Data center, as an important carrier of digital infrastructure, has been growing rapidly in the number and scale, which results in increasing system energy consumption and operating costs. The integrated energy system (IES) is a complex system which takes the power system as the core and covers the production, supply and marketing of energy such as electricity, gas, cooling and heat. Its optimized operation can realize comprehensive management and economic dispatch of multiple energy sources. Using IES to replace the original energy supply system is an effective solution to reduce the operating cost of the data center. At present, the typical architecture of IES has not been clearly defined, and there is little research on the optimization operation of IES based on data center. Firstly, this paper introduces the typical physical framework of IES and data center-based IES. Secondly, it analyzes the physical model of the coupled operating equipment and the application of energy hub and its extended model in the system. On this basis, this paper summarizes and sorts out the current status of IES and the optimized operation of IES for data centers, and proposes an operation method based on data centers, including multiple energy input, coupling and output. Finally, this paper looks forward to the key development directions and potential research contents of energy uncertainty and load forecasting, energy dynamic modeling, optimized scheduling strategies, and benefit evaluation in IES based on data center, with a view to provide reference for future related project landing, simulation research and development.
This paper studies the distribution network planning problem which considers high proportion of photovoltaic power and large-scale electric vehicles (EVs). Firstly, the paper quantifies the flexibility of the EV virtual power plant, and then establishes a distribution network planning model that comprehensively considers the flexibility of the EV virtual power plant and the high proportion of photovoltaic access. The model aims to minimize the annual investment cost of distribution network lines, while taking into account renewable energy consumption, investment costs of energy storage system and flexibility compensation costs of EV virtual power plant, in order to improve the economy of distribution network planning and realize peak shaving and valley filling of power system and improve the photovoltaic output consumption rate. Finally, the IEEE 24-node distribution network system is taken as an example for simulation verification. The example shows that the planning model proposed in this paper uses the flexibility of the energy storage system and electric vehicles to reduce the planning and operation cost of the system and improves the consumption rate of the photovoltaic power station in the distribution network. The model can provide reference for future power system planning including high penetration of renewable energy and virtual power plant flexibility resources.
Under the background of “double carbon”, the access to distributed power generation in the distribution network is gradually increasing. How to decide the output setting value of distributed power generation to make the economic operation of the distribution network becomes a key issue of research. In this paper, an in-depth study on distributed power management and control is carried out. On the basis of a time-varying optimization model, an online optimization decision algorithm for distributed generations considering voltage prediction is proposed to solve the problem that the existing algorithm does not consider the influence of the decision of distributed power in the current period on the decision of distributed power in the future period. Firstly, the network node voltage after the load change should be predicted according to the load change of each node in the distribution network. The predicted node voltage values in distribution network can be further used to decide distributed power generation. Then, the change of the decision setting value of the distributed power source is used to predict the network power flow state after the decision of the distributed power source, and so on to solve the decision setting value of the distributed power source in all periods. For the 502-node distribution network example system, the online optimization decision algorithm proposed in this paper is used to verify the optimal power flow convergence of the distribution network, the convergence of node voltage regulation, and the validity of charging and discharging prediction of the energy storage power supply.
In the study of low-carbon economy dispatching of park integrated energy system (PIES), it is difficult to accurately estimate the carbon emissions caused by purchasing electricity from outside, which poses a severe challenge to the accuracy of PIES low-carbon economy dispatching. In this paper, the carbon flow density of the external power grid is used to accurately calculate its hidden carbon emissions, and a two-layer optimization model of PIES low-carbon economic dispatching is established. Finally, an example is analyzed in a typical PIES. The results show that, compared to PIES economic scheduling alone, the proposed optimization model can reduce carbon emissions of the PIES by about 21.37% while increasing operating costs by 4.82%. At the same time, the proposed model can optimize the carbon flow distribution of PIES to eliminate the weak link of the system and greatly improve the environmental protection of PIES.
With the increasing integration of PV, the peak and valley of the source and the load are inconsistent in the distribution networks, and the PV generation is intermittent and fluctuating. These characteristics have brought about the problem of overvoltage. This paper proposes a divisional and hierarchical voltage regulation model based on the sensitivity analysis of holomorphic embedding method (HEM). First, the reactive-power output model for energy storage and PV are established, and the voltage sensitivity matrix of distribution network based on HEM is derived. A network partition method is further studied using sensitivity matrix. Then, a two-layer voltage optimization strategy is proposed. The upper-level model dispatches the grid-connected energy storage, PV and volt-var regulating equipment to regulate the voltage in the entire network. The lower-level voltage regulation model further explores the reactive power potential of small grid-connected distributed PV and energy storage to adjust the voltage in each region according to the voltage sensitivity. Finally, the proposed strategy is simulated on a 35kV distribution network. The simulation verified that the proposed sensitivity calculation method is applicable and the voltage regulation strategy is efficient in solving the voltage problem.
In the optimal configuration of distributed generation in active distribution network, the output and load timing fluctuation characteristics of wind power cause too many scenes, which will not only increase the difficulty of model solution, but also affect the optimal configuration results. For this reason, this paper first introduces Davies-Bouldin index (DBI) and K-means algorithm to cluster and reduce the full-hour scenes. Secondly, a two-level optimal allocation model of active distribution network considering the time series characteristics of wind power and load is established. The upper layer aims to optimize the annual comprehensive economic cost of the system, and the decision variables are the locations and capacities of wind power units connected to the distribution network; The lower level takes the optimal operation voltage level of the system in each period as the objective, and the decision variables are the operation costs under the active management measures of the distribution network; Considering the time sequence outputs of wind power and the time sequence response of load sides, the economic planning of distributed generation in active distribution network is realized. Finally, an improved adaptive genetic algorithm is used to solve the model. The IEEE 33-node system example shows that the proposed method and model not only ensures the diversity and rationality of scenario selection, but also takes into account the wind power output and load timing fluctuation characteristics, and improves the system operation level and reduce the system operation cost.
After a large number of distributed power sources such as distributed photovoltaic power and energy storage devices are connected to an active distribution network (ADN), the multiple uncertainties of photovoltaic output, load and rotating reserve capacity bring new challenges to the reliability of economic dispatching of ADN. Considering the uncertainty of photovoltaic output and load, this paper puts forward an economic optimization scheduling model of ADN with the objective function of minimizing operation cost. In this paper, the chance constrained programming (CCP) model is transformed into a deterministic mixed integer linear programming (MILP) model, which is easy to solve, by using CPLEX solver. The proposed method is verified by IEEE 33-node distribution system. The calculation results show that the proposed strategy significantly reduces the system calculation time and achieves better optimization effect, and the balance between the reliability and economy of ADN can be achieved by appropriately setting the confidence level of the probability constraint of rotating reserve capacity.
In order to realize the commercialization of cloud energy storage business, on the premise of ensuring the best overall benefit of users, a cloud energy storage service mechanism considering multi-agent power transactions is proposed, and the value of matching energy storage resources with renewable energy is explored. This paper provides a new way of thinking for the commercialization of cloud energy storage by introducing a competition mechanism in electric energy trading, to encourage users to actively participate in services and improve user autonomy and transaction efficiency. Firstly, the architecture and business process of cloud energy storage service mechanism are designed according to the user’s energy consumption behavior. Secondly, a day-ahead power dispatching model and strategy with the goal of minimizing the net energy cost of users are constructed. Then, an electric energy trading bidding and matching strategy based on multiple rounds of collective bidding is established. On this basis, a default processing and fee settlement model is designed to clarify the benefits of each participant. Finally, a park is taken as an example to verify the effectiveness of the proposed service mechanism. The results show that this mechanism can realize the maximum local consumption of renewable energy, reduce the energy consumption cost of users, enable healthy competition among users, and achieve a win-win situation for all parties.
At present, the transmission grid with a high proportion of photovoltaic and wind power has a generally high rate of curtailment of wind and solar power. As a potential utility, the configuration of energy storage equipment is favored in current research. Aiming at the problem of energy storage configuration in the transmission network, this paper proposes a configuration method based on a two-layer programming model for the hydrogen-electric hybrid energy storage system in order to reduce the comprehensive cost and the curtailment rate of wind and solar energy. Firstly, the characteristics of the hydrogen-electricity hybrid energy storage system are modeled, and on this basis, a two-layer planning model for the power and capacity allocation of the hydrogen-electricity hybrid energy storage system in the transmission network is established. The upper-layer model aims to minimize the annual comprehensive cost after the energy storage is configured; the lower-layer model aims to minimize the curtailment rate of wind and solar energy. Then, according to the different characteristics of batteries and hydrogen energy storage, a hydrogen-electricity hybrid energy storage system coordination strategy is proposed. Finally, according to the nonlinear multi-objective characteristics of the model, the two-layer iterative particle swarm optimization algorithm is combined with power flow calculation to solve the model. In the simulation part, taking a transmission network with a high proportion of wind and solar power access as an example, the capacity and location of hydrogen-electricity hybrid energy storage system are optimized to verify the feasibility and effectiveness of the model and the proposed solution algorithm. Ultimately, the transfer ability of photovoltaic and wind power output in time series is improved, the consumption of high-carbon fossil energy is reduced, and the curtailment rate of wind and solar energy is reduced.
The continuous increasing of penetration rate of photovoltaic power reduces the rotational inertia of the new power system and brings new challenges to the frequency stability of the system. Therefore, it is imperative to combine photovoltaic power and energy storage to form a combined PV and energy storage power station to participate in the frequency regulation of the new power system. The active power output of the combined PV and energy storage power station participating in the frequency regulation of the system is not only affected by the randomness and volatility of the photovoltaic output, but also limited by various output constraints. The ability to handle multiple constraints is also poor. In this context, this paper proposes a new frequency regulation control strategy based on model predictive control for combined PV and energy storage power stations in power systems. The control strategy aims to minimize the sum of the system frequency deviation and the rate of change of frequency, and takes into account the constraints of the active power output and total power generation of the combined PV and energy storage power stations to obtain the optimal active power output of the combined PV and energy storage power stations, and participate in the system frequency regulation by controlling the active power output of the combined PV and energy storage power stations. Finally, simulation examples verify that, the proposed method has a better effect in frequency regulation than the traditional control strategy, which can quickly and accurately determine the output of the PV and energy storage power station under the condition of considering multiple constraints, and improve the frequency stability of the system.
Power supply structure has undergone tremendous changes with the increasing proportion of renewable energy in the power system. The high permeability, randomness and uncertainty of the output of renewable energy cause problems increasingly prominent, such as demand of power system peak-shaving resources on both power supply and load side. Firstly, this paper summarizes temporal and spatial distribution characteristics of renewable energy, and explains the peak-shaving problem of power system with high proportion of renewable energy. Then, the equipment, characteristics and development trend of various peak-shaving energy storage types are summarized. The paper also elaborates the peak-shaving modes of the demand side in the power system, including demand response construction, power market and auxiliary peak-shaving service market, and enhancing power transmission capacity. Finally, the future research prospects of peak-shaving of power system with high proportion of renewable energy are discussed and prospected.
As the proportion of renewable energy power generation continues to increase, the inertia of modern power system continues to decrease, and the frequency characteristics and power transmission appear unstable. This paper analyzes the influence of renewable energy access on the dynamic and steady-state characteristics of traditional power grid frequency and power, and clarifies that renewable energy needs to participate in system frequency regulation. Considering the coordination of the output of traditional synchronous generators and renewable energy, and the performance of the system frequency characteristics, three ways of renewable energy participating in the system frequency regulation are analyzed, and the influences on the frequency and power characteristics under load step are compared. A coordinated frequency regulation scheme for the output of traditional units and renewable energy is proposed. The hardware simulation experiment based on RTDS shows that the frequency regulation scheme proposed in this paper ensures the dynamic and steady-state characteristics of the frequency response of the system, and solves the problem of coordination between the synchronous generator and the renewable energy power.
The randomness of wind power output makes it difficult to balance the robustness and computational efficiency of microgrid grid connection planning. A wind power output scenario construction method matching with typical daily load scenarios is proposed. The daily load trend and the location of peak and valley periods should be considered in the microgrid planning. The daily load curve trend and peak and valley period information are extracted by using the membership function, and combined with the improved ordered clustering, a typical daily load selection method is proposed; In the effective time of typical daily load, using the maximum increase and decrease of wind power output, combined with interpolation method, a wind power scene construction method is proposed. Then an evaluation index system is established to evaluate the selection of typical daily load and the construction effect of corresponding wind power scenarios. Finally, the effectiveness of the proposed model is verified by power grid data.
In order to solve the problems of long decision-making time, high cost of trust and privacy security faced by microgrid in the traditional centralized trading, a decentralized market trading system based on multi-agent deep deterministic policy gradient (MADDPG) algorithm and smart contract is proposed for microgrid. Firstly, after dividing the multiple agents in the microgrid market, a decentralized transaction mechanism for microgrids that is suitable for all entities to participate in distributed transactions is designed to protect the interests of market entities. Secondly, in order to realize the optimization of the transaction strategy of the microgrid market entities in the transaction confirmation stage, the multi-agent deep deterministic policy gradient algorithm is used to solve the bidding model that each entity pursues the most benefits. Finally, the feasibility and economy of the MADDPG algorithm in the optimization process of the transaction strategy of the microgrid market entities under the smart contract is verified by example simulation.
Considering that the game relationship between load aggregators as the organizers and users as the members of the alliance directly affects the overall external bidding and profitability of the alliance, this paper takes the scenario of load aggregator’s participation in the peak-shaving market, factors such as load peak regulation effect and load aggregator’s clearing ability are quantified and used as bargaining chips for different game subjects based on Nash’s bargaining theory. A revenue distribution model between load aggregators and various types of users is established. According to the different attitudes of peak electricity price and electricity, the user satisfaction index is constructed. The simulation results of the numerical example show that the proposed model can ensure the stability of the alliance and increase the overall income of the alliance, and can reasonably and effectively distribute the income of the main bodies. The strategy of formulating a floating electricity price can form positive incentives for user demand response and improve the income of both parties. The overall and individual win-win results verify the rationality and superiority of the method proposed in this paper.
