With the increasing penetration of photovoltaic (PV) systems, the problems posed by the inherent intermittency of small-scale PVs are becoming more severe. To address this issue, it is critical to involve the uncertainty of PV generation in the look-ahead periods in a comprehensive framework. To this end, a direct deep learning architecture for probabilistic forecasting of solar generation is proposed in this paper. An end-to-end deep learning architecture as a novel mixture density network (MDN) is designed based on the combination of a convolutional neural network and a gated recurrent unit. Furthermore, a new loss function and training process based on adversarial training is proposed to enhance the accuracy in direct contracting of the probability density function. Then, several deep and shallow networks are implemented, and the results are compared with the proposed architecture. The effectiveness of the proposed MDN in providing complete statistical information is verified through comparison with Monte Carlo dropout, non-parametric kernel density estimation, and the proposed MDN without adversarial training.

The increasing penetration of distributed generations enables an innovative operation paradigm that allows islanded operation to enhance the resilience of the distribution grid. In this study, a cyber-physical oriented islanding strategy is proposed by coordinating centralised and distributed control to achieve seamless islanding transition and operational flexibility in emergency conditions. A cyber-physical control structure is developed to mitigate various disturbances (e.g. emergencies or fluctuations) according to different operation conditions. Specifically, the distributed fault isolation and seamless islanding transition are coordinated to mitigate the outage caused by unplanned islanding, while a secondary control is proposed to support primary control by reducing the power fluctuations during islanded operation. With a rapid response speed, the local cyber-physical devices are coordinated to accomplish islanding separation by selecting a feasible islanded area even under an unplanned islanding situation. A field test was conducted on a practical distribution network in China, and the results demonstrated the effectiveness and feasibility of the proposed islanding strategy.

The new grid operation constraints, allied to the substation automation and intelligence, require the state estimation (SE) to work even more efficiently, with reliability and accuracy. The proposed method for optimally allocating existing phasor measurement unit (PMU) channels and their resources at substations ensures reliability in the SE process even in cases of equipment malfunctioning, and guarantees the quality improvement of the SE. The proposed algorithm has two stages. In the first stage, a genetic algorithm optimally allocates the PMUs. In the second stage, a heuristic method optimises measurement and channel allocation. Equipment failure, missing data, observability issues, methods to reduce critical sets in SE results to preserve the system's observability, are also discussed.

With the development of low-wind-speed technology, it becomes a trend that low-wind-speed wind turbine generators (LWTGs) are integrated into a microgrid. However, the frequency stability of the microgrid has thereby been challenged since the increased penetration of wind power lowers the inertia of the microgrid. In order to investigate how LWTGs can effectively participate in suppressing the frequency fluctuation of the microgrid, virtual inertia control, over-speed control, as well as droop control, is applied to LWTG. Moreover, the de-loading ratio of over-speed control, along with the control parameters of virtual inertia control and droop control are all optimised under different wind speeds by virtue of the deep belief network, whereas the problem of over-speed control failure with the scheme of fixed de-loading ratio becomes more pronounced under low-wind speeds, which is defined as a blind area problem. To solve this problem, on the one hand, the strategy of the variable de-loading ratio is adopted under low-wind-speeds. On the other hand, the concepts of the minimum and maximum critical wind speed are deduced through theoretical analysis, which greatly restricts the number of feasible solutions of de-loading ratio under different wind speeds so as to improve the optimisation efficiency about 50%.

This study proposes an integrated framework for coordinated optimisation of the interdependent microgrid (MG) and electric vehicle (EV) fleet entities using the normalised normal constraint approach. By considering the active/reactive power management option of the bidirectional charger enabled EVs in the proposed model, the authors investigate the effectiveness of EV's integration in the presence of the techno-economical objective functions. This work concentrates on the trade-off analysis of two conflicting objectives, including the economic objective of the MG's operation cost minimisation and the technical objective of the MG's voltage deviation. Besides, they consider several uncertainty sources, e.g. wind, EV and solar panel (PV) power provision, as well as market price fluctuations in the proposed model affecting the aforementioned techno-economic trade-off solution. The proposed model is a stochastic multi-objective mixed-integer non-linear programming problem where the authors apply the designed integrated framework on a modified IEEE 18-bus test case in GAMS software. Through numerical results, they demonstrate MG optimal operation changes due to different MGO priorities and study the positive effects of EVs integrated energy management on the bi-objective operation.

This paper proposes a novel approach to designing technical and financial protocols needed to support the penetration of distributed energy resources (DERs). It first formulates a complex, hard-to-implement, centralized decision-making objective for providing end-to-end electricity service. It then introduces a new taxonomy of an end-to-end interactive operations planning framework. The taxonomy rests on the dynamic monitoring and decisions systems (DyMoNDS) principles for supporting interactive protocols of (i) end-to-end interactions within a complex, multilayered multi-voltage power system; (ii) dynamic energy resource management system (DERMS) interactions with their DERs as well as with the bulk power system (BPS) operators; and (iii) DERs interactions with DERMS. The distributed model predictive control for creating physically implementable cost functions is essential. Also, the minimal coordination of different layers utilizes an AC optimal power flow that is essential for ensuring power flow delivery. We next provide a proof-of-concept illustration on the IEEE 14 bus system augmented by two standardized microgrids of the proposed interactive protocols, and their potential use for enhancing dynamic host capacity (DHC). While novel, this approach is a natural outgrowth of the existing industry operations: It only requires enhancing decision-making tools by the stakeholders, and carefully-defined protocols for implementing their interactions.

In this study, energy management and damping improvement of a DC microgrid are proposed by using an interconnection and damping assignment–passivity-based control (IDA–PBC) technique. The control technique is applied for (i) the grid connected inverter and (ii) the DC–DC converters connected with fuel cell and battery of the DC microgrid to control the DC bus voltage. The IDA–PBC generates the control laws with the desired energy function for the current controller of the converters based on the generation from the renewable power generators and load demand. In addition, an integral action is added with the IDA–PBC control laws to reduce the steady-state error in the DC bus voltage. The parameters of the IDA–PBC technique are tuned based on the state of charge of the battery and grid availability for a smooth transition between the operating modes and better energy management. It is proved that the derived energy function satisfies the Lyapunov stability criterion. Simulations are carried out in MATLAB/Simulink to test the performance of the proposed control technique. The results show that the proposed technique provides the desired damping and energy management.

This study describes a synchrophasor estimation method and its hardware implementation applicable at the distribution level of the power system, under steady-state or dynamic conditions. The method proposed is based on the discrete Fourier transform and uses the Savitzky–Golay filter in the frequency estimation process. The main goal of this study is to present the signal processing algorithm for phasor estimation and some important issues regarding its implementation in real-time systems. The performance of the micro-phasor measurement unit (-PMU) is validated using a waveform generator and real-time power system simulator.

The variable step size modified least mean square (VSSMLMS) learning algorithm based multifunctional double-stage grid interactive photovoltaic (PV) system is implemented in this paper which feeds power to the grid and the load, ensures unity power factor operation and compensates reactive power. Along with this, it mitigates high grid neutral current during unbalanced loading condition. This algorithm uses variable step sizes to have a smooth dynamic and steady state performances. The boost converter (booster) is used in the first stage to extract peak power from a PV array using perturb and observe peak power point tracking technique. Next to this, a four leg voltage source converter (VSC) is placed to interlink the solar PV system to the grid and the loads. VSSMLMS-based control technique is used to filter the fundamental component of load current and hence gate pulses are generated for the four-leg VSC. This system is simulated in SIMULINK/MATLAB software under several working conditions and the obtained simulation results are validated on a developed laboratory prototype. The performance of the VSSMLMS-based control technique is compared with that of the traditional LMS-based technique. All results validate the IEEE-519 standard.

This study proposes a novel flexible hybrid state estimation (SE) algorithm when a realistic communication system with its irregularities is taken into account. This system is modelled by the Network Simulator 2 software tool, which is also used to calculate communication delays and packet drop probabilities. Within this setup, the system observability can be predicted, and the proposed SE can decide between using the static SE (SSE) or the discrete Kalman filter plus SSE-based measurements and time alignment (Forecasting-aided SE). Flexible hybrid SE (FHSE) incorporates both phasor measurement units and supervisory control and data acquisition-based measurements, with different time stamps. The proposed FHSE with detailed modelling of the communication system is motivated by: (i) well-known issues in SSE (time alignment of the measurements, frequent un-observability for fixed SE time stamps etc.); and (ii) the need to model a realistic communication system (calculated communication delays and packet drop probabilities are a part of the proposed FHSE). Application of the proposed algorithm is illustrated for examples with time-varying bus load/generation on two IEEE test cases: 14-bus and 300-bus.

This study presents an improved double closed-loop current (fundamental and harmonics current) control method for inductive-capacitive-inductive (LCL) type shunt active power filter (SAPF), which is designed to enhance the robustness of the SAPF system to adapt weak grid application condition. Due to the variation of weak grid impedance, fundamental current control loop based on grid current feedback control method may become unstable, a robust parameters design method in discrete z-domain according to amplitude-frequency and phase-frequency characteristics is proposed to fit the grid impedance variation. The harmonics current of non-linear load exists in a wide frequency range, normal resonant controller will cause the system unstable due to the negative −180° cross of phase at high-frequency range, a digital phase-lead resonant controller (PL-RC) is proposed to extend the bandwidth of the harmonics current control, then the quantities of the PL-RCs used to suppress harmonics can be increased considerably compared with normal resonant controller. Experimental results are presented to verify the effectiveness of the robust parameters design method and the proposed PL-RC.

A hybrid AC/DC distribution network (HDN) is formed after multiple hybrid AC/DC microgrids (MGs) and distributed generators (DGs) are integrated into the distribution network. Voltage/var control (VVC) in this evolved system constitutes a big challenge to system operators as a large variety of voltage control devices with quite different control characteristics are expected to be co-managed. This study proposes a coordinated VVC scheme to regulate voltage in an HDN by integrating power management model of hybrid AC/DC MGs into HDN's VVC model. The devices at HDN's level and those in MGs are allocated into two different models, and two models are linked via an approach called mathematical programmes with equilibrium constraints to achieve system-wide coordination. In the proposed VVC, mechanical devices such on-load-tap changer (OLTC) and shunt capacitor (SC) are scheduled every 2 h to save their lifetime, while the electronically interfaced DGs and MGs are scheduled every 30 min to leverage their fast power support. Case studies on a modified IEEE 33 nodes distribution system validate that the proposed VVC can effectively coordinate MGs' power support with OLTC, SCs, and DGs, and it constitutes significant improvements on power loss reduction and voltage quality compared with traditional VVC.

Since the distribution system operator (DSO) cannot directly control prosumers with controllable resources, this study proposes an optimal dispatch method of using three-phase distribution locational marginal prices (DLMPs) as effective economic signals to incentivise prosumers' behaviours. In the proposed three-phase DLMP model, DLMPs for active power demand, active power output and reactive power output are calculated. To alleviate the imbalance, congestions and voltage violations in active distribution networks (ADNs), the DSO and prosumers should be coordinated. The authors develop such a coordinated control scheme for the DSO and prosumers, in which the DSO generates and broadcasts three-phase DLMPs as price signals to induce prosumers' behaviours. They prove that given the DLMPs as settlement prices, the optimal dispatch of the ADN will also maximise the surplus of prosumers. Therefore, the power output of rational prosumers will match the optimal dispatch, resulting in better operational conditions of ADNs. Then the three-phase imbalance, congestions and voltage violations will be well reduced. Numerical tests demonstrate the effectiveness of the proposed approach.

This study presents a new protection scheme for a series compensated transmission line, which prevents maloperation of distance relay during severe stressed conditions. The proposed method depends on the calculation of the net angular difference of positive sequence phasors of voltage and current signals acquired from phasor measurement units (PMUs). For fully observable power system network, the minimum number of PMUs required with a maximum number of system observability redundancy index has been identified by the suggested method using binary chemical reaction optimisation technique. The validity of the suggested technique has been confirmed by generating different severe stressed conditions along with faulty situations by modelling IEEE-39 bus system using RTDS/RSCAD software. The proposed technique offers effective discrimination between severe stressed conditions (symmetrical/asymmetrical power swing, load encroachment, voltage instability) and faulty situation even against wide variation in fault type, fault resistance, fault inception angle, degree of compensation and location of the series capacitor (middle or both ends). The accomplished results disclose higher discriminating capability between severe stressed conditions and faulty situation of the proposed technique in comparison with those of the several existing techniques.

This study presents a methodology to solve simultaneously the alternating current (AC) transmission network expansion and reactive power planning problems, considering multiple stages and operating conditions. A mixed-integer non-linear programming model for the proposed planning problem is presented and rewritten with semidefinite structures. Then, the generalised Benders decomposition is used to separate the overall problem into an upper-level (master) problem and several lower-level (slaves) problems. The master problem is a mixed-integer linear programming problem that optimises the investment cost and constraints of the multistage expansion. Each slave problem minimises the operating costs associated with each stage and operating condition (normal operation or contingency), considering the AC power flow via semidefinite relaxation. With the proposed methodology, the global optimality of generalised Benders decomposition can be preserved due to the use of semidefinite relaxation in each slave problem. Garver's 6-bus system and an IEEE 118-bus system are used to show the precision and convergence to near-global optimal solutions with small relaxation gaps through the proposed approach.

Ensuring a stable and reliable operation of current and future distribution networks represents a major challenge for system operators aggravated by the global proliferation of plug-in electric vehicles (PEVs). While the introduction of a controlled charging process would be advantageous to minimise the impacts PEVs cause in the system, a suitable, efficient and ready to be implemented solution is still missing. The present work addresses this issue by proposing a smart charging management solution capable to simultaneously combat the main network impacts derived from the energy needs of the vehicles. This is done by means of an agent-based hierarchical real-time algorithm which combines a local decentralised nodal voltage management with a centralised thermal control conceived to minimise the impact upon participating users. The effectiveness of the proposed system is tested both using a simulation environment considering multiple PEV penetration levels and employing commercially available charging stations and cars through hardware-in-the-loop simulations. The results reveal how all network violations are successfully attenuated by peak shaving the total aggregated charging demand and ensuring a correct system operation for all penetration scenarios while inflicting no impact on the participating users.

A microgrid (MG) is a cluster of small-scale sources known as micro-sources, the energy storage devices, and the loads. MGs are promising entities that allow a considerable amount of renewable energy penetration into the system. A stochastic optimisation model is presented in this study to find the optimal size of the distributed energy resources (DERs). The uncertainties are considered with the help of multiple scenarios of random variables viz. wind output, solar output, and load demand data. A particular case of Beta distribution is used to create these variations. A cost-based, multi-variable constrained non-linear objective function is formulated. The considered test MG system is more generalized with solar and wind energy penetration. The results show the effect of operational as well as the planning aspects together under various reliability conditions for an isolated MG. The formulated problem is solved to obtain a global optimal solution using a sequential programming approach and a comparative assessment of results with a hybrid particle swarm optimisation (PSO) approach has also been presented. It is found that the results obtained from the sequential programming approach are better compared to the hybrid PSO approach.

Virtual power plants (VPPs) of secondary distribution network have come into existence due to the installation of generators in the premises of domestic consumers and the recent legislative emphasis on net-zero energy operation. A predetermined operation of VPP as required by a system operator is possible only when its power output is assessed beforehand. A simple approach for assessment based on load profiling using descriptive statistical parameters such as z-scores of the load data is proposed in this study. The proposed method collates historical customer data according to seasons, days of the week and specified time-intervals of the day. Then, z-scores are computed for each of the historical load and generator data. The calculated z-scores are subsequently grouped into specified clusters. The grouped data in each of the clusters are then employed to predict the VPP output. Demonstration of the assessment of VPP output using available Pecan street data is presented to validate the proposed approach.

In the field of transformer failure diagnosis, the potential correlation between different characteristic parameters and failures is difficult to detect using traditional methods. Further, the quantities of inspection data have not been fully utilized. To improve the accuracy of transformer diagnosis, this study establishes a diagnosis model based on fuzzy association rules combined with case-based reasoning (CBR) to evaluate the failure types, fault locations, and cause of breakdown in power transformers. First, the inspection data of transformers are collected from several substations over 10 years. Then, the pre-processed data are randomly separated into training and testing sets. For the training set, fuzzy association rules are built for multiple parameters to narrow the search scope of base case preliminarily. Next, CBR is applied to determine the most similar cases. The failure information of the target transformer can be obtained in detail along with the most similar base case. Finally, the accuracy of the model is validated shown in case studies using the testing data set. The result demonstrates that the diagnosis model provides a higher accuracy than the classic IEC 60599 three-ratio method used in the current industry, which means that this diagnosis model has better performance on diagnosis accuracy.

With the rapid development of machine learning technology, a new tool is provided for real-time stability evaluation in power systems. The training of a machine learning-based model is inseparable from a large number of training samples. However, compared with stable samples, unstable samples in power systems are infrequent. The results of the model evaluation are biased due to the imbalance of training samples. Faced with such a problem, a framework based on deep imbalanced learning is proposed. Firstly, for each sample, the samples nearby in the opposite class are applied to calculate its space information. Based on the space information of all samples, the spatial distribution characteristics of the training samples are obtained. And then, in order to obtain balanced training samples, unstable samples are generated according to their spatial distribution characteristics. Finally, stacked sparse denoising auto-encoder (SSDAE) based model, which has the ability of anti-noise, is established as the classifier. Simulation results in IEEE 39-bus system show the high performance of the proposed imbalanced correction scheme and evaluation scheme.