System integrity protection schemes (SIPSs) are installed at substations to preserve power system stability by preventing bottlenecks in transmission and distribution networks and improving the overall reliability. SIPSs are also referred to as remedial action schemes, emergency control systems, special protection systems, and wide-area control schemes. Typical SIPSs have centralised computing architectures with applications to ensure reliable power system responses during transmission outages, remediate regional transmission bottlenecks caused by delays in the construction of new lines, prevent severe instability and blackouts due to faults and inadvertent disconnections, prevent line damage resulting from thermal limits, avoid voltage collapse, and protect against other critical events in the power system. This study proposes a distributed computing architecture for SIPSs, including algorithms that use DC and AC optimal power flow. The authors classify SIPS applications based on control types and propose novel algorithms for next-generation SIPSs using synchrophasors and real-time technology. The authors also propose testing requirements and architectures for validating SIPSs prior to field installation. Lastly, the authors share results obtained using the proposed SIPS for a transmission overload condition caused by excessive wind generation.

The ever-increasing power demand coupled with limited power transfer corridors and minimum redundancy have forced power system operation close to its security limits. Further, increase in renewable penetration, market deregulation, and mal-operation of protective devices under stressed conditions has imposed serious threats to power system stability. To overcome these operational challenges, a new response-based system integrity protection scheme (SIPS) is proposed here. The proposed scheme works towards preserving system integrity in two ways. First, an apparent impedance-based relay security index (RSI) is introduced to prevent the zone 3 mal-operations of distance relays under stressed situations and achieve accelerated zone 3 operation during faults. Second, a system stability index (SSI) computed from the generator rotor angles is forecasted using a predictive analytic to foresee system instability, if any, and take corrective actions in real time. The scheme can be used for area-based monitoring and control of system stability, does not require coherency group identification, remains unaffected by changes in network topology, and can effectively function for first swing stable contingencies. The scheme is tested on IEEE 39-bus New England system and the results obtained validate the reliability of scheme in maintaining system integrity.

Transmission lines are a vital part of power systems and are prone to a variety of short-circuit faults. It is imperative to identify the fault location to restore the system as quickly as possible for maintaining the system integrity. Impedance-based fault location algorithms are most commonly used but they are prone to several sources of error such as fault resistance and mutual coupling between the lines and may not provide accurate results in the presence of complex network configurations such as three-terminal lines. Though information about mutual coupling or knowledge of complex network configuration is known, they cannot be directly used in impedance-based fault location algorithms. This study proposes a novel automated generic framework for model-based fault location approach which uses the power system model to overcome the shortcomings of impedance-based fault location methods. A combination of power system simulation software and a control and processing software provide the flexibility to be able to incorporate all the available information about the system to accurately estimate the fault location. Furthermore, this study provides key insights to implement the proposed framework in an effective and efficient manner by using reduced equivalent circuits.

System Integrity Protection Scheme (SIPS) aids the normal protection scheme in case of certain special contingencies like tripping of important corridors/flow gates etc. to mitigate the voltage or angular instability, load generation imbalance and finally large-scale cascading catastrophes. A recent blackout incident in an ultra-mega-power project (UMPP) in India has motivated the authors to ponder upon a SIPS as a solution. As part of designing the SIPS, an efficient solution is formulated using synchrophasors for a well-known issue nonetheless an issue which still disturbs the Indian power system network with blackouts – i.e. unintended distance relay operation in the event of power swing. Here, real-time voltage and current phasors and the angular separation between the buses connecting faulted line, acquired by phasor measurement unit is utilised to derive a differential power coefficient (DPC). This study proposes a modified logic scheme for blocking/de-blocking of distance relays i.e. change the existing scheme with ‘block all zones except Zone-1’ to ‘block all zones on power swing detection’. Rate of change of DPC is derived here which aids in differentiating stable and unstable power swings. A special protection scheme is finally formulated based on the modified logic.

This paper proposes a novel method to locate faults in an AC-meshed microgrid. To this end, a set of features is first extracted and selected from the measured signals and fed to a Support Vector Machine (SVM) to detect the occurrence of fault. Then, the Distributed Generator (DG) with the lowest amount of fundamental voltage, which is the closest one to the fault, injects an appropriate voltage/current harmonic. As the faulted section has the lowest impedance value from the Point of Common Coupling of the DG, the harmonic current of the corresponding line has the highest value. Based on this fact, the first candidate DG sends a notification signal to the second candidate DG, in which the fault occurs between them. Finally, the impedances in the injected frequency are measured from these two DGs and fed into a multi-class SVM to locate the faulted line. The proposed method has the ability to locate faults for islanded and grid-connected microgrids with variable configurations. Real-time simulation results are taken by OPAL-RT to show the effectiveness of the proposed method in the meshed microgrid.

The study presents a new protection scheme to enhance the zone-3 operation of the distance relay. The proposed scheme develops an index based on the rate of change of active power and angle derived from synchronised phasor measurement units (PMUs). It is found that it clearly distinguishes faults from several stressed conditions those are likely to occur, and takes necessary action. The proposed scheme is developed and validated for a modified Western System Coordinating Council 9-bus, 3-machine system and New England 39-bus system with different fault scenarios on a MATLAB/Simulink platform. Further to test the robustness, the proposed method is tested on a real-time digital simulator platform using phasors received from gigabit transceiver network PMUs. Extensive performance testing for faulted as well as stressed conditions such as stable power swing, load encroachment and voltage stressed conditions indicate that the proposed scheme is highly reliable in enhancing the zone-3 operation of distance relay.

Prediction of the transient stability in real-time can be a boon in preventing the occurrence of catastrophic events in a particular area that could potentially affect the entire grid and cause blackouts. Synchronised voltage angle measurements procured from the installed phasor measurement units in the network have been used to devise an index for assessment of transient stability in the system. The behaviour of the trend of the acceleration at the instant of the critical clearing angle has been used in this study as the basis for differentiating between a stable and an unstable power swing. Upon detection of an unstable condition, the generator, which is most likely to be going out of step is detected, and appropriate control action is suggested for mitigating the predicted out of step condition. Moreover, the distance relay maloperations caused during a stable power swing can also be a reason for triggering cascading outages. Suitable remedial measures to avoid such maloperations of the distance relay have been proposed. The efficacy of the proposed philosophy has been tested on a single machine infinite bus system, WSCC-9 bus system and New England 39 bus system modelled in Power Factory/DIgSILENT.

Discovering early signatures of voltage instability can enhance situational awareness and facilitate emergency control. Such signatures usually appear as decaying trends in load voltages and increase in reactive power output of generators. Discrete control actions such as frequent switching of load tap changer (LTC) and capacitor banks, that are otherwise unmonitored, can also be crucial indicators of a voltage problem. Hence, the authors propose a unified early warning scheme (EWS) for detecting voltage instability. To quantify and correlate spatial, temporal (spatio-temporal) trends and level changes, the authors use a multivariate, adaptive trend and level (T&L) filter. A unique feature of the proposed scheme is the signature authentication, by verifying simultaneity in T&L changes across multivariate time series. The authors show that, by extrapolating trends, limit violations in bus voltages and generator reactive power output can be estimated well in advance. Further, the level-change detection logic is shown to have the ability to detect frequent LTC-actions directly from the bus voltage time series. While the T&L filtering scheme is validated on field phasor measurement unit (PMU) datasets, EWS and emergency control is demonstrated on the classical 10-bus voltage stability test system.

A modified masking empirical mode decomposition (EMD) for effectively improving the mode separation capability of standard EMD while analysing power system signals with closely spaced modes of frequency is proposed here. A multi-frequency test signal resembling the dynamic power system signal and the actual data recorded by wide area measurement systems (WAMS) from Northern Grid, India, on 1 June 2010 due to generation loss are analysed by both the existing method and the proposed method. Various components of modal frequency extracted by masking signal base EMD are compared with extracted components by the proposed modified masking EMD. Further, the time–frequency representation of extracted modes by Hilbert spectral analysis is implemented to know mode behaviour with time reference. Simulation results prove that the EMD with proposed modification is capable of separating various modes of frequencies present in the WAMS signal.

The emergence of cyber-physical attacks brings a key challenge to the existing system integrity protection schemes (SIPSs) of smart grids. As one of typical cyber-physical attacks, false data injection attacks (FDIAs) can bypass the existing Kalman filter-based -detector detection techniques in SIPSs. To improve the detection performance against the FDIAs in SIPSs, this study proposes an unknown input observer (UIO)-based detection and isolation method. Taking the stealthy characteristics of FDIAs into account, this study presents a set of UIOs to detect the FDIA based on the internally physical dynamics. Furthermore, a UIO-based detection and isolation algorithm against the FDIAs is proposed based on the feature of residuals generated by UIOs. To detect the cyber attacks quickly and avoid missing detection, an adaptive threshold is designed to replace the precomputed threshold by taking the model linearised error and disturbance into account. Finally, comprehensive simulation results on the proposed algorithm are carried out, and the effectiveness of improving the detection performance in SIPSs is verified.

Electric energy consumption is progressively increased requiring new investments in the network to meet. Demand response programs (DRPs) are of practical tools to defer and/or decrease these investments. In this study, a real-time DRP is proposed to promote consumer participation in the energy delivery. The proposed DRP is designed on behalf of the Load Serving Entity (LSE) and aims to maximise its profit. The LSE purchases energy from the up-stream wholesale market and offers to the consumers in the retail market. The consumers, with the objective of maximising the profit, have the choice to purchase energy from the retail market or the LSE. The problem is formulated as a mixed integer linear bi-level programming model wherein LSE and aggregators, on behalf of the consumers, act as the leader and followers of the problem. Also, to avoid creating a new peak profile in light load periods promoted by the designed LSE prices, the upper level problem is tailored as a multi-objective formulation wherein flattening load profile in addition to the maximising LSE profit is considered. The proposed model is implemented on a test case and the results show that the both LSE and consumers benefit is increased with a smoother load profile.

This study proposes a non-linear stabiliser to suppress the current/voltage fluctuations of DC micro-grids (DC-MGs) caused by large transients. These transients typically occur following a reconfiguration of DC-MGs, such as disconnection/reconnection of electrical sources, or rearrangement in the structure of DC-MGs. In the DC-MG under consideration, a number of the DC sources are implemented in the form of hybrid power conversion systems (HPCSs) consisting of a parallel combination of a super-capacitor (SC), a fuel cell and a photovoltaic (PV) system. Due to the fast dynamics of the SC units, the stabilisation function of the DC-MG during transients is performed by these units using a supplementary signal provided by the stabiliser. The proposed stabiliser is intended to operate in a decentralised manner. To this end, a novel Lyapunov function is proposed to adjust the stabiliser parameters based on local data relevant to each HPCS. The control scheme employed is based on a simple structure facilitating the implementation of the proposed stabiliser. Finally, time-domain simulations are carried out demonstrating the effectiveness of the proposed control framework in a multisource DC-MG.

In this study, an interval arithmetic-based method has been proposed for contingency analysis of line outage in DC distribution networks. This method incorporates the uncertainties in source/load which become considerable in the system with increasing renewable generations and time-varying loads. Two-stage contingency analysis approach is adopted in this method. For the first stage of contingency screening, a contingency filter has been developed based on interval matrix norm to quickly estimate the bounds of operating state in the post-outage period. For the second stage of detailed analysis to the selected contingencies, the interval analytical model is first transformed into a non-linear programming-based equivalent model. To solve this model accurately and efficiently, successive linear programming and reformulation-linearisation techniques are employed. The proposed method is benchmarked against Monte Carlo simulations and line outage distribution factor-based method.

This study presents an approach to calculate average technical losses (TLoss) and non-technical losses (NTLoss) in distribution grids using an equivalent operational impedance (EOI) calculated from a load flow solution for the transformers’ average loads, obtained from customers’ electric bills. If transformers are supervised, i.e. have available active and reactive power injections measurements or estimates from a state estimator program, the transformers’ average technical and non-technical losses can be calculated by running a load flow for the grid supplied by the transformer and using the definition of EOI presented in this study. For unsupervised transformers, i.e. no measurements or estimates are available, pseudo measurements of average active and reactive power injections can be obtained by a procedure that uses active and reactive power measurements at the feeder substation. The use of this methodology by electric utilities allows the calculation of technical and non-technical losses more accurately, as well as the inclusion of non-technical losses in the distribution networks operation planning. Results using the IEEE 13-bus test system and an urban real distribution grid are presented to demonstrate the effectiveness of the proposed methodology.

This study aims to monitor the temperature inside power cable joint, with strong robustness to variable thermal environments and uncertain thermal parameters of the joint. The model consists of two sequential steps, radial-direction temperature inversion in the cable and axial-direction temperature inversion in the conductor. The former part was improved by coating the cable with a heat insulated layer, whereas the latter one was optimised through a proper parameter selection. Afterwards, an outdoor cable joint temperature-rise test was carried out and the temperature obtained from the proposed approach agrees well with the measured one. With the great change in ambient temperature and wind speed during the test, the model accuracy remains almost constant, and this is due to the heat insulated layer and low sensitivity coefficient of the algorithm. Besides, if the thermal conductivity and volumetric specific heat of waterproof compound in cable joint are assumed to increase by nearly five times, the maximum changes in temperature monitoring are only 1.4 and 0.4°C, respectively. The results indicate that this approach can effectively overcome the effects of variable thermal environments and thermal parameter dispersion in joint, thus promoting practical applications in temperature monitoring for cable joints.

The purpose of optimal power flow (OPF) is to optimise an objective function subject to a set of operating constraints. It remains an active research area because of using the bus-wise power balance equations, which leads to a non-linear solution space, resulting in OPF yielding local optimal solutions, thereby causing significant economic loss. In this study, first, a new line-wise OPF (LWOPF) formulation is proposed. Thereafter, a maximum loadability factor, as a voltage collapse indicator, is derived and combined with LWOPF constraints to form a voltage stability constrained LWOPF (VSCLWOPF) model. As the line-wise power balance equations are based upon the square of voltage magnitudes, it results in significant improvement in the solution space and lower-order terms in all computational steps. The LWOPF and VSCLWOPF formulations, are solved using non-linear optimisation technique, tested on several benchmark and real power systems. Results show that the proposed LWOPF is accurate, provides monotonic convergence, and scales well for large systems. It provides a better solution and is consistently faster, up to twice the speed of MATPOWER, due to reduced computational needs. Results of VSCLWOPF show that, for the same voltage stability level, the solution costs less than that obtained by classical bus-wise OPF.

One way to represent the physical behaviour of the transformer winding under high frequencies is to construct an equivalent circuit using modernised methods. This study represents a luxurious and time-efficient computation method based on particle swarm optimisation with simulated annealing (PSO-SA) method to construct an equivalent circuit of a transformer winding, from its frequency response analysis. The proposed method uses a simulated annealing method as a local search tool to improve the convergence and the characteristics of particle swarm optimisation in fast calculation. Furthermore, a comparative study of PSO-SA, crow search algorithm and genetic algorithm method is made. Different case studies of a transformer winding with an air core are considered. According to the results, the PSO-SA has higher efficiency and faster convergence.

Volatile impact of intermittent renewable energy sources (RESs) on the one hand and the uncertainties of loads and market prices, on the other hand, make the bidding strategy of microgrids (MGs) too risky and high-computational problem. To cope with these challenges, the bidding problem of MGs based on a three-stage hybrid stochastic/interval optimisation (HSIO) is devised in this study, which provides a trade-off between covering the volatilities by means of the MG potential flexibilities resources or by means of the energy provision from the real-time market (RTM). To tackle the uncertainties of the day-ahead market prices, the cost-effective stochastic programming (SP) is applied to maximise the profit of MG in the day-ahead stage of decision-making. In order to handle the volatilities of RESs production and uncertainties of RTM prices, a flexibility scheme based on the robust and low-computational interval optimisation (IO) approach is designed to minimise the balancing cost of MG in the real-time stages. Comprehensive numerical results are provided to compare the effectiveness, robustness, and computational complexity of the proposed model. Results show that the HSIO model takes advantage of the cost-effective solution from the SP model, and the robust solution with computational simplicity from the IO model.

Vehicle-for-grid (VfG) is introduced as a mobile energy storage system (ESS) in this study and its applications are investigated. Herein, VfG is referred to a specific electric vehicle merely utilised by the system operator to provide vehicle-to-grid (V2G) and grid-to-vehicle (G2V) services. The advantages of VfGs over the ESSs and plug-in electric vehicles (PEVs) include mobility of the VfGs across the distribution system and their complete availability for the system operator, respectively. In this study, VfGs are utilised by the distribution company (DISCO) to minimise the daily operation cost of the distribution system by providing the V2G and G2V services in optimal buses of the feeders. In addition, VfGs are applied by the generation company (GENCO) to minimise the daily operation cost of the generation system by providing the V2G and G2V services at optimal time periods. It is demonstrated that optimal application of VfGs has a considerable potential for cost reduction for both DISCO and GENCO. In fact, the DISCO and GENCO are benefitted because of the minimisation of feeders’ power loss and deferring the expensive generation units, respectively. Additionally, it is proven that cooperation of GENCO and DISCOs in utilisation of the VfGs has more benefit for them.

In this study, an improved crow search algorithm (ICSA) is proposed for the optimal allocation of renewable energy sources (RESs) and distributed static compensator in a practical distribution network (PDN) to improve system voltage, reduce line losses, maximise economic benefit, and decrease pollutants’ emission. To improve the performances, the control parameters of ICSA are logically tuned to make them adaptive with the iteration. Voltage stability index is used to identify healthier and weaker buses (HBs and WBs) of the system and comparative performance analysis is conducted after placing the devices at the identified HBs and WBs. The proposed technique is applied for both the cases of reduced search space (only WBs) and full search space (all the load buses) to establish that the optimal placement is always obtained at the WBs of the network. The optimal allocation for each type of device is conducted considering the time-varying characteristics of load demand and RES's power output, and their optimal performances are analysed in details, which will be helpful for the distribution companies to identify the most suitable device for incorporating in PDN. Furthermore, to show the effectiveness, the proposed algorithm is also compared with other algorithms.

This study focuses on the equitable loss allocation method for radial distribution networks integrated with distributed generators (DGs). As the traditional Shapley value method may cause a combinational explosion problem, the authors propose a sampling method for estimating the actual Shapley value. They use a stratified sampling method (SSM) to reduce the number of samples of Shapley value method with the subject to the overall equilibrium constraint. To determine the number of samples drawn from each stratum, they use the Neyman optimum allocation to minimise the variance of the sample mean. A reinforcement learning algorithm is introduced to estimate the standard deviations of the strata needed for the optimal stratified sampling. The proposed method is applied to a modified 17-bus distribution network and an actual distribution network in Zhejiang Province, People's Republic of China. The simulation results show that the proposed method can not only resolve the combinational explosion problem of Shapley value method but retain its desirable characteristics. The proposed method can dramatically reduce computational time by implementing a SSM. Therefore, the efficiency and superiority of the proposed method with regards to loss allocation are verified for radial distribution networks that include DGs.

In the market, the aggregators offer various proposals for an electrical customer, who selects one among all of them. The proposals may include different numbers of levels of uncontracting capacity rates, demand and energy prices, measuring intervals and billing periods. To schedule the best contract between user and electricity supply provider for upcoming months, two new approaches are presented here to solve the multi-optional multi-level demand-contracting problem. First, the feasible region is curtailed to the certain edged points and then, the optimal proposal and contracting demand (CD) are obtained by either direct or indirect proposed method. Moreover, a robustness analysis on the solution due to having errors in the forecasted maximum demands and changing the generations of the connected energy resource (ER) or installing a new one is presented. Some useful stability indices are also proposed. Likewise, updating the optimal solution due to having such errors is presented. Various numerical tests are taken place to analyse the influence of each parameter of the problem on the optimum and compare the proposed techniques. The results highlight the efficient capability of the proposed methods in fast obtaining the optima and stability investigation on the solution.

This paper proposes an identification method of multiple harmonic sources in the presence of different inverter-based distributed generations (DG) such as photovoltaic (PV), wind turbine with doubly fed induction generator (DFIG), and microturbine (MT) at the point of common coupling (PCC). To classify the linear loads, non-linear loads (NL) and inverter-based DGs, K-nearest neighbours (KNN) classifier was employed. The features were extracted using empirical mode decomposition (EMD), just from voltage waveforms. To reduce the redundant data, dimension of features vector, and time, the Relief-F method was implemented on the extracted features. Utilising only the voltage waveform for extraction of features increases the speed of the process and decreases the number of measurement equipment. To verify the effectiveness of the proposed method, a number of scenarios such as different harmonic sources with various harmonic orders, load and DG levels were simulated. The results on two test systems showed that the proposed method has a high accuracy even in the presence of different inverter-based DGs. It can be used as an added tool for power-quality engineers and can be integrated into monitor instruments. Also, the speed and precision of this method make it suitable for real-time applications in the power-quality issues.

With the large-scale access to distributed generation and energy storage (ES), the rapid growth of DC loads, and the development of power electronics technology, the DC distribution network is becoming a new type of high-reliability power supply. Detection and switching control of islanding is important to ensure the stability and robustness of the DC distribution network. In this study, first, for a ‘hand-in-hand’ DC distribution network, an unintentional islanding detection method, which is a combination of the main criterion based on switch status and an auxiliary criterion based on no-current and difference-voltage criteria, is proposed. Second, an unintentional islanding control strategy based on the coordination of the switching strategy of ES controller and DC solid-state transformer is proposed. Finally, the proposed detection and switching control scheme is validated through simulations using the real-time digital simulator platform. Simulation results indicate that the proposed scheme for the DC distribution network can accurately identify an island, and effectively switch control mode to achieve stable islanding control.

Here, small-signal stability issue associated with multi-scale uncertainty excitation is investigated by using differential inclusion theory. Specifically, a polytopic linear differential inclusion (PLDI) model for large-scale uncertainty system is developed, in which a modified non-sequence Monte Carlo method is introduced to identify a series of time-variant operation states. Additionally, a simplified small-signal model of renewable energy resources (RES) is proposed, the outputs of RES are modelled as time-varying elements in PLDI model to reflect the uncertainty in the linearised matrix. The stability criterion for the stochastic time-varying system is mathematically deduced based on convex hull Lyapunov function (CHLF). The criterion is then utilised to design robust stabilisers for stochastic system. Simulation, utilising two-area four-machine system and New England 39-bus test system, demonstrates the benefits of the proposed model in describing system stochastic characteristics, designing additional controllers and reducing computational burden.

In this study, a voltage instability prediction algorithm is proposed based on Thevenin impedance matching theorem. This algorithm consists of three important parts: an adaptive network zoning algorithm, the Thevenin-based voltage instability risk predictor (VIRP) and a third-order extrapolation technique. The zoning algorithm, based on the concept of electrical distance, is proposed to divide the main network into several zones after the occurrence of any large event. Each zone consists of some load buses and generating units have the most influences on their voltage stability. The proposed VIRP is defined similarly to the Thevenin indicator multiplied with a contribution factor. This factor is calculated for each zone, based on the lack of the reactive power reserve in its generating units, to increase the predictor value, especially in the proximity of the instability. So, VIRP could reduce the instability detection time with respect to the Thevenin indicator. The further reduction in the instability detection time is obtained by the well known cubic spline extrapolation technique. This function is applied to the values of VIRP to predict the instant of the instability. The proposed algorithm is tested on the modified New England 39-bus test system supplied with the voltage-source converter-high-voltage dc link.

The conventional distribution network is undergoing structural changes and becoming an active grid due to the advent of smart grid technologies encompassing distributed energy resources (DERs), aggregated demand response and electric vehicles (EVs). This establishes a need for state estimation-based tools and real-time monitoring of the distribution grid to correctly apply active controls. Although such new tools may be vulnerable to cyber-attacks, cyber-security of distribution grid has not received enough attention. As smart distribution grid intensively relies on communication infrastructures, the authors assume in this study that an attacker can compromise the communication and successfully conduct attacks against crucial functions of the distribution management system, making the distribution system prone to instability boundaries for collapses. They formulate the attack detection problem in the distribution grid as a statistical learning problem and demonstrate a comprehensive benchmark of statistical learning methods on various IEEE distribution test systems. The proposed learning algorithms are tested using various attack scenarios which include distinct features of modern distribution grid such as integration of DERs and EVs. Furthermore, the interaction between transmission and distribution systems and its effect on the attack detection problem are investigated. Simulation results show attack detection is more challenging in the distribution grid.

The reactive loss generated from the geomagnetically induced current (GIC) flowing through transformers has a characteristic of simultaneity in the whole power grid, which will result in a lack of reactive power, bring about voltage fluctuation and even threaten the safe operation of power grid. To solve the voltage fluctuation problem generated by GIC, this study presents a solution to restrain the voltage fluctuation by placing grounding small resistances in transformer neutral points. Aimed at minimising the voltage offset and the configuration budget and taking account into the limit of resistance value, the generator's reactive power output and the constraint of adjustable-ratio of transformers, a multiple-objective optimisation strategy known as NSDEA based on the elitist non-dominated sorting genetic algorithm combined with the differential evolution algorithm is proposed in this study, whose feasibility has been demonstrated by applying it in the placement schemes of 750/330 kV Gansu Provincial power grid and the 750 kV UHV power grid of northwest region in China. Results indicate that the optimal configuration scheme can effectively restrain the voltage fluctuation caused by GIC with the voltage offset reduced obviously, thus helps to improve the resistance strength of the large-scale power grid to the risk of geomagnetic storm.