according to individual transactions, and this leads to allocate the network losses among all network transactions. This approach is based on the share of the loads and generation units on active and reactive power flows passing through the lines according to transactions. The implementation of the method on large networks is simple and fast. To approve the performance of the presented method, the proposed loss allocation technique has been applied to a three-bus case study system and the IEEE 14-bus test system. Also, several different transaction modes are used to show the performance of the presented method for transmission loss allocation based on transactions.

This study proposes an efficient local energy management system (LEMS) based on the generalised power prediction model for the uncertain operation of renewable distributed generations (DGs)-based microgrid. Photovoltaic with battery energy storage, and wind power generation are considered as primary DGs to compensate intermittency. Conventional direct power prediction models are limited to specific DG applications, where the plant data acquisition system is a necessity. Solar irradiance and wind speed are considered here as prediction targets to cope with such additional expenditure for a microgrid. To ensure a robust reduction in prediction error (e_p ), a short-term prediction model is developed by virtue of the proposed robust regularised random vector functional link network. A maximum-likelihood estimator using Huber's cost function is employed to attain the robustness of this model. Further, a direct renewable energy source-power calculation is opted to address model accuracy under local uncertainties. The LEMS operation is completed by compensating e_p with distributed adaptive droop-based primary controllers for multi-DG based microgrid. To ensure the performance of the prediction model, solar irradiance, wind speed and power at different atmospheric conditions (seasonal volatility) and time span (i.e. 5, 10 and 60 min) have been implemented in MATLAB and real time.

The peak value of continuous operating voltage (PCOV) has a direct influence on the reference voltage of valve arrester, which appreciably influences the manufacturing cost of the line-commutated converter (LCC). Since simulations with high accuracy based on broadband models are memory intensive and time consuming, it is important to determine the operating conditions for PCOV calculation to reduce computational burden. This study establishes piecewise expressions of ideal valve voltages in different converter operation modes, which is the basis for the investigation of commutation overshoots. A simplified circuit is established to evaluate the valve voltage stress, with a modification introduced on the calculation of voltage peaks across blocking valves, which is neglected by previous literature. Then, this study proves theoretically that the 6-pulse rectifier operation mode and the long-term overload power transmission with the maximum delay angle should be the corresponding operating condition for PCOV calculation. Conclusions derived from the analytical analysis have been validated by simulations based on broadband models for practical projects.

Dissolved gas analysis (DGA) is commonly used to identify the fault type in power transformers. However, the available DGA methods have certain limitations because every method depends on the concentration of the dissolved gases. Therefore, in this work, hybrid feature selection–artificial intelligence–gravitational search algorithm (GSA) techniques were proposed to determine the fault type of power transformers based on DGA data. The artificial intelligence (AI) methods applied include support vector machine and artificial neural network. Both AI methods were optimised by GSA to enhance the accuracy of the results. Feature selections using stepwise regression and robust regression were applied to utilise only significant gases. The accuracy of the results was tested with various ratios of testing and training data. Comparison of the results using the proposed method with other optimisation methods and the previous works was performed to validate the performance of the proposed technique. It was observed that the proposed hybrid feature selection–AI–GSA technique yields reasonable accuracy although fewer types of dissolved gases were used. Therefore, the proposed method can be recommended for the application of automated power transformer fault type detection based on DGA data in practice.

The accuracy of power system state estimation (PSSE), its robustness against bad data and the speed of its algorithm are crucial to economic and secure system operation. On the other hand, observability and redundancy considerations mandate PSSE to take advantage of traditional supervisory control and data acquisition (SCADA) measurements along with available phasor measurement unit (PMU) measurements. This set of hybrid PMU/SCADA inputs has traditionally made the problem formulation non-linear, and hence time-consuming to solve due to the iterative process of solution. This study addresses the foregoing challenges by proposing a novel linear least-absolute-value (LAV) estimation, without the need for an initial guess of the system state. The linearity of the proposed PSSE formulation is guaranteed regardless of whether PMU-only, SCADA-only or hybrid SCADA/PMU measurements are utilised. This facilitates the fast and non-iterative solution of the LAV estimation of system state based on linear programming. The LAV estimator outperforms the weighted-least-squares estimator in dealing with erroneous measurements, by automatically rejecting bad data of any size. An extensive number of simulation studies carried out on test systems of different sizes confirm the superiorities of the proposed method in comparison with other existing PSSE methods.

Identifying defects in electrical power systems during field inspections is a difficult task, since faults are generally not visible and may be intermittent. To find possible adverse conditions, specific inspection equipment is used. The ultrasound detector is the equipment normally used to inspect outdoor insulating systems; however, using it demands operator experience. To improve the defect condition classification, artificial intelligence techniques are applied to assist the operator in the decision task and thereby facilitate the identification of faulty insulating devices in the grid. The training of artificial neural network (ANN) models is an important step in solving the classification problem. This study aims to evaluate the training capacity in terms of the performance of different optimisation methods for the calculation of the mean square error after convergence. Traditional methods such as Gradient Descent and its variations will be presented, as well as methods that employ high computational effort such as quasi-Newton and Levenberg–Marquardt. In order to base these concepts, a review will be presented on the use of these algorithms and on the problem of classification of insulators in distribution networks. The results show that there is a considerable performance difference between the calculation methods.

In this work, the development of sign regressor least mean mixed norm (SRLMMN) control technique for a distribution static compensator (D-STATCOM) is presented. This control technique extracts fundamental weight components from the non-sinusoidal load currents and generates reference grid currents. D-STATCOM is developed for harmonics eradication, reactive power injection and load balancing and its performance is investigated in several operating modes. The performance of SRLMMN control is compared with recursive least square (RLS) and variable step least mean square (VSLMS) control techniques in terms of convergence, steady-state error, harmonics elimination, sample time and computation complexity. The major advantages of SRLMMN control technique, are fast convergence, less steady-state error, low total harmonic distortion (THD) and offers less computation complexity when compared with RLS and VSLMS. A laboratory-scale prototype of compensator is realised using a voltage source converter with the controller implemented in the dSPACE-MicroLabBox. Both MATLAB simulation and experimental results are included to demonstrate the performance of shunt compensator under steady-state and dynamic loadings. The developed SRLMMN control technique mitigates power quality problems and effectively suppresses THD observed in the grid current with reference to IEEE Standard 519–2014.