Abstract:The virtual power plant（VPP）holds great potential in integrating distributed energy resources（DER）, which provides an efficient solution for the flexible management and effective utilization of grid resources. A flexible resource aggregation operation boundary estimation method for VPP is proposed to facilitate VPP participation in system dispatch. First, the active and reactive power operation constraints of DER are considered and expressed in a unified linear inequality form. Then, all heterogeneous DERs on the same distribution network node are aggregated into one DER aggregator by a geometric computation method. Finally, considering the network constraints, a VPP operation boundary estimation model is established, and the operation boundary is calculated using the boundary point search method. The case study results show that the proposed method can effectively estimate the multi-time operation boundary of VPP by considering the power coupling and time coupling characteristics.

Abstract:The regulation boundary of adjustable load is one of the key data that needs to be mastered in advance for the safe operation of the power grid. For process-oriented processing industry（PI）, the accurate evaluation of its whole-process load regulation capacity is limited by multiple energy requirements and process coupling, as well as the process requirements and production tasks that must be met. To this end, a method to evaluate the whole-process load regulation capacity of PI considering the refined modeling of processes is proposed,which provides a boundary reference for the power grid to formulate the regulation plan. Firstly, the production process of typical PI is analyzed, and an equivalent model of generalized PI considering process classification is established based on the adjustment characteristics of processes, and a refined modeling method considering process classification is proposed. Secondly, considering the production constraints such as process coupling, multi-energy utilization, and production planning, the model to evaluate the whole-process load regulation capacity of PI is established. Finally, the effectiveness of the proposed method is verified by taking the cement industry as an example,and the effect of process regulation on the whole-process load regulation capacity of PI is explored.

Abstract:The power industry is key to achieve the goal of peak carbon neutrality. Green electricity consumption effectively reduces the carbon emissions of electricity, and local users actively carry out green electricity trading to assume the responsibility of emission reduction. The current accounting system does not distinguish green electricity from traditional thermal power, and the environmental value of green electricity has not been reflected. Based on the principle of shared proportion, carbon flow tracking is formed by tracking the trend,and the user carbon emission accounting model is constructed. Then, an equivalent network is established considering the financial characteristics of green electricity trading, the carbon emission reduction benefits of green electricity trading are quantified, and the revised user carbon emission accounting model is established. Finally, the accuracy and rationality of the model method are verified by the IEEE30node calculation example.

Abstract:Due to the multimodal and nonlinear nature of photovoltaic（PV）models, parameter identification is a challenging problem. In view of the limitations faced by traditional algorithms in the field of PV model parameter identification, such as insufficient reliability, low accuracy, easy to fall into local optimal solutions and premature convergence, a improved complex valued encoding symbiotic organisms search（ICSOS）is proposed for PV model parameter identification. In order to enhance the optimization ability of the traditional symbiotic organism search algorithm, a complex valued encoding is introduced, which expands the original one-dimensional real number coding to a two-dimensional complex coding space, in order to expand the search range of the population and enhance the optimization ability and speed of the algorithm. Simulation validation shows that the proposed improved algorithm has good applicability in the process of parameter identification in single diode model, and PV module model, and compared with other optimization algorithms, the ICSOS algorithm is able to obtain lower root mean square error（RMSE）values and can quickly find the optimum to effectively reduce the prediction error and improve the accuracy of parameter identification.

Abstract:Aiming at the economic stability of substation integrated energy system, a multi-time scale optimization scheduling strategy based on model predictive control（MPC）is proposed, which combines ground source heat pump and demand response. Firstly, the system architecture is constructed and the MPC algorithm flow is described. In day-ahead scheduling, by considering the randomness of TOU and new energy, an optimization model aiming at minimizing operation cost is established, and the optimal output and energy storage plan of the unit are determined. Intra-day scheduling uses MPC algorithm to correct the day-ahead schedule in real time to reduce the impact of uncertainty on the power grid and improve the economy. The demand response characteristics and energy efficiency level of ground source heat pump are discussed by comparing different schemes through example analysis. The results show that the proposed model can optimize the system operation under the demand response environment, improve the equipment utilization rate, reduce the operating cost, and promote the consumption of renewable energy.

Abstract:In order to further reduce the load pressure during peak and valley periods in the integrated energy system, improve the low-carbon economy of system operation, and reduce system operation risks, the differences in demand response of different loads and the uncertainty of demand response are considered. The uncertainty of demand response is fuzzily processed, and a low-carbon economic dispatch model for the integrated energy system considering the uncertainty of differentiated demand response is established. Firstly, differentiate pricing for different types of loads and use a tiered carbon emission cost model to constrain carbon emissions. On this basis, a planning model is established with the goal of minimizing the total operating cost of the system, including the operation and maintenance cost, energy sales revenue, and carbon emission cost. Then, the squirrel algorithm is improved and solved using the chaotic squirrel optimization algorithm. Through actual case analysis, the results show that considering the uncertainty of differentiated demand response can improve system reliability, smooth the load curve, and improve the low-carbon economy of the system. This verifies the effectiveness of the established scheduling model, and the data also shows that the chaotic squirrel algorithm has better optimization ability.

Abstract:Micro-grid optimization scheduling is an effective technical means to absorb intermittent distributed energy. In order to achieve the dual goals of economic and environmental optimization of micro- grid, a micro- grid optimization model including photovoltaic cells,fans, micro-gas turbines, diesel generators and batteries is established. The model is solved by an improved whale algorithm. In view of the problems of slow convergence and easy to fall into the local optimal solution of the traditional whale algorithm, Tent map is used to initialize the population, adding the sine and cosine operators to improve the bubble net attack phase of the whale algorithm, and Levy flight is introduced to enhance the global search ability. The improved whale algorithm is applied to the microgrid. Compared with traditional particle swarm optimization and basic whale algorithm, the results show that the improved whale algorithm with mixed sine and cosine operators has faster iteration speed and better economy, and has a good solution effect for the optimization scheduling problem of microgrid.

Abstract:Flexible load resources can respond to power grid dispatching quickly without significant impact on user comfort because of their rapid response and flexible regulation. As the core part of flexible load, air conditioning load can reduce the peak power demand through scientific control strategy, and then relieve the pressure of power supply. In view of the nonlinear and fuzzy characteristics of air conditioning load data, a model of air conditioning load prediction based on modal decomposition and neural network is proposed. First,Pearson correlation coefficients are used to construct similar weekly load sequences. Then the load is decomposed by adaptive noise complete set empirical mode decomposition and variational mode decomposition（VMD）. In the VMD section, the original time series signal is input into the VMD layer and decomposed into multiple eigenmode functions（IMFs）by the VMD algorithm. These IMFs are input into convolutional neural network respectively, and their local features are extracted by convolutional, activation and pooling operations. These feature vectors are then fed into a bidirectional long short-term memory network, which uses its bidirectional propagation capability to capture long-term dependencies in the sequence. The improved whale algorithm is used to optimize the hyperparameters, and the regulation potential of the load is further discussed on the basis of the output forecast load sequence. The experimental results show that this method not only has high forecasting speed and accuracy, but also can reveal the adjustment potential of load more clearly.

Abstract:In order to further reduce the forecasting error of electric load data, a short-term power load forecasting method based on load decomposition and identification is proposed. First, for the electric power load data of each industry, the polynomial fitting error of temperature-sensitive load to the temperature series is taken as the objective function, and the load decomposition is transformed into a mathematical optimization problem, and the total load of each industry is decomposed into the weekly load based on load identification component and the temperature-sensitive load component. Second, the short-term load prediction is performed for the temperature-sensitive load component based on the long short-term memory network. Finally, the temperature-sensitive load prediction results are superimposed with the weekly load based on load identification component to obtain the complete load forecast results. The results show that the short-term load forecasting method based on load decomposition and identification proposed can effectively reduce the short-term load forecasting error.

Abstract:Power load forecasting is the basis of power system development planning and power generation plan. The load data of power grid is huge, complicated in structure and diverse in statistical scope, and the factors affecting load change are changeable. The large-scale access of new energy further increases the difficulty of power load forecasting. A multi-level load forecasting method of power system based on improved long short term memory（LSTM）is proposed, which establishes the time series relationship of multi-level load at the provincial,municipal and substation levels, takes historical load data, meteorological data and regional economic data of different load levels as input of the forecasting algorithm, and classifies the load factors at each level. Constraints are added from the perspective of planning development, and the three-layer stacked neural networks prediction model based on improved LSTM is used to complete the overall prediction of each level of load. The simulation example is based on the actual power load data and PV output data of S province and Y city in East China. The results show that the proposed method has a good effect on improving the prediction accuracy of multi-level power load.

Abstract:With the continuous increase in the penetration rate of photovoltaic power generation, data quality have become a key factor affecting the intelligent operation and grid connection research of photovoltaic power plants. The existence of bad data not only affects the accuracy of predictions, but may also lead to deviations in photovoltaic system status monitoring and fault diagnosis. To improve the integrity and reliability of photovoltaic power plant data, this paper proposes a method for identifying and reconstructing bad data in photovoltaic power plants based on multi-source heterogeneous data correlation. Firstly, analyze the data characteristics and correlation between multisource parameters of the photovoltaic system under normal operation, and select the historical data with the most similar characteristics to the data to be reconstructed as input. Secondly, the multi density clustering algorithm based on relative density is used to identify and clean power poor data. Finally, based on the correlation of environmental data, a photovoltaic system combination long short-term memory data reconstruction model is established to achieve high-precision reconstruction of the data. The calculation results show that the proposed method can effectively identify the bad data of photovoltaic power station output and accurately reconstruct it.

Abstract:Under the background of the transformation of distribution network to low-carbon, the high proportion of green energy and electric vehicles powered by electric energy are developing continuously, and the operation and planning of distribution network are facing new challenges and opportunities. Based on this, an optimal planning model of distribution network including electric vehicle charging load and distributed photovoltaic is proposed. First, the electric vehicle charging load and distributed photovoltaic timing characteristics are modeled. Second, considering the economy, low carbon and stability of the distribution network, a multi-objective planning model is constructed with the annual comprehensive benefit, carbon emission and load variance of the distribution company as the objective function,and the multi-objective coati algorithm is used to solve the above model, and the global Pareto optimal solution set is retained based on the Pareto dominant principle. The equilibrium decision function is introduced to evaluate the distribution equilibrium of the optimal solution.Finally, the feasibility of the proposed model and method is verified by an example analysis on IEEE-33 node power distribution system.

Abstract:In order to improve the distributed photovoltaic carrying capacity of the distribution network, a distributed photovoltaic carrying capacity improvement method considering a variety of regulation and control methods is proposed. Firstly, considering the reactive power control of the inverter on the power supply side, the additional reactive power compensation on the distribution network side and the network reconstruction, a multi-objective optimization planning model with the largest photovoltaic access capacity and the best economy in the distribution network is established. Secondly, aiming at the uncertainty of PV-load in the distribution network, a typical joint timing scenario considering the correlation between PV-load timing is constructed. Finally, the NSGA-II multi-objective optimization algorithm is used to solve the model, and the optimization results are obtained. The simulation results show that the combination of three control methods, namely inverter reactive power control, additional reactive power compensation and network reconstruction, can effectively improve the distributed photovoltaic carrying capacity of the distribution network under the premise of taking into account the economy.

Abstract:To reduce the phase lag existing in the control process of energy storage system（ESS）in the photovoltaic fluctuate suppression scenario, an innovative ESS controller based on second-order filtering is proposed. The effect of the phase delay characteristic of the traditional smoothing filter on the ESS capacity is analyzed. The larger the phase delay is, the greater the ESS capacity demand is. In order to eliminate this effect, the lowest order of the controller that can eliminate phase delay is the second order, and an ESS controller based on the second order filter is derived, and the control parameter design method is proposed to effectively reduce the phase delay in the smooth control process and minimize the ESS capacity requirements. The controller is simple in design and contains only two control parameters,which can be applied to practical projects. Finally, the power data of a 5MW photovoltaic power station are used to verify the effectiveness and superiority of the proposed ESS controller.

Abstract:In the context of China’s electricity market reform, the effective integration of mid-to-long-term and spot markets is studied in depth. Firstly, the composition of the electricity market system is outlined, and the main points of convergence between mid-to-long-term and spot markets in terms of trading, scheduling and settlement are analyzed. On this basis, a series of integration mechanisms are proposed, including the mid-to-long-term curve decomposition mechanism, the physical security guarantee mechanism and the price linkage mechanism. The effectiveness of the proposed mechanisms in actual operation is verified through the analysis of arithmetic examples. Finally, a coordination scheme between mid-to-long-term and spot markets is proposed, aiming at promoting the complementarity of the two markets and realizing the smooth operation and sustainable development of the electricity market.

Abstract:In the context of source grid load collaborative scheduling, a reliability evaluation method for active distribution networks con sidering load recovery strategies is proposed to achieve accurate assessment of distribution network reliability. Firstly, a time-varying fail ure rate model for components and a multi state model for distributed generators（DG）are constructed based on Markov processes. Based on this, an improved Distflow power flow model suitable for fault reconstruction and islanding partitioning is proposed. The model is linearized using the big M method and a linear radial constraint method is introduced to model the load recovery strategy as a mixed integer linear programming（MILP）problem. Secondly, power outage consequences analysis is conducted for various fault scenarios to achieve reliability assessment of active distribution networks. Finally, the impact of load recovery strategy on reliability was analyzed through IEEE standard examples, and the results showed that considering load recovery strategy can effectively tap into the power supply recovery capability of active distribution networks and further improve power supply reliability.

Abstract:Aiming at the problems of increased operating cost, insufficient line capacity and high operating risks caused by short-term peak loads after distributed photovoltaic is connected to the distribution network, a two-layer planning method of distribution network including distributed power is proposed considering the influence of short-time peak load. The upper-layer planning model is based on particle swarm optimization algorithm, and takes PV, energy storage cost, location and line expansion planning cost as the objective function to solve the problem. The lower-layer planning model uses self-organizing mapping method to select typical days of different load types and takes the minimum daily operating cost of distribution network as the objective function to solve the problem. Then, using the established two-stage planning method, taking the IEEE33-node distribution network system as an example, the influence of short-time peak load on the distribution network is analyzed. Finally, through scenario comparison, the necessity of considering the impact of short-term peak loads is further verified.

Abstract:The mean of power consumption elasticity coefficient is an important indicator that reflects the relationship between power con sumption and economic development, and used for power demand forecasting. At present, the calculation of the mean in China mostly adopts the arithmetic mean method, which is more susceptible to extreme year data fluctuations,and cannot avoid significant errors. To ob tain a more accurate mean, a time series fitting regression method based on the double logarithmic model is proposed. which selects“Gross Domestic Product”and“Electricity Consumption of the Whole Society”as samples to establish time series. and performs regression fitting analysis on them under the framework of linear relationships. By conducting tests on the regression model and comparing it with the geo metric mean method example, it is statistically demonstrated that the mean error of the power consumption elasticity coefficient calculated based on the double logarithmic model is smaller and more reliable.