Abstract:In the context of climate change, the global generation ratio of renewable energy keeps growing, and the power system with an ultra-high proportion of renewable energy becomes the future blueprint, in which the low-capacity-factor wind and photovoltaic generation dominates (their electricity penetration exceeds 50%). However, the severe power generation fluctuation poses the dual risks of load loss and power abandonment, and the flexibility supply-demand balance becomes crucial and complicated. This paper proposes a new analytical framework for the flexibility supply-demand balance of the power system with an ultra-high proportion of renewable energy. The major structural and operational features of the power system with an ultra-high proportion of renewable energy are first analyzed. The basic principles and key challenges of flexibility supply-demand balance are then studied. Three key issues of concern are finally pointed out, including the evaluation and simulated calculation of flexibility supply-demand balance and the simulation scenario construction taking different meteorological processes into account. The corresponding solutions are provided as the probabilistic evaluation method combining spatial correlation modeling and chronological connection relationship of component status, the rolling long time-sequence simulation method considering the risk of insufficient flexibility, and the weather scenario construction method based on extreme value theory.

Abstract:The access of massive distributed generators, microgrids, energy storage devices, and other flexible resources to the distribution network makes a resilient distribution network possible. However, the potential of these resources in cooperatively resisting external shocks and restoring the normal operation of the power grid needs to be further explored. The deep spatial-temporal coupling of massive resources makes its structure and control more and more complex in the distribution network. Fluctuations on any side of the source-grid-load can trigger chain failures and affect the safe operation of the distribution network. For this purpose, this paper proposes a resilience research framework for the distribution network with massive resources considering uncertainty disturbances. The framework incorporates resilience to external hazards and suppression of internal fluctuations into the research scope. Firstly, this paper expands the connotation of distribution network resilience, summarizes the existing flexible resources of distribution networks, proposes the structure of a highly resilient distribution network with massive resources, and analyzes its characteristics. Secondly, structural resilience indexes and system resilience indexes are proposed from static topology and dynamic response perspectives respectively to realize quantitative analysis. Then, the uncertainty disturbance is divided into external shocks and internal fluctuations. And the research framework of distribution network resilience is proposed. Its data foundation, theoretical support and resilience enhancement technology, and application scenarios are sorted out and discussed in detail. Finally, an outlook on the highly resilient distribution network with massive resources is presented.

Abstract:There are many types of sources and loads in a regional integrated energy system, and there are correlation characteristics between different energy sources, between loads, and between energy sources and loads. Therefore, a planning method for regional integrated energy systems considering the correlation of sources and loads is proposed. Firstly, Nataf transform and Cholesky decomposition are used to improve Latin hypercube sampling, and sample matrices are obtained considering the correlation of wind power, photovoltaic and power, electricity, gas, cooling and heat loads. Then, the typical source and load scenarios of the regional integrated energy system are obtained by improving cluster centers. Secondly, a planning model of regional integrated energy systems considering the correlation of multiple heterogeneous sources and loads is established. The objective function of the model is to minimize the sum of investment, operation, external energy transaction cost and curtailment cost of wind power and photovoltaic power. Finally, a test example based on the combination of the IEEE 39-bus power grid system, Belgium 20-bus gas grid system, and 14-bus cooling/heat grid system is simulated to verify the necessity of considering the correlation of multiple heterogeneous sources and loads in the planning of the regional integrated energy system and the feasibility of the proposed method.

Abstract:The correlation between distributed renewable energy and load has a crucial impact on the distribution network planning. Firstly, based on the truncated R-Vine Copula model, the joint probability distribution model of multiple loads, distributed wind power and photovoltaic generation is established. Then, a two-stage stochastic expansion programming model of the distribution network and its corresponding chance-constrained programming model are proposed. In this two-stage model, the first stage is the investment decision model of the distribution network, in order to minimize the total cost of the distribution network to determine the appropriate expansion and upgrade scheme of feeders, substations and capacitors. The second stage is the operation model of the distribution network in multiple scenarios. The objective of this stage is to minimize the network loss and the maintenance cost of the distribution network, and the optimal operation scheme is determined with the safe operation of the distribution network as the boundary condition. The correlation and randomness of loads and distributed generators are introduced into the model by the scenario and its realized probability. Considering that the solution of the scenario-based stochastic optimization model and the chance-constrained models is difficult, a new bilinear Benders decomposition algorithm is proposed, which can effectively solve the stochastic optimization model containing a large number of second-order cone power flow equations. Finally, taking a 38-bus distribution network in Nantong of China as an example, the proposed model and algorithm are verified.

Abstract:As the number of electric vehicles (EVs) gradually increases, the integration of EVs to power grids brings the difficulties to the operation and control of power grids. At the same time, the new power system with renewable energy as the main body faces the challenges of power balance. Based on complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and bi-directional long short-term memory (BiLSTM) neural network, a multi-time-scale scheduling method for regional power grid considering EV flexibility and wind power accommodation is proposed. Firstly, the intrinsic mode functions at different frequencies are obtained by CEEMMDAN of historical wind power and load data. Secondly, the intrinsic mode functions are reconstructed according to the criterion of the number of maximum value points. Thirdly, BiLSTM is used to predict the reconstructed components to obtain the predicted data of wind power and load data. Based on the predicted data, the multi-time-scale scheduling model for the regional power grid is established based on the model predictive control (MPC) method. Finally, the simulation results show that the proposed prediction method has universal applicability. The proposed multi-time-scale scheduling method is effective and economic, which can not only suppress the load fluctuation and reduce the influence of wind power grid-connection, but also use the flexibility of EVs to perform the real-time deviation compensation of wind power prediction, and to maintain the system balance.

Abstract:The access of large-scale distributed generators (DGs) such as distributed photovoltaics to the power grid increases the demand for dispatch flexibility of distribution network and the difficulty of traditional centralized control modes. Therefore, a two-stage intraday distributed optimal dispatch method for the distribution network considering the robust balance between the flexibility supply and demand is proposed. In the first stage, considering the uncertainty of net load and the flexibility supply capacity of the distribution network, a robust balance index on the flexibility supply and demand of DG cluster (DGC) is proposed, and a DGC partition model of the distribution network is established combined with the modularity index. In the second stage, according to the DGC partition results, an intraday distributed optimal dispatch model suitable for DGC is established, and a synchronous alternating direction method of multipliers (SADMM) is used to solve this problem. Finally, the effectiveness of the proposed method is verified by taking the IEEE 33-bus system as an example.

Abstract:Traditional day-ahead robust optimization algorithm can only provide the specific output value of equipment at each moment. Due to the influence of time coupling constraints such as randomness of the renewable energy output, the climbing constraints of equipment and state of charge constraints, it is difficult to provide an adjustment scheme for the intraday dispatching plan. In this paper, when considering the economic dispatching problem of the park-level integrated energy system (PIES), the interval variable of optimal power output is added into the traditional min-max-min three-layer optimization problem. The time coupling constraints are decoupled. Besides, the demand response constraint is decoupled by introducing the concept of accumulated load change. Due to the introduction of charging and discharging flag variables in energy storage, the nested column and constraint generation algorithm is used to deal with the three-layer optimization problem, and the actual carbon emission expression is relaxed and piecewise linearized. The optimal power output interval, upper and lower bounds of the demand response can be obtained, so that, during intraday dispatching, it can be adjusted arbitrarily within the interval according to the renewable energy output. Case study shows that, the method based on the optimal power output interval has a stronger ability to deal with uncertainty during intraday dispatching. At the same time, due to the decoupling of the state of charge constraints for the energy storage, the frequency regulation capacity of the energy storage can be determined according to the price of the frequency regulation market capacity to obtain benefits, and further reduce total cost. Finally, the influence of carbon trading price on the carbon trading cost and carbon dioxide emission reduction in the park is analyzed, so as to briefly explain the operation mechanism of the carbon trading market.

Abstract:The integrated energy system (IES) of electricity-gas-heat interconnection presents the characteristics of multiple time scales in the process of energy transmission, and the dynamic characteristics of the gas-heat network contain abundant flexible capacity. Fully tapping the potential flexibility of IES can effectively improve the consumption level of renewable energy. Thereby, an optimal dispatching model of IES considering the flexibility of sub-hourly-time-scale operation is proposed. Based on the analysis of the system operation regularity in the net load variation interval, the mathematical expression of power system flexibility demand and resources is derived. Then, the dynamic characteristics of gas network and heat network are analyzed, and the mathematical models of gas network and heat network which provide operation flexibility are established, respectively. Based on this, an optimal dispatching model of IES considering the flexibility constraints of sub-hourly-time-scale operation is constructed, and the nonlinear constraints in the model are linearized. Finally, the case studies on the modified electricity-gas-heat 24-20-16 bus system demonstrate that the dynamic characteristics of the gas-heat network can enhance the system flexibility in a short time scale and significantly improve the economy of system operation.

Abstract:In the context of operation flexibility improvement of thermal power plants in northern China, the traditional power balance method—determining power generation by heat is no longer applicable, so a coupling balance analysis model of provincial integrated electricity-heat energy system with multi-type flexible resources is constructed. The model first aggregates similar equipment to simplify the complexity of the problem, and considers system safety and stability constraints such as system backup and minimum operation mode through heuristic unit commitment with daily cycle. Then, the balance analysis without flexibility resources is carried out at the hourly granularity to determine the power surplus, and the electricity-heat flexibility resources are called in an orderly manner according to the surplus power to realize the electricity-heat coupling balance calculation. Finally, by rolling calculation day by day, the balance result on the long-term scale is obtained. Based on the actual power grid data, the calculation example verifies the validity of the model, and the model is used to analyze the effect of flexibility resource allocation on the improvement of renewable energy accommodation.

Abstract:Demand response is one of the important ways to stimulate the flexibility of thermal load and promote the accommodation of renewable energy. Aiming at the problem that the power granularity and time granularity of heterogeneous thermal loads may not be considered in the traditional demand-side management, which may cause power deviations in the dispatch and control, a coordinated optimal dispatch and control method for the thermal load considering the multi-granularity attributes of power consumption is proposed. Firstly, the coordinated response mechanism of thermal loads with different power granularity or time granularity is revealed, and the thermal load model considering the multi-granularity attributes of power consumption is established. Secondly, the priorities of dispatch and control are evaluated according to the granularity attribute difference of each thermal load, and the demand response is used to guide the thermal load to actively participate in the load adjustment, so as to achieve the purpose of coordinated complementarity of multi-granularity thermal load. Finally, by taking the penalty cost of the net exchange power of the system, the social cost of carbon emission per unit, the cost of electricity purchase, the economic subsidy cost of users with different priorities, and the penalty cost of user satisfaction reduction as the optimization objectives, the participation of multi-granularity thermal load in the balanced dispatch and control of electricity of regional power grid is simulated and analyzed to improve the flexibility of the regional power grid.

Abstract:The AC/DC hybrid distribution network is an important branch of the active distribution network, and the research on its flexibility operation region is of great significance to realize the active interaction of transmission and distribution networks and efficient and economic dispatch. Considering the flexibility operation characteristics of photovoltaic, energy storage and controllable load, a power flexibility operation region model of AC/DC hybrid distribution network based on DistFlow power flow model is constructed. Firstly, in order to reduce the influence of the randomness of photovoltaic output, the intersection of the operation regions in multiple typical scenarios is used as the flexibility operation region of the AC/DC hybrid distribution network, and the ellipses are used to approximate operation regions of active-reactive power of root nodes at the AC side and the interactive active power of the voltage source converter (VSC) for multiple consecutive periods. Secondly, the model is transformed into a single-layer optimization model by adding outer approximate polygon vertex constraints to maximize the area of the ellipse while ensuring the feasibility of its decomposition. Finally, the 33-node and 93-node AC/DC hybrid distribution networks are used as examples for analysis, and the influence of energy storage, controllable load, VSC control mode and different photovoltaic penetration rates on the power flexibility operation region is studied. The Monte Carlo method is used to verify the decomposition feasibility of the ellipse flexibility operation region, and the solution speed and accuracy are compared with the traditional method.

Abstract:With the increasing proportion of renewable energy generation, improving the flexible regulation ability of power system and maintaining the frequency stability of the power system are the key issues in the research of new power systems. As a flexible resource with fast response characteristics, how to fully guide electric vehicles (EVs) to participate in frequency regulation is a current research hotspot. Existing studies mostly use aggregators to manage EV clusters and participate in the frequency regulation auxiliary service market, but this mode has high risks in the data privacy security of EV users. The third party aggregators also have conflicts of interest with users in frequency regulation transactions. There are problems of unequal status and information asymmetry among power grid dispatching centers, aggregators and EV users. There is no channel for users to participate in frequency regulation auxiliary service market without relying on aggregators. Therefore, the definition of virtual aggregator is given in this paper, and a virtual aggregation frequency regulation framework is proposed. Based on the consensus mechanism of consortium blockchain, smart contract and other technologies, services such as frequency regulation resource aggregation and frequency regulation instruction decomposition required by power grid dispatching center and EVs are realized. The example shows that the proposed virtual aggregation frequency regulation framework can realize decentralized distributed frequency regulation optimization in the consortium blockchain network, and is more fair and reasonable in revenue distribution, which is conducive to encouraging EV users to maintain a high response level and promoting the flexible resource utilization of EVs.

Abstract:High fluctuation and randomness caused by the integration of high proportion of renewable energy lead to the rising demands for flexibility resources in power systems. The flexible ramping product (FRP) can effectively improve the flexibility of the power system. However, the existing research on FRPs focuses on the supply side, and the research on trading strategies on the demand side for FRPs remains to be studied. For managing distributed FRPs provided by the demand side, a virtual power plant (VPP) is regarded as the integration mode to uniformly purchase FRPs from users and sell them to the main grid after the integration using the flexible price. To balance the profit of VPP and demand side, the energy transaction bi-level optimization model of demand-side VPP is proposed considering the trading of FRPs. Meanwhile, a stochastic ranking based surrogate-assisted evolutionary algorithm is proposed to solve the model. To overcome the difficulty in obtaining user private information, the radial basis function network is applied to fit the proposed bi-level model. In addition, the stochastic ranking mode avoids the optimization deviation caused by the excessive error from the single surrogate model. The testing results of the case studies on the PJM electricity price data sets verify the accuracy and feasibility of the proposed method, which provides a theoretical decision basis for the trading of FRPs on the demand side.

Abstract:After the high-permeability distributed generators are connected to the load side, the structure and operation forms of the distribution network will significantly change. In this background, this paper studies a scheduling method and optimal model of new power system in the framework of edge-cloud collaboration. Cooperative game is adopted as the theoretical basis for the collaborative and interactive optimization of the edge-cloud collaboration. The scheduling optimal model of flexible resource at demand side and the clearing optimal model at distribution network side considering network topology constraints are established on both sides of edge and cloud, respectively. In addition, considering the packet loss in communication process, a double-layer closed-loop control model which regards the edge service platform as a hub is established. The economy and reliability of the interactive scheduling scheme and control model are verified by numerical simulation.

Abstract:The resilience of the distribution network refers to the ability to withstand extreme disasters, reduce fault losses, and restore power supply as soon as possible. In order to avoid large-scale power outages caused by insufficient resilience of the distribution system, this paper proposes a resilience enhancement strategy for the distribution system based on mobile energy storage (MES) sharing in a peer-to-peer (P2P) transaction mode. Firstly, a framework for P2P transactions is constructed. In this framework, users in the distribution system can manage and distribute their electric power according to the load demand, and realize the source-load-storage coordination among power generation, consumption, and energy storage. Secondly, this paper proposes the MES sharing mechanism to realize the spatial-temporal transfer of MES and improve the resilience of the system by optimizing the driving path and charging and discharging power of MES. Thirdly, this paper proposes a cost equalization mechanism of MES sharing based on the proportional distribution method. Finally, the effectiveness of the proposed strategy is verified by simulating and analyzing the 15-bus, IEEE 33-bus, and IEEE 69-bus radial distribution systems based on Julia, respectively.

Abstract:In view of the problem that distribution network lines are covered with ice frequently and the overload de-icing is prone to power flow over limit, the flexible adjustment ability of the soft open point (SOP) and the mobility of the mobile de-icing device (MDID) are fully considered in this paper. An optimization method of multi-time de-icing strategy in the distribution network is proposed. Firstly, the variation pattern of ice-cladding thickness during de-icing is depicted based on the heat balance equation and meteorological data. Secondly, the influence of the transportation network on the MDID movement time is considered. To minimize the risk of line breakage and power flow over limit, this paper establishes a bi-level optimization model of the distribution network based on the coordinated cooperation of SOP overload de-icing and MDID path scheduling. The model is solved using an elite-preserving genetic algorithm. Finally, the analysis is carried out by using the hand-in-hand IEEE 33-bus distribution network and 114-bus distribution network. Simulation results verify the effectiveness of the proposed method.

Abstract:The influencing factors such as the grid connection of distributed energy sources, the mixed operation of radial and toroidal topologies, and the measurement errors make it difficult to accurately obtain the online topology of distribution networks. For this purpose, this paper proposes an identification method for the distribution network topology based on two-stage feature selection and Gramian angle field (GAF). Firstly, the two-stage feature selection method based on XGBoost and maximal information coefficient is used to filter out the important measurement data that do not contain redundant topology feature information. Secondly, the one-dimensional time-section measurement data of the distribution network are transformed into two-dimensional GAF based on GAF feature transformation. While preserving the nodal voltage distribution pattern of the time section, the topological feature information in the nodal voltage magnitude distribution is effectively characterized. Finally, a three-convolutional-layer neural network model for topology identification is designed. With the operation characteristics of convolution and pooling and noise immunity, the topological feature information embedded in GAF is stably extracted. Thus, the mapping relationship from the nodal voltage-amplitude distribution to the topology of the distribution network is accurately established. The effectiveness of the proposed method is verified by the IEEE 33-bus system, and the adaptability of the proposed method to different noise levels, missing data ratios, and other scenarios is analyzed.

Abstract:In order to achieve the joint planning with robustness of wind-photovoltaic-energy storage system (ESS) to deal with the negative impact of the uncertainties of renewable energy output, a robust-driven multi-scenario confidence gap decision theory (MCGDT) is proposed. Considering the optimization objectives of maximizing the wind/photovoltaic power consumption rate and minimizing the total investment cost, a wind-photovoltaic-ESS joint robust planning model based on MCGDT is established. Furthermore, the complex chance constraints are transformed into the equivalent determinate constraints according to the uncertainty theory, and a novel cross entropy-radar scanning differential evolution (CE-RSDE) algorithm is proposed to solve the model efficiently. MCGDT can not only make up for the inadequacy of conventional robust optimization and information gap decision theory (IGDT) that fails to measure the robustness accuracy and embody the polymorphism of randomness, but also solve the problem of non-interval ergodicity of typical scenarios in stochastic programming. Therefore, a more reasonable and accurate uncertainty planning can be realized. The validity and superiority of the proposed theory and method are verified by case study.

Abstract:The randomness of power disturbance leads to the uncertainty of the frequency regulation energy of wind power,resulting in that the setting parameters of frequency regulation may not match the actual frequency regulation capability and may be unknowable. At the same time, the virtual inertia and the primary frequency regulation have the same energy source of frequency regulation, which cannot achieve a reasonable natural distribution, making the power grid unable to obtain equivalent parameters that reflect the frequency regulation capability of the wind farm. Therefore, from the perspective of grid operation, a linked distribution mechanism of frequency regulation energy of wind power is constructed, so that the virtual inertia and primary frequency regulation are linked up to distribute the energy of the two frequency regulation links under the total constraint. On this basis, considering the capture loss of the mechanical energy of the operation point offset, an expression model of the net energy that can be released by the wind turbine is obtained according to the wind speed information. Furthermore, the equivalent frequency regulation parameters are calculated according to the energy distribution model, so that the wind power can participate in the entire process of rapid frequency regulation at low and medium wind speeds, which avoids energy imbalances between grid frequency regulation links, and has clear equivalent frequency regulation parameters that match the time-varying wind speed and actual frequency regulation capabilities. Finally, the application modes of the proposed method on the wind farm side and the power grid side are clarified. The simulation results show that the equivalent frequency regulation parameters of the wind power can be updated by obtaining the value of changing wind speeds, which has the characteristics of linkage and knowability, and is time-varying following the changing wind speeds, and can effectively feedback the actual frequency regulation capability of wind power, so that frequency regulation capability levels of the wind farm and the whole network are timely obtained for the grid operation.

Abstract:With the large-scale integration of wind power into the power grid, the demand for probability prediction of wind power is becoming more and more urgent. In order to realize the short-term probability distribution prediction of wind power, a short-term wind power probability prediction method based on the improved gradient boosting machine (GBM) algorithm is proposed. Firstly, the problems of the GBM algorithm applied to short-term wind power probability prediction are analyzed. Secondly, the negative log-likelihood loss function is used as the loss function in the GBM algorithm, and the Fisher information matrix is used to modify the gradient of the loss function in the parameter space of probability distribution and convert the gradient into the natural gradient of the probability distribution space. Then, based on the natural gradient, an improved GBM algorithm suitable for the short-term wind power probability distribution prediction is proposed. Finally, the proposed algorithm is compared with the traditional GBM algorithm and other methods. The results show that the training process of the proposed algorithm converges faster and has better prediction performance, which verifies the practicability and effectiveness of the proposed algorithm.

Abstract:In order to better mine the spatial-temporal dynamic characteristics of electric vehicle charging load under the situation of strong grid-transportation network coupling and improve the accuracy of charging load forecasting, a framework of graph WaveNet based charging load forecasting for electric vehicles is proposed. First, the charging stations in the coupled grid-transportation network are regarded as charging load nodes. Then, by regarding the charging load data of the charging stations as the characteristic information of the nodes, all the nodes are constructed into a graph, and the graph containing the spatial-dimension information of charging loads and the time-dimension information of charging loads are input into the adaptive graph WaveNet framework for forecasting. Finally, taking the charging station load data in an urban area of a city in China as an example, the forecasting results based on the adaptive graph WaveNet framework are compared with the forecasting results of the existing methods, and the correctness and effectiveness of the proposed method are verified.

Abstract:During June 15 to 24, 2022, the Australian National Energy Market (NEM) was suspended, which has attracted widespread attention in the industry. Firstly, combined with the operation of NEM before and after the suspension, starting with the analysis of market operation data, this paper deeply analyzes the emergency response mechanism of NEM suspension and focuses on the response measures and implementation procedures of Australian Energy Market Operator (AEMO). Further, the paper summarizes and combs the cases and response mechanisms of electricity market emergencies around the world in recent years, including the relevant provisions of emergency identification standards and intervention means, and classifies and summarizes them to find common mechanism elements. Finally, based on international experience, this paper puts forward relevant policy suggestions for emergency response to China's current electricity market construction promotion, including: identification standard, operation mechanism, pricing and compensation mechanism, market recovery mechanism, etc.