This study provides a comprehensive literature review of techniques for harmonic related power quality improvement of electrical generation systems. Increasing interest in these aspects is due to ever more stringent power quality requirements, deriving from new grid codes and compliance standards, aimed at limiting waveform harmonic distortion at all points of the distribution network. Although a wealth of literature is available for such techniques, it has never been compiled into a handbook incorporating all the solutions aimed at both electrical machine and power system engineers.

Transcontinental grid interconnection and clean energy development for sustainability are the prime objectives to address through global energy interconnection (GEI) platform. Key issues can be solved through GEI such as environmental pollution, climate change, resources scarcity and unbalanced development. The growing concern of fossil fuel depletion leads to the exploitation of renewable energy which is mostly located on Arctic and Equator zones. This study addresses about the importance, current projects and research directions of clean energy, smart grid, ultra-high voltage transmission, grid interconnection trends around the world in order to globally interconnect the future grid. The review results provide comprehensive background knowledge to all the researchers in order to investigate further into the field.

The existing research on furfural analysis mainly focuses on the relationship between the furfural content in oil and the degree of polymerization (DP) of the paper, which are always obtained by accelerated thermal aging at high temperatures. However, the operating temperature of a winding is relatively low, approximately 60°C to 90°C. Thus, the relationship established at high temperature may be not suitable for the operating temperatures. In this research, aging tests of oil-paper systems at different temperatures ( from 100°C to 130°C) were carried out, with the DP, furfural content in oil and furfural content in paper being measured regularly to analyze the effect of temperature variation on the production and diffusion behavior of furfural in oil–paper insulation. The results show that temperature influences the production of furfural, i.e., there is no obvious difference in the amount of furfural produced by paper in the initial aging periods; however, a higher temperature results in more furfural being produced when the DP decreases to values below 600. An obvious diffusion process occurs after furfural is produced in paper in accelerated aging experiments, resulting in most of the furfural content remaining in the cellulose paper in the initial aging period.

This study describes the use of entropy model for some systems used in the dielectric technique. In this manner, the changes considered in the breakdown voltage are similar to those described by the probability entropy formula. This is particularly the case due to the radius ratios of spherical and cylindrical electrode systems modelled in the high-voltage technique. The breakdown voltage variations in the spherical and cylindrical electrode systems are consistent with the Shannon formulation of Information. These information curves demonstrate how the system should be operated under optimal conditions and these curves exhibit entropy changes due to voltage and field strength.

For its high coefficient of performance and zero local emissions, the heat pump (HP) has recently become popular in North Europe and China, which shows potential in absorbing local photovoltaic (PV) generation. In this study, the authors describe a distributionally robust optimisation (DRO)-based HP day-ahead operational schedule model (HP-DOSM) to match the PV power generation, which can well capture the uncertainties of weather, PV, and load prediction errors. Moreover, this DRO-based HP-DOSM can be transformed into a tractable deterministic model. The DRO method they proposed is suitable for linear expectation constrained optimisation whose ambiguity set is constructed using Kullback–Leibler divergence, which could be further transformed into deterministic conic/linear constraints. Compared with robust optimisation (RO) models, the authors’ model is less conservative since more statistical information on the uncertainties is utilised. Numerical tests were conducted to demonstrate its performance, compared with the RO model via Monte Carlo simulations.

The static synchronous compensator (STATCOM) has attracted considerable attention because it can stabilise severe transients that are caused by power system disturbances. This study presents a novel interval type 2 fuzzy control system (IT2-FCS)-based controller for the STATCOM to stabilise bus voltages that are caused by faults or forced wind farm outages in a smart grid. Two IT2-FCSs are presented to tune increments of the proportional integral (PI) controller, which are optimised by the gradient descent method. The IT2 fuzzy rules, involving upper and lower membership functions, result in a fast and stable system response. Since many possible scenarios may arise in the power system, the Taguchi method is used to design experiments using an orthogonal array, in which all scenarios are mutually independent. A power system that consists of a wind farm and STATCOM is studied. Comparative studies show that the proposed method is superior to traditional PI and type 1 FCS methods.

When a wind energy conversion system (WECS) based on a doubly fed induction generator is operating in a different mode than maximum power tracking, there exist two different modes of operation. Here, it is shown that such modes satisfy the torque balance condition between the WECS and the electric network, which is described by a transcendental equation in terms of the desired angular velocities. The analytic solution to such equation is the newly found generalised Lambert W-function. Using a real WECS, the authors make an analysis of the lower and upper torque modes of operation. Finally, the authors propose an effective formula to estimate the upper solution which is universally applicable to this class of WECS.

An efficient method for the real-time contingency analysis of meshed high-voltage direct current (HVDC) power grids fed by voltage source converter (VSC) stations is introduced here. A linearised AC/DC grid model is initially determined considering the control strategies of the various VSC units. This lays the foundations for the determination of linear sensitivity factors with which the contingency analysis is carried out to evaluate the real-time N−1 criterion in AC/DC grids, as demanded by system control centres. Distribution and power-injection factors are subsequently derived for efficiently assessing the impact of AC/DC transmission line outages and load/generator disconnections on the HVDC grid. Conversion factors are also derived to estimate the impact of the loss of VSC stations on the AC/DC network, this being another inherent contribution of this work. The efficiency and validity of this timely approach, which finds practical applicability to the real-time operation of HVDC power grids, is confirmed using a meshed DC network fed by three VSC stations. The disconnection of AC and DC transmission lines, generators, and VSC stations are dynamically simulated using Simulink and their post-disturbance steady-state conditions are compared against those computed by the introduced method where it is confirmed that both solutions concur very well with each other.

Due to current transformer (CT) saturation, the secondary current waveform may contain distortions which results in mal-operation of the protective system. This study presents a framework that attempts to detect and reconstruct saturated secondary current of the CT. To such aim, first an algorithm is proposed which detects and discriminates between normal and abnormal condition. This algorithm which is based on high-order derivatives and statistical criterion, classifies load change, inrush current and fault condition. In addition, this algorithm is designed so that it detects the first time that CT becomes saturated. After fault detection and phase fault selection, based on analytical and comprehensive formulations, the parameter of fault current signal is calculated in the sub-cycle. Moreover, the dynamic behaviour of the CT is considered in the waveform reconstruction. Several case studies are carried out to investigate the performance of the proposed method. Simulation results show that proposed method can effectively classify the inrush current and fault signal even in the CT saturation condition. Also, under different CT saturation condition, proposed method quickly and precisely detects and reconstructs saturated waveform without sacrificing details.

Condition monitoring data in the form of oil test results are important for transformer ageing assessment and asset management. Analysis of variance-based factorial analysis was demonstrated as a data mining technique for identifying the influence of different factors on transformer ageing. Using a UK oil test database of 33/11(or 6.6) kV transformers, studies were performed to assess the influence of in-service age, manufacturer, year of manufacture, load and environment on the measured oil acidity. This methodology can reduce the workload of manually performing pairwise comparisons to understand the influence of each factor. In addition, it can also consider any interaction effect present among the factors. As a demonstration, influence of the year of manufacture on oil acidity was revealed, which was confirmed by identifying a change in design for late 1960s units and an early degradation phenomenon for late 1980s units.

With the development of energy internet technology and smart power distribution & utilisation technology, multiparty interaction based on the complementarity of multi-energy demand has become an alternative solution to avoiding power shortage and improving comprehensive energy efficiency. The industrial integrated energy system with CCHP (combined cooling heating and power) as a generation source of heating and cooling energy is studied. Based on the traditional electrical demand response (DR) mechanism, the demand for electricity, heating and cooling is incorporated in the scope of generalised demand side resources. Considering the difference on price, demand and supply characteristics of multiple energy resources, the multi-energy-based DR scheme and the corresponding optimisation model are established to minimise dispatching expenses and improve the interaction between electricity companies, CCHP and industrial consumers. The numerical analysis shows that the proposed scheme could effectively engage CCHPs and consumers in multi-energy interaction, and the overall expenses can be significantly reduced.

The existing technique of fault state recognition for adaptive single-phase reclosure is severely impacted by the transient beat frequency oscillations after the secondary arc being extinguished in a shunt-reactance compensated line, due to the resonance between the equivalent capacitors and compensating inductors. The characters of fault phase terminal voltage and neutral small reactance voltage are analysed, and the differences of fault phase terminal voltage between the two fault states at recovery voltage stage are compared. Considering that there are some misjudged zones caused by beat frequency oscillation on the fault phase voltage, the neutral small reactance voltage is taken as a base value, and the ratio which uses the difference of calculated and measured on fault phase terminal voltage to neutral small reactance voltage is detected through a beat frequency cycle. Only when all ratios are >1 at the sampling points, a permanent fault is identified. If the ratio of a sampling point is <1, the transient fault is identified, and the reclosing command is issued after a determined delay. The simulation results show that the algorithm can distinguish the fault state under different compensation degrees effectively, and be free of different loads, fault locations and fault resistances.

One of the challenging issues in a restructured power system is fair allocation of transmission losses to its users. This study proposes two new algorithms to trace the active power produced (consumed) by the generators (loads) for assigning active losses to the generators or loads separately. In the proposed methods, there is no need to convert the study system to a system for which the active power at both ends of each transmission line are the same. In these algorithms, based on AC power flow calculations, new matrices called outflow-line and inflow-line are first created. Then, according to the properties of these matrices, four other matrices, which express active power flows of the transmission lines in terms of active power productions (consumptions) of the generators (loads), are then derived. These matrices are employed not only to allocate the active loss of every transmission line to each generator or load, but also to determine the contribution of each generator to the active power consumption of each load. The results obtained by implementing the proposed algorithms on a five-bus system and two IEEE standard systems indicate its superiority over some earlier methods.

Distribution networks are evolving into active meshed networks with bidirectional power flow as the penetration of distributed generation (DG) sources is increasing. This necessitates the use of directional relaying schemes in these emerging active distribution networks. However, conventional directional overcurrent (OC) protection will not be adequate to protect these networks against the stochastic nature of renewable DGs and the changing network architectures. Hence, this study proposes an adaptive directional OC relay algorithm that determines optimal protection settings according to varying fault currents and paths induced by the DGs in active meshed distribution networks. The proposed algorithm consists of a two-phase approach that deduces: (i) optimal floating current settings through a fuzzy decision-making module, and (ii) optimal floating time settings through an optimisation algorithm. Extensive case studies are implemented on the modified power distribution networks of IEEE 14-bus and IEEE 30-bus by varying the type, location, and size of DGs. The results validate the ability of the proposed protection scheme to capture the uncertainties of the DGs and determine optimal protection settings, while ensuring minimal operating time.

For the reliability assessment of a distribution system with microgrids (MGs), considerable computation time has to be cost to ensure convergence of an applied Monte-Carlo method (MCM) to estimating indices of interest, in the context of adopting multi-state models of constituent components. Towards achieving an efficient and accurate reliability assessment as much as possible, multi-state models of distributed generation resources, especially the battery energy storage system, are proposed. Based on the proposed models, effect analysis of a given sampled MG state is conducted in a proposed analytical manner. Moreover, an accelerated MCM nesting the proposed effect analysis methodology is also presented, dedicated to reliability evaluation of a resulting multi-state distribution system from the multi-state constituent components. Case studies on a modified IEEE-RBTS Bus6 F4 system demonstrate the effectiveness of the proposed method, against the plain MCM. In addition, impact of an important parameter, termed total state number of state-of-charge, on the convergence and speed of the proposed method has been fully analysed.

This study proposes an energy management framework for cooperative operation of multiple coupled microgrids, in electricity market environment. The framework focuses on a scalable profit maximisation approach across the entire smart grid for coordinated control and management of the coalition forming microgrids and the utility. The optimal control problem only uses the power exchange schedules as control signals between microgrids. Uncertainties pertaining to renewable power generation and load demand are described via scenarios. In this stochastic framework, conditional value-at-risk is considered as a risk measure to lessen the danger to which the aggregator is exposed to because of fluctuating power transactions induced by such uncertainties. The framework is further extended to evaluate the impact of cooperative operation on the portfolio returns of participating microgrids over the project lifetime. Several case studies are simulated in-order to validate the proposed risk-constrained framework. The results obtained clearly show that the proposed framework is effective in limiting the risk of the profit variability of the microgrids, at the cost of a small reduction in their expected profit.

For the mixed-voltage quadruple-circuit transmission lines on the same tower, its special construction form makes the cross-voltage faults, i.e. multi-circuit faults that occur between different voltage level systems, easy to occur. Although the phase selector based on phase-to-phase superimposed currents (PTPSC) shows excellent performance in the single transmission line, it cannot guarantee to identify faulted phases when confronting cross-voltage faults. In this study, the adaptability of phase selectors based on PTPSC is analysed and a new method of phase selection is put forward. First, the PTPSC measured by the relay can be obtained on the basis of cross-voltage short-circuit calculation. Then, the influencing factors and possible selection results are comprehensively analysed, and the corresponding measure has been proposed. Finally, a fault model of mixed-voltage quadruple-circuit transmission lines is constructed in PSCAD and simulation results agree well with the theoretical analysis, verifying the validity of this method.

Power system stabilisers (PSSs) are supplementary controllers connected to the excitation system of synchronous generators to damp electromechanical oscillations. Multi-band PSSs are reported as advanced PSSs with the ability to damp out all oscillation modes present in the power systems. This study presents the design of a robust fractional-order multi-band power system stabiliser (Fo-MBPSS) using a meta-heuristic hybrid algorithm for dynamic stability improvement of multi-machine power systems. The large bandwidth, memory effect and flat phase contribution in the frequency response of fractional-order controllers are exploited to make the Fo-MBPSS perform well against a wide range of system uncertainties. The parameter tuning problem of Fo-MBPSS is transformed to an optimisation problem that is solved using a hybrid algorithm by combining a dynamic genetic algorithm (DGA) with a standard particle swarm optimisation (PSO) algorithm. The performance of the proposed DGA-PSO-Fo-MBPSS is evaluated through eigenvalue analysis, non-linear time-domain simulations and some performance indices, in two different multi-machine systems under different loading conditions and disturbances. The results are compared with PSO-based conventional MBPSS and PSO based Fo-MBPSS (PSO-Fo-MBPSS) to establish the fractional parameter effect on the improvement of the system dynamic response and the relevance of the proposed hybrid optimisation technique in achieving robustness.

A distribution static compensator (DSTATCOM) is implemented at distribution level for overcoming several power quality problems commonly encountered today. Some of these problems considered here deal with issues of harmonics, unbalancing of loads, meeting reactive power demand of the load for improving power factor to unity and regulating the voltage. A new technique is proposed based on affine projection control algorithm and its design and implementation on a shunt compensator are discussed. The proposed control approach estimates the fundamental weight values of load currents. These weights are utilised to compute the magnitude and phase of reference grid currents and consequently gating pulses for three-phase voltage source converter (VSC). The control approach is based on the convergence of the weighted values and is independent of the property of the input signal. This control approach has characteristic of updating weights on the basis of multiple, delayed input signal vectors. Fast convergence of active and reactive weighted values is added advantage of the proposed control approach with the above-mentioned characteristics. A working prototype of the DSTATCOM is implemented using three-phase VSC and digital signal processor (DSP) (dSPACE 1104). The affine projection control approach is developed in MATLAB/SIMULINK and verified on a prototype using DSP.

Day-ahead peak load forecasting is a crucial factor for the electrical utility for daily power planning and distribution. A comparative study of the peak load forecasting at distribution feeder-level circuits with high penetration level of PV generation are investigated. First, two years of private load data from the local utility and two years of public load data from Texas utility are analysed. The correlation analysis is applied to peak load and its driver factors. Next, Bayesian additive regression trees (BART) is employed to do peak load forecasting. Since the BART method takes an amount of time to generate the forecasting, the composite kernel methods based on Gaussian process regression (CKGPR) are designed. Then these methods are compared with multiple linear regression method and the support vector regression method based on the residential area and the business area load data. Thorough comparison results are presented based on five forecasting measurements. The BART has the best forecasting accuracy among all the indices, and the CKGPR also has counterpart forecasting results but with less computation time. Meanwhile, the forecasting accuracy difference between two areas is analysed. Lastly, influential driver factors are summarised.

This study evaluates a number of uncertain parameters that affect the accuracy of distribution system state estimation (SE), and ranks their importance using an efficient sensitivity analysis technique, Morris screening method. The influence of the uncertain parameters on SE performance is analysed globally and zonally. Furthermore, the dependence structure between the critical variable and SE accuracy is analysed using copula to establish their relationship at different section of the bivariate space. The sensitivity of the critical parameter at different ranges is also studied and ranked using Morris screening methods to present the variation of SE performance when the critical variable is allocated at different sections within the feasible range. Accurate assessment of the importance of various uncertain parameters and the analysis of the dependence structure can inform power system operators which parameters will require the greatest levels of mitigation or increased monitoring accuracy in order to have satisfactory performance of distribution system SE.

Owing to mixed-integer and non-linear properties, the distribution network reconfiguration (DNRC) problem has been widely addressed with meta-heuristic algorithms. To accelerate the solution process, two essential components of meta-heuristic algorithms are investigated in this study: solution representation and fitness evaluation. Instead of the popular binary and integer numbers, decimal encoding is employed. Decoding is based on the proposed probability-based loop destruction strategy. The fitness evaluation is based on the power flow calculation of radial network. Different from backward/forward sweep method, the advantageous direct solution technique is utilised, where the matrix generation process has been accelerated. Both improvements are based on the graph theory and fully explained with illustrative examples. Case studies are implemented on five benchmark systems. The superiority of the proposed methods over their advanced counterparts has been established with intensive comparisons. Finally, these methods are integrated into a standard particle swarm optimisation framework for the solution of DNRC. Results indicate that the proposals significantly improve the solution efficiency without the loss of quality.

Demand baseline estimation (BE) is key to the impact assessment of a demand response event in a power system. While many BE techniques exist in literature and are implemented by utilities, these are either inaccurate, or computationally intensive, and only provide point estimates of the demand baseline. This study presents a simple, single-stage, similarity-based BE technique. The authors posit a new definition of similarity that includes weather covariates, and therefore eliminate the need for a subsequent adjustment. A novel growth rate assumption for the demand, combined with an optimised exponential smoothing technique results in a higher accuracy for the proposed BE technique. Additionally, an L-order iterated bootstrap is used to generate confidence intervals to account for prediction uncertainties. The proposed BE technique is tested for the Singaporean National Electricity Market, and is shown to be consistently more accurate than other conventional BE techniques.

Quick fault detection and isolation of faulty section are desired in DC microgrid due to the presence of power electronic converters and low cable impedances. Owing to need of fast disconnection, limited time and data are available for online fault distance estimation. Some of the existing techniques consider source capacitors connected at only one end of the cable; therefore, assume that the fault current is contributed by only one end of the cable. This may not be true in the case of multi-source DC microgrids, where fault current would be supplied from both the ends. Further, existing communication-based techniques require either data synchronisation or fast communication network. To address these issues, this study proposes an online fault location method for multi-source DC microgrid without using communication. The mathematical model of faulted cable section connected to sources at both the ends is derived. This model is used along with the measurements to determine the fault distance. The model consistency with the measurements is quantified using the confidence level based on the residual analysis. A ring-type multi-source DC microgrid system is considered and simulated on real-time digital simulator to demonstrate the effectiveness of the proposed algorithm.

Temperature is a key factor for failure analysis of power transformers. Conventionally, transformer failure rate is calculated with hot spot temperature induced from IEEE empirical equations. This article firstly introduces a spatial and temporal related failure model based on three-dimensional thermal simulations of transformer and the related environment. The proposed thermal model is established with computational fluid dynamics for ventilation calculation and heat generation equations for power device simulation. Cooling strategies and mutual heating effect of power equipment are considered for an accurate temperature distribution prediction. By incorporating the three-dimensional thermal model into the service life-dependent and temperature-dependent model, the failure rate of each spatial point in power transformer could be calculated according to Arrhenius theory and Weibull distribution. The simulation results show that the proposed model clearly improves the accuracy of failure analysis and can be used for thermal and ventilation design of transformer room.

Detection of shorted DC faults and high-resistance faults on ring type low-voltage DC (LVDC) micro-grids imposes an elusive challenge. This research work proposes an efficient and reliable protection scheme for DC system in order to boost up LVDC system adoption on large scale, despite lack of proper standards and mature experience in DC system protection. In order to ameliorate these deficiencies, a wavelet transform-based fast fault protection scheme using local measurements has been proposed for eliminating threat to voltage source converters during fault. The features extracted using wavelet coefficients are utilised as key indicators to identify the fault with wide variations in operating conditions. The LVDC micro-grid system has been simulated using real-time digital simulator and the test results studied show a great potential ability in performing DC fault detection and classification task efficiently. Furthermore, the robustness of the proposed protection scheme is capable of fault line identification and setting up desired relay selectivity in the LVDC distribution system.

With the traditional vertically integrated electric market transforming into a liberalised one, the generation side consisting of various generation companies (GENCOs) is separated from the grid. This transformation gives opportunity for GENCOs to enhance their own profits by expanding business. One example is to establish bilateral contracts directly with the end users. This study considers a thermal-wind GENCO trading with the grid in the pool-based electricity energy market and meanwhile signs bilateral contracts with a battery swapping station (BSS). The operation framework for this GENCO to maximise its profit under uncertain information is proposed. A two-stage stochastic model is adopted to formulate the profit of the GENCO with the uncertain wind power, electricity market price and stochastic BSS electricity demand. The stochastic problem is transformed into deterministic mixed-integer linear programme and solved by CPLEX. The impact of bilateral contract capacity, contract price, degree of uncertainties and the wind power penetration on the final profit is analysed. This work provides a novel operation framework for GENCOs, and provides feasible bidding strategy and scheduling under uncertain conditions, especially for GENCOs owning renewable energy.

To make an appropriate compromise between the solving speed and solution quality for a distribution network with distributed generations (DGs), an efficient reconfiguration algorithm based on vector shift operation (VSO) considering DGs is proposed. The nodal equivalent power model considering the influence of DGs is established. Power moment is selected as an index to decide whether load flow calculation or VSO should be conducted. In the former case, all bus voltages are updated and then be regarded as the initial condition for VSO. In the latter case, the power and resistance vectors are first formulated based on the structure and parameters of the distribution system, and the variation of power loss after branch exchange is then calculated based on these vectors instead of the time-consuming load flow calculation. The power and resistance vectors are updated by simple element shift operation after each step of branch exchanging. The final optimal solution is obtained when the open-branch set remains unchanged after branch exchanging. The proposed VSO-based reconfiguration algorithm is applied to a practical distribution system in Taiwan. The influences of the DGs and network reconfiguration on power loss and nodal voltages are illustrated. Test results show that the proposed reconfiguration algorithm is feasible and efficient.

Concerning multi-source oscillation scenes, an oscillation centre identification method based on the variation trend of bus frequency is proposed. First, according to the amplitude and phase relationship between voltage and current at any point in the system, the functional expression of the frequency of voltage and current is derived. Then, the variation trend and distribution of the frequency of voltage and current in multi-source oscillation scenes are analysed. On this basis, the relationship between system frequency distribution and the oscillation centre and desynchronising centre is revealed, and an oscillation centre identification scheme for multi-source oscillation scenes is put forward. Simulation results based on DIgSILENT/PowerFactory demonstrate that the proposed scheme can effectively identify the location of oscillation centre in multi-source oscillation system.

Here, it is presented a novel Cuckoo Search (CS) algorithm called Cuckoo-GRN (Cuckoo Search with Genetically Replaced Nests), which combines the benefits of genetic algorithm (GA) into the CS algorithm. The proposed method handles the abandoned nests from CS more efficiently by genetically replacing them, significantly improving the performance of the algorithm by establishing optimal balance between diversification and intensification. The algorithm is used for the optimal location and size of distributed generation units in a distribution system, in order to minimise active power losses while improving system voltage stability and voltage profile. The allocation of single and multiple distribution generation units is considered. The proposed algorithm is extensively tested in mathematical benchmark functions as well as in the 33-bus and 119-bus distribution systems. Simulation results show that Cuckoo-GRN can lead to a substantial performance improvement over the original CS algorithm and others techniques currently known in literature, regarding not only the convergence but also the solution accuracy.

Conventional distribution network departs to the smart grid. The parking lot will have an important role in the smart grid as a distributed generation. Due to the output power of parking lots is uncertain, More accurate modeling of parking lot output power is necessary for the future of distribution network studies such as Distribution Network Expansion Planning (DNEP). In this paper, a systematic method based on the Z-number concept is utilized to represent the uncertainty of Vehicle to Grid's (V2G's) presence. In order to investigate the impact of V2Gs uncertainty on the DNEP, we proposed a Probabilistic–Possibilistic DNEP in the presence of V2Gs referred to as P-PDNEPV2G. If the V2Gs historical data is incomplete, the proposed structure can significantly consider the effects of V2G on the DNEP. In P-PDNEPV2G, parking lots output power is described as a probabilistic–possibilistic variable by Z-number method. The optimization of P-PDNEPV2G is executed by the Non-Dominated Sorting Genetic Algorithm (NSGA-II). A 24-bus test system and the real 20 kV distribution network of Ghale-Ganj city of Kerman province in Iran are used to demonstrate the effectiveness of the proposed methodology. Eventually, several analyses are conducted to investigate the impact of probabilistic–possibilistic V2G model on the DNEP problem.

The emerging complex circumstances caused by the essence of the new economic normal and the low-carbon development of power industry lead to the demand for electricity from the rapid growth phase to a ‘new normal’ pace for China. To better capture the more and more non-linear and non-stationary characteristics of electricity demand, Tapio decoupling model is applied to research the decoupling of electricity demand from economic growth, and T's correlation degree is used to measure the influencing degree and reveal the role direction of each driver on electricity demand. On these bases, a novel improved particle swarm optimisation-extreme learning machine (IPSO-ELM) hybrid forecasting model is proposed to predict electricity demand. In IPSO-ELM, a mutation operator of genetic algorithm is introduced into standard PSO to ensure the overall convergence and enhance the accuracy of convergence, and the improved PSO is adopted to optimise the input weights and hidden bias of traditional ELM which are randomly generated. Finally, the case study of China demonstrated the efficacy and feasibility of IPSO-ELM. Simultaneously, the comparisons with ELM, back propagation neural network, support vector machine, GM (1, 1), and logistic model showed that IPSO-ELM has a better forecasting performance in electricity demand.