A method for characterizing dynamic flexible domains of active distribution networks with massive distributed flexible resources

Abstract

The present invention discloses a method for characterizing the dynamic flexible domain of an active distribution network containing massive distributed flexible resources. In the method for characterizing the dynamic flexible domain of an active distribution network of the present invention, according to the modeling form of the distributed flexible resources connected to the active distribution network and the DistFlow flow model for processing the flow calculation process, an active distribution network flexible domain optimization model based on the DistFlow flow calculation is established to optimize the upper and lower limits of the power injected into the substation node. The QuickFlex algorithm is adopted to explore the area slight increase approximate search strategy of the flexible domain boundary to characterize the multi-time dynamic flexible domain of the active distribution network. The present invention takes into account the differences in the flexibility of the distributed flexible resources connected to the active distribution network, and visualizes the data of the distributed flexible resource operation strategy and the power grid flow by characterizing the multi-time dynamic flexible domain, thereby improving the intuitiveness of the expression of the flexibility of the active distribution network.

Description

Technical Field

[0001] The present invention belongs to the technical field of power distribution optimization operation, and relates to a method for characterizing a dynamic flexible domain of an active distribution network, and in particular to a method for optimizing power injected into a substation node in an active distribution network considering distributed flexible resources. Background Art

[0002] The construction of modern energy system requires that the proportion of distributed flexible power sources reaches about 24%, and emphasizes the further improvement of terminal electrical rate. This development goal poses a challenge to the control optimization technology and dispatching mechanism of power grid. As an emerging form of power system, active distribution network has attracted extensive attention from researchers and engineers. It is considered to be an important way to build smart grid with its characteristics of efficient use of distributed energy, integrated intelligent control and management. In order to achieve the planning goals, tapping and utilizing a large number of flexible resources to participate in grid regulation has become a key link in the construction of active distribution network. However, the current mainstream active distribution network optimization strategy tends to emphasize economic considerations, such as minimizing costs or losses, or focusing on the safety of distribution network trends, and cannot fully tap and utilize the flexibility potential of active distribution network brought by widely accessed distributed flexible resources. Therefore, while actively carrying out research and construction work related to active distribution network, in order to improve the flexibility and intuitive performance of active distribution network, it is necessary to understand the response ability of distribution network in response to various load changes from the perspective of power flexibility of distribution network, so as to provide theoretical basis and practical guidance for more efficient energy management and optimization.

[0003] In the development and research of distribution networks, distribution network domain theory (such as feasible domain, safety domain, etc.) as another important field besides network power flow problems has attracted the attention and research of many scholars. The research on these domain theories helps to provide theoretical support for the optimal dispatching of distribution networks. On the basis of ensuring the safe operation of distribution networks, the flexibility of distribution networks has received more and more attention, and the exploration of flexible domains has also been expanded. Nowadays, the main algorithms used for exploring the flexible domain of distribution networks are based on random sampling (Monte Carlo) methods, Constraint methods and methods based on iterative optimization algorithms. However, the Monte Carlo method requires a large number of random sampling to obtain statistically significant results. For complex power systems, a large number of simulations and optimization calculations need to be performed. Moreover, since this method relies on random sampling, the results may contain certain random errors, which will affect the accurate characterization of the flexible domain boundary. The constraint method will have a heavy computational burden when the system is large or when there are many objectives and constraints. There are high requirements for the setting of the value. If it is set too large or too small, it will not be able to accurately capture the flexible domain boundary of the system. Algorithms based on iterative optimization have gradually gained wide recognition and application, but in order to further improve the performance and efficiency of the algorithm, it is urgent to develop a more accurate and efficient flexible domain characterization algorithm that shows higher efficiency and stability in the iterative optimization process to meet the increasingly complex practical application needs.

[0004] In view of the problems mentioned in the background technology and the shortcomings of the prior art, the following problems need to be solved urgently:

[0005] 1) Widely distributed flexible resources are connected to the active distribution network. Most optimization schemes consider cost or loss, but have not yet optimized the flexibility of the distribution network by adjusting distributed flexible resources from the perspective of power flexibility.

[0006] 2) The mathematical model of the flexibility of the active distribution network is not intuitive enough, and it is impossible to visualize the data of the flexibility resource operation strategy and power grid flow to provide a more intuitive method for optimizing the operation strategy of the active distribution network.

[0007] 3) When exploring the flexibility of distribution networks, commonly used algorithms need to perform a large number of simulations and operations on complex power systems, which has a heavy computational burden and cannot accurately capture the flexible domain boundaries of distribution networks. Summary of the invention

[0008] The technical problem to be solved and the technical task proposed by the present invention are to improve and perfect the existing technical solutions, and provide a method for characterizing the dynamic flexible domain of an active distribution network containing massive distributed flexible resources, so as to achieve the purpose of improving the flexibility of the active distribution network and enhancing the intuitiveness of the flexibility of the distribution network. To this end, the present invention adopts the following technical solutions.

[0009] The characterization method of the present invention intuitively expresses the flexibility of the active distribution network in the form of a dynamic flexible domain, and improves the flexibility of the distribution network by using the active regulation capability of the active distribution network on distributed flexible resources. An active distribution network flexible domain optimization model based on DistFlow flow calculation is established to optimize the upper and lower limits of the power injected into the substation node. The dynamic flexible domain is characterized by dividing a typical day into 24 moments and solving it using the QuickFlex algorithm, and then using an iterative search method to characterize the dynamic flexible domain of the active distribution network.

[0010] A method for characterizing a dynamic flexible domain of an active distribution network containing massive distributed flexible resources includes the following steps:

[0011] Step (1) Build a distributed flexible resource model.

[0012] Due to the access of distributed flexible resources, the power range of substation nodes will become more flexible and changeable. Therefore, in order to accurately describe and predict this change, it is necessary to establish a corresponding flexible resource model based on the output characteristics and behavior patterns of different resources.

[0013] Step (2) Construct a wind-solar-load uncertainty scenario model. Using the scenario analysis method, based on historical observation data, the power boundary curves of wind power, photovoltaic power and curtailable load are obtained.

[0014] Step (3) proposes a flexible domain optimization model for active distribution network based on DistFlow power flow calculation and optimizes the maximum and minimum power injected into the substation node.

[0015] The model uses the second-order cone relaxation technique to convexly relax the power flow equation and transforms it into a linear problem using the linear programming method. The optimization goal of the model is to maximize the flexibility of the active distribution network, that is, to optimize the maximum and minimum active power and reactive power injected into the substation node.

[0016] Step (4) Use the QuickFlex algorithm to characterize the flexible domain of the active distribution network. This method uses the optimized maximum and minimum active power and reactive power injected into the substation node to construct the initial convex hull of the flexible domain. By exploring the area increment approximate search strategy of the flexible domain boundary, each line segment of the current convex hull is visited clockwise to iteratively search for new points that constitute the flexible domain and update the flexible domain interval until the algorithm stopping criteria are met to characterize the final flexible domain.

[0017] Step (5) divides a typical day into 24 moments, and uses the QuickFlex algorithm to characterize the flexible domain of the active distribution network state at each moment, and combines them to form a multi-moment dynamic flexible domain of the active distribution network.

[0018] The beneficial effects of the present invention are as follows:

[0019] (1) Different from the existing distribution network optimization that mostly considers cost or loss, the proposed active distribution network flexible domain optimization model takes the maximum and minimum values ​​of active and reactive power injected into the substation as the objective function. From the perspective of power flexibility, the flexibility of the distribution network is optimized by adjusting distributed flexible resources.

[0020] (2) As the mathematical model of the flexibility of the active distribution network is not intuitive enough, the concept of dynamic flexibility domain is proposed to visualize the data of the flexibility resource operation strategy and power grid flow to make it more intuitive.

[0021] (3) Compared with the efficiency and accuracy of existing algorithms for flexibility calculation, the use of the QuickFlex algorithm to characterize the flexible domain has a faster calculation speed and higher accuracy, providing a more reliable way to characterize the flexible domain. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A dynamic and flexible domain characterization framework diagram for an embodiment of the present invention;

[0023] Figure 2 A QuickFlex algorithm solution flow chart of an embodiment of the present invention;

[0024] Figure 3 A schematic diagram of the characterization process for a single-moment flexible domain;

[0025] Figure 4 The improved IEEE33 node active distribution network structure diagram is constructed;

[0026] Figure 5 To inject power optimization results into substation nodes;

[0027] Figure 6 This is the change curve of the slight increase rate of the flexible domain area at the 13th moment;

[0028] Figure 7 This is a schematic diagram showing that the algorithm meets the stopping criteria at the 13th moment;

[0029] Figure 8 The flexible domain diagram of the active distribution network depicted at the 8th, 13th and 19th moments. DETAILED DESCRIPTION

[0030] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings.

[0031] like Figure 1 As shown, the present invention comprises the following steps:

[0032] Step (1) Construct a distributed flexible resource model. Due to the access of distributed flexible resources, the power range of substation nodes will become more flexible and changeable. Therefore, in order to accurately describe and predict this change, it is necessary to establish a corresponding flexible resource model based on the output characteristics and behavior patterns of different resources.

[0033] Step (2) Construct a wind-solar-load uncertainty scenario model. Using the scenario analysis method, based on historical observation data, the power boundary curves of wind power, photovoltaic power and curtailable load are obtained.

[0034] Step (3) proposes a flexible domain optimization model for active distribution network based on DistFlow power flow calculation and optimizes the maximum and minimum power injected into the substation node. The model uses the second-order cone relaxation technology to convexly relax the power flow equation and transforms it into a linear problem using the linear programming method. The optimization goal of the model is to maximize the flexibility of the active distribution network, that is, to optimize the maximum and minimum active power and reactive power injected into the substation node.

[0035] Step (4) Use the QuickFlex algorithm to characterize the flexible domain of the active distribution network. This method uses the maximum and minimum values ​​of active and reactive power injected by the optimized substation nodes to construct the initial convex hull of the flexible domain. By exploring the area increment approximate search strategy of the flexible domain boundary, each line segment of the current convex hull is visited clockwise to iteratively search for new points that constitute the flexible domain and update the flexible domain interval until the algorithm stopping criteria are met to characterize the final flexible domain.

[0036] Step (5) divides a typical day into 24 moments, and uses the QuickFlex algorithm to characterize the flexible domain of the active distribution network state at each moment, and combines them to form a multi-moment dynamic flexible domain of the active distribution network.

[0037] The active distribution network flexibility domain at 24 moments of a typical day is depicted separately, with active power as axis, reactive power as Axis, time As Axis, each flexible domain is combined to depict the dynamic flexible domain of the three-dimensional active distribution network.

[0038] The present invention takes into account the differences between the outputs of distributed flexible resources at multiple times, generates a series of typical scenarios through cluster analysis technology, and describes the dynamic flexible domain of the active distribution network at multiple times. This characterization method makes the flexibility characteristics of the active distribution network more intuitive, thereby significantly improving the adaptability of the active distribution network under the background of widespread access to distributed flexible resources. In addition, this method not only helps to optimize resource allocation, but also improves the operating efficiency and stability of the system.

[0039] The technical solution of the present invention focuses on power flexibility, and realizes the flexibility optimization of the active distribution network integrating distributed flexible resources and the characterization of multi-time dynamic flexible domains. For the flexibility optimization problem, the output characteristics of each distributed flexible resource are analyzed and the optimization modeling is constructed accordingly; for the flexible domain characterization problem, the QuickFlex algorithm based on the iterative optimization algorithm is used to achieve characterization. The method has higher computational efficiency and accuracy, and provides a reference for the exploration of the flexibility of the active distribution network. This technical solution not only improves the flexibility management level of the active distribution network, making the flexibility of the distribution network more intuitive, but also provides a valuable reference for the future exploration of the flexibility of the distribution network.

[0040] The objective function of the flexible domain optimization model of the active distribution network is to maximize the flexibility of the distribution network, that is, to optimize the maximum and minimum active power and reactive power injected into the substation node. Its mathematical expression is:

[0041]

[0042]

[0043]

[0044]

[0045] , , , They are the minimum and maximum values ​​of active power and reactive power injected into the root node substation respectively;

[0046] for Time Node The active power of the connected photovoltaic power generation;

[0047] for Time Node The active power of the connected wind turbine;

[0048] , for Time Node The active and reactive power output by the connected micro gas turbine;

[0049] , for Time Node The active and reactive power of the connected load reduction;

[0050] , for Time Node The charging and discharging active power of the connected energy storage;

[0051] , They are Time Node Active and reactive power of connected substations;

[0052] , It is the active and reactive power injected into the root node substation.

[0053] The active distribution network flexible domain optimization model is built on the DistFlow power flow model, which includes its power flow calculation method and various power flow constraints. The constraints of the active distribution network flexible domain optimization model include:

[0054] 1) Power flow equation constraints;

[0055]

[0056]

[0057]

[0058] Among them, the collection For active distribution networks The set of head nodes of the branch that is the terminal node; For active distribution networks The end node set of the branch of the head node; and Respectively Time Branch Active and reactive power at the head end; and Respectively Time Branch Active and reactive power at the head end; for Time Node Voltage; for Time Node Voltage; , Node Active and reactive power of connected loads; , Line Resistance and reactance on.

[0059] First, perform SOCP convex relaxation to obtain:

[0060]

[0061]

[0062]

[0063]

[0064] in, , They are Timeline , The square of the upper current amplitude; , They are Time Node , The square of the voltage amplitude; , Line Resistance and reactance on.

[0065] Then after LP linear approximation, we get:

[0066]

[0067]

[0068]

[0069] Among them, the collection For active distribution network and node The set of all nodes connected by branches.

[0070] 2) Network security constraints;

[0071]

[0072]

[0073] in, , Node Voltage upper and lower limits; for Timeline The trend on For Line The upper limit of the trend.

[0074] 3) Flexible resource constraints;

[0075]

[0076]

[0077]

[0078]

[0079]

[0080] in, , Access Node The active power boundary of photovoltaic power generation; , Access Node The active power boundary of wind power generation; , Access Node The upper and lower limits of the active power of the micro gas turbine; , Access Node The upper and lower limits of the reactive power of the micro gas turbine; is the power factor angle of the load; Access Node The limit of load active power that can be reduced; , Access Node The upper and lower limits of the energy storage charging active power; , Access Node The upper and lower limits of the energy storage discharge active power; , Access Node Energy storage charge and discharge state 0-1 variable; for Time Node the capacity of the connected energy storage; , Access Node The charging and discharging efficiency of energy storage; Access Node The maximum capacity of energy storage; is the sampling time, defined as 1.

[0081] 4) Substation power constraints;

[0082]

[0083]

[0084] in, , For Node The upper and lower limits of active power of the connected substation; , For Node The upper and lower limits of the reactive power of the connected substation.

[0085] Figure 2 A flowchart of the QuickFlex algorithm for describing the flexible domain is given. Figure 3 A schematic diagram of the QuickFlex algorithm characterizing the flexible domain process is given. The QuickFlex algorithm specifically includes the following steps:

[0086] Step 1: Initialize and construct the initial convex hull.

[0087] First, data is read and initialized. Read the network flow data of the active distribution network and the available flexibility resource data. Initialize the flexible area for , and set a tolerance error .

[0088] Secondly, construct the initial convex hull. The maximum and minimum active power and reactive power injected into the substation node obtained after optimization in step 3, and their operating states correspond to the four "active-reactive" feasible power vertices of the required flexible domain. The four vertices are connected to form an initial convex hull, which is the convex hull Initialization.

[0089]

[0090] in, The flexible domain depicted is the convex hull; To constitute the first flexible domain of active distribution network A feasible operating point, that is, a feasible active-reactive power point injected into the substation node; It is the actual flexible domain of the active distribution network.

[0091] Step 2: Sequential Update .

[0092] Visit each line segment of the current convex hull in a clockwise direction and find an optimal point for each line segment ( ). This point maximizes the vertical distance from the line segment to the feasible flexible domain. At the optimal point ( ) creates two new line segments between the two endpoints of the line segment and , and expand the flexible domain coverage.

[0093]

[0094] in, The distance from the search point to the current search segment; For the The best point found.

[0095] The step of maximizing the perpendicular distance is to find a new vertex on the boundary of the convex hull that will increase the coverage of the convex hull. The "perpendicular distance" is the shortest distance from the current line segment to a point that is outside the current convex hull and within the feasible flexible domain. To find the point that maximizes the perpendicular distance, the QuickFlex algorithm will perform the following steps:

[0096] Select the current line segment: Starting from a vertex of the convex hull, select a line segment in a clockwise direction. This line segment is defined by two adjacent vertices of the convex hull.

[0097] Calculate the perpendicular distance: For each point outside the current convex hull and within the feasible flexible domain, calculate its perpendicular distance to the current line segment. The perpendicular distance is calculated using the point-to-line distance formula, which takes into account the coordinates of the point and the endpoints of the line segment.

[0098]

[0099] in, , They are the horizontal and vertical coordinate values ​​of the search point respectively; is the distance from the search point to the line segment; , are the slope and intercept of the search segment respectively;

[0100] Maximize vertical distance: Search for a point in the feasible flexible domain outside the current convex hull so that the vertical distance from the point to the current line segment is the largest, in order to maximize the coverage of the selected feasible operating point. The maximization search is completed by selecting the optimal point, with the goal of maximizing the vertical distance, while satisfying the constraints of the feasible flexible domain of the active distribution network.

[0101] Determine the new vertex: After finding the point that maximizes the vertical distance, use this point as the new vertex of the convex hull. This point not only maximizes the distance to the current line segment, but also still meets the operational constraints of the active distribution network, such as node voltage output constraints, line flow constraints, etc.

[0102] Update convex hull : Add the new vertex to the convex hull, create two new line segments between the new vertex and the two endpoints of the corresponding line segment, and update the boundary of the convex hull.

[0103] Continue searching: Repeat the above steps, searching for the optimal point for each line segment of the convex hull in a clockwise direction. After completing one round of clockwise search, perform the second iteration, and perform the same operation for each new line segment in a clockwise direction until the stopping criteria is met.

[0104] Update by sequence , the flexibility area of ​​the active distribution network can be gradually constructed, and each iteration increases the coverage of the convex hull, thereby more comprehensively capturing the potential flexibility of the distribution network.

[0105] Step 3: Determine the stopping criteria and eliminate the line segment search.

[0106] Use the area increment approximate search strategy to update the convex hull range each time the optimal point is found, and calculate the area increment rate between the new convex hull and the previous convex hull. .

[0107]

[0108] in, To generate the The area increment rate when a convex hull is generated, which is used for stopping criteria and line segment search elimination; For the The area of ​​the convex hull.

[0109] If the optimal point cannot be found for a line segment, the area increment rate of the flexible area is greater than the tolerance error. ,Right now , no further search is performed for that line segment. Repeat the sequential update for the next line segment The operation is repeated until all line segments have completed the search for the optimal points and finally a flexible domain is formed.

[0110] Step 4: Output the final characterization of the flexible domain.

[0111] The result output by the QuickFlex algorithm is a flexible domain formed by a set of points that meet the algorithm's stopping criteria. In addition, the QuickFlex algorithm always selects new points from the boundary of the feasible flexible domain of the distribution network, and does not select points inside or outside the feasible flexible domain, which has the fast calculation speed.

[0112] This embodiment is further described in detail below in conjunction with a specific implementation plan. This embodiment uses an IEEE 33-node distribution network connected to various distributed flexible resources for simulation. Figure 4 The grid structure of the system is given. The distribution network includes 3 photovoltaic generators, 3 wind generators, 4 micro gas turbines and 2 energy storage devices. Among them, the photovoltaic generators are connected to nodes 14, 25 and 30; the wind generators are connected to nodes 7, 17 and 20; the micro gas turbines are connected to nodes 5, 11, 22 and 33; and the energy storage devices are connected to nodes 10 and 28. The output power limits of each distributed power source are shown in Table 1. Among them, the upper limit of photovoltaic and wind power output is the predicted value of the real-time power of the unit. The annual fluctuations of wind speed, light intensity and load that can be reduced are taken from historical statistical data.

[0113] Table 1 Distributed power generation power limit

[0114]

[0115] Figure 5The optimized upper and lower limits of active and reactive power injected into the substation node obtained by the flexible domain optimization model of the active distribution network based on DistFlow power flow calculation are given. On the basis of meeting the network security of the active distribution network, the access to various distributed flexible resources brings great flexibility to the operation of the active distribution network. When the output of the distributed power source is at a low level, the active and reactive power injected into the substation node will reach its upper limit. In this scenario, the substation needs to increase its output power to ensure the normal operation of the distribution network load. However, if multiple distributed power sources are connected to the distribution network and the installed capacity of these power sources is sufficient to support the load demand of the distribution network, the active distribution network will have the ability to reversely transmit power to the substation, so that the active and reactive power injected into the substation node reaches or approaches its lower limit. At this time, the active distribution network essentially plays the role of a power station, providing power output to the upper power grid. Among them, the fluctuation of active power is mainly affected by load changes. As the load increases, in order to meet the load demand, the substation needs to increase its output accordingly, which leads to an increase in the upper limit of the substation's active power. At the same time, in the case of such load growth, the distributed flexible power source first ensures that the load's power supply needs are met, and then outputs the remaining power to the superior power grid. This priority of power supply sequence results in that when the load increases, the lower limit of the substation's active power also increases accordingly to ensure the stable operation of the power grid.

[0116] Figure 6 The curve of area slight increase rate at the 12th moment is given. Figure 7 The algorithm convergence diagram that meets the stopping criteria is further given. The tolerance error is set to ,After five rounds of iterative calculations by the QuickFlex algorithm, all ,search segments satisfy the algorithm segment search elimination, thus meeting the ,algorithm stopping criteria, and it takes 273 s to generate 94 operable ,operation points to form the flexible domain of the active ,distribution network at that moment, which verifies the feasibility and efficiency of the ,algorithm.

[0117] Figure 8The flexible domain diagrams of the active distribution network depicted at the 8th, 13th, and 19th moments are given. The boundary characteristics of the flexible domain of the substation node power at different moments are analyzed. The shape of the flexible domain boundary at time 8 presents a relatively regular shape, and the boundary line is relatively smooth. This boundary shape reflects that the feasible operating range of the substation node power at this moment is relatively balanced and within a moderate range. At the 13th moment, due to the significant increase in the output of distributed power sources such as photovoltaics and wind power, the shape of the flexible domain boundary becomes more complex, and its boundary range is also expanded accordingly, reflecting that the active distribution network has higher flexibility at this stage. Further observation of the flexible domain boundary at time 19 shows that due to the peak load, but the output of distributed power sources is reduced, its boundary range is wider than that at time 8, but compared with the 13th moment, its boundary shape appears to be more stable. The characterization of multi-moment dynamic flexible domains not only shows the intuitive single-moment flexible domain characteristics, but also can deeply explore the dynamic changes of the flexibility of the active distribution network at different moments. This characterization method provides a comprehensive perspective to analyze and understand the flexible adjustment capabilities of the active distribution network at different time scales.

[0118] In summary, the method proposed in this patent optimizes the flexibility of the active distribution network connected to distributed flexible resources, optimizes the resource configuration of the distribution network, and significantly improves the adaptability of the active distribution network. At the same time, it depicts a more intuitive multi-time dynamic flexible domain, provides a flexible domain characterization algorithm with higher computational efficiency and higher accuracy, and provides a new visualization idea for the research and exploration of the flexibility of the active distribution network.

[0119] above Figure 1 The method for characterizing a dynamic flexible domain of an active distribution network containing massive distributed flexible resources shown is a specific embodiment of the present invention, which has embodied the substantial characteristics and progress of the present invention. According to actual use needs and under the guidance of the present invention, equivalent modifications in shape, structure, etc. can be made to the method, which are all within the protection scope of this scheme.

Claims

1. A method for characterizing the dynamic flexible domain of an active distribution network containing massive distributed flexible resources, characterized in that: The following steps are involved: Step (1) constructing a distributed flexible resource model; Establish corresponding flexible resource models based on the output characteristics and behavior patterns of different resources; Step (2) constructing a wind-solar-load uncertainty scenario model; using a scenario analysis method, based on historical observation data, obtaining power boundary curves of wind power, photovoltaic power, and load that can be reduced; Step (3) proposes a flexible domain optimization model for active distribution network based on DistFlow power flow calculation and optimizes the maximum and minimum power injected into the substation node; The model uses the second-order cone relaxation technique to convexly relax the power flow equation and transforms it into a linear problem using the linear programming method. The optimization goal of the model is to maximize the flexibility of the active distribution network, that is, to optimize the maximum and minimum active power and reactive power injected into the substation node. Step (4) Use the QuickFlex algorithm to characterize the flexible domain of the active distribution network; this method uses the optimized maximum and minimum active power and reactive power injected into the substation node to construct the initial convex hull of the flexible domain; through the area slight increase approximate search strategy of exploring the boundary of the flexible domain, each line segment of the current convex hull is visited clockwise to iteratively search for new points constituting the flexible domain and update the flexible domain interval until the algorithm stopping criteria are met to characterize the final flexible domain; Step (5) divides a typical day into 24 moments, uses the QuickFlex algorithm to characterize the flexible domain of the active distribution network state at each moment, and combines them to form a multi-moment dynamic flexible domain of the active distribution network; The objective function of the flexible domain optimization model of the active distribution network is to maximize the flexibility of the distribution network, that is, to optimize the maximum and minimum active power and reactive power injected into the substation node. Its mathematical expression is: P 0,t , Q 0,t , They are the minimum and maximum values ​​of active power and reactive power injected into the root node substation respectively; is the active power of photovoltaic power generation connected to node i at time t; is the active power of wind power connected to node i at time t; is the active and reactive power output by the micro gas turbine connected to node i at time t; is the active and reactive power of the load reduction connected to node i at time t; is the charging and discharging active power of the energy storage connected to node i at time t; P i,t , Q i,t are the active and reactive power of the substation connected to node i at time t respectively; P 0,t , Q 0,t The active and reactive power injected into the root node substation; The QuickFlex algorithm is used to characterize the flexible domain of the active distribution network, which specifically includes the following steps: Step 1: Initialize and construct the initial convex hull; First, data is read and initialized; network flow data and available flexibility resource data of the active distribution network are read; the flexible area R0 is initialized as And set a tolerance error ε; Secondly, construct the initial convex hull; the maximum and minimum active power and reactive power injected into the substation node obtained after optimization in step 3, and their operating states correspond to the four "active-reactive" feasible power vertices of the required flexible domain, and the four vertices are connected to form an initial convex hull, which is the initialization of the convex hull R; Among them, R represents the flexible domain depicted, namely the convex hull; (p n ,q n ) is the nth feasible operation point constituting the flexible domain of the active distribution network, i.e., the feasible active-reactive power point injected into the substation node; F is the actual flexible domain of the active distribution network; Step 2: Update R sequentially; Visit each line segment of the current convex hull in a clockwise direction and find an optimal point (p k ,q k ); This point maximizes the vertical distance from the line segment to the feasible flexible domain; at the optimal point (p k ,q k ) creates two new line segments (p k ,q k )-(p n ,q n ) and (p k ,q k )-(p n+1 ,q n+1 ), and expand the coverage of the flexible domain R; Where d is the distance from the search point to the current search segment; (p k ,q k ) is the kth optimal point searched; The step of maximizing the perpendicular distance is to find a new vertex on the boundary of the convex hull that will increase the coverage of the convex hull; "perpendicular distance" refers to the shortest distance from the current line segment to a point outside the current convex hull and within the feasible flexible domain; to find the point that maximizes the perpendicular distance, the QuickFlex algorithm will perform the following steps: Select the current line segment: Starting from a vertex of the convex hull, select a line segment in a clockwise direction; this line segment is defined by two adjacent vertices of the convex hull; Calculate the perpendicular distance: For each point outside the current convex hull and within the feasible flexible domain, calculate its perpendicular distance to the current line segment; the perpendicular distance is calculated using the point-to-line distance formula, which takes into account the coordinates of the point and the endpoints of the line segment; Among them, P0 and Q0 are the horizontal and vertical coordinate values ​​of the search point respectively; d is the distance from the search point to the line segment; k and b are the slope and intercept of the search line segment respectively; Maximize vertical distance: Search for a point in the feasible flexible domain outside the current convex hull so that the vertical distance from the point to the current line segment is the largest. The purpose is to maximize the coverage of the selected feasible operating point. The maximization search is completed by selecting the optimal point. The goal is to maximize the vertical distance, while satisfying the constraints of the feasible flexible domain of the active distribution network. Determine the new vertex: After finding the point that maximizes the vertical distance, use this point as the new vertex of the convex hull; Update the convex hull R: add the new vertex to the convex hull, create two new line segments between the new vertex and the two endpoints of the corresponding line segment, and update the boundary of the convex hull; Continue searching: Repeat the above steps to search for the optimal point for each line segment of the convex hull in a clockwise direction; after completing one round of clockwise search, perform the second iteration and perform the same operation for each new line segment in a clockwise direction until the stopping criteria is met; Step 3: Determine the stopping criteria and eliminate the line segment search; Use the area increment approximate search strategy to update the convex hull range each time the optimal point is found, and calculate the area increment rate between the new convex hull and the previous convex hull. in, is the area increment rate when generating the kth convex hull, which is used for stopping criteria and line segment search elimination; H k is the area of ​​the kth convex hull; If the optimal point cannot be found for a line segment, the area increment rate of the flexible region is greater than the tolerance error ε, that is, Then no further search is performed on the line segment; the operation of sequentially updating R is repeated for the next line segment until all line segments have completed the search for the optimal point, and finally a flexible domain is formed; Step 4: Output the final characterization flexible domain; The output of the QuickFlex algorithm is a flexible domain formed by a set of points that meet the algorithm's stopping criteria.

2. The method for characterizing a dynamic flexible domain of an active distribution network containing massive distributed flexible resources according to claim 1 is characterized in that: The flexible domain optimization model of the active distribution network is established on the DistFlow power flow model, which includes its power flow calculation method and various power flow constraints; the constraints of the flexible domain optimization model of the active distribution network include power flow equality constraints, network security constraints, flexible resource constraints and substation power constraints.

3. The method for characterizing a dynamic flexible domain of an active distribution network containing massive distributed flexible resources according to claim 2 is characterized in that: The power flow equation constraints are as follows: Wherein, set a(i) is the set of head nodes of the branch with i as the terminal node in the active distribution network; set b(i) is the set of terminal nodes of the branch with i as the head node in the active distribution network; P hi,t With Q hi,t They represent the active and reactive power at the head end of branch hi at time t respectively; P ij,t With Q ij,t Respectively represent the active and reactive power at the head end of branch ij at time t; V h,t is the voltage of node h at time t; V i,t is the voltage of node i at time t; are the active and reactive power of the load connected to node i; r ij 、x ij are the resistance and reactance on line ij respectively; r hi 、x hi are the resistance and reactance on line hi respectively; First, perform SOCP convex relaxation to obtain: v j,t =v i,t -2(r ij P ij,t +x ij Q ij,t )+[(r ij ) 2 +(x ij ) 2 ]l ij,t Among them, l hi,t , l ij,t are the squares of the current amplitudes on lines hi and ij at time t; v i,t 、v j,t are the squares of the voltage amplitudes at nodes i and j at time t respectively; Then after LP linear approximation, we get: v i,t -v j,t =2(r ij P ij,t +x ij Q ij,t ) The set c(i) is the set of all nodes in the active distribution network connected to the node i through branches.

4. The method for characterizing a dynamic flexible domain of an active distribution network containing massive distributed flexible resources according to claim 2 is characterized in that: The network security constraints are as follows: in, V i are the upper and lower limits of the voltage at node i, respectively; I ij,t is the power flow on line ij at time t; is the upper limit of the power flow on line ij.

5. The method for characterizing a dynamic flexible domain of an active distribution network containing massive distributed flexible resources according to claim 2 is characterized in that: Flexible resource constraints are as follows: in, are the active power boundaries of photovoltaic power generation connected to node i respectively; are the active power boundaries of wind power generation at access node i respectively; are the upper and lower limits of the active power of the micro gas turbine connected to node i; are the upper and lower limits of the reactive power of the micro gas turbine connected to node i; γ is the power factor angle of the load; is the active power boundary of the load that can be reduced at access node i; are the upper and lower limits of the energy storage charging active power connected to node i, respectively; are the upper and lower limits of the energy storage discharge active power of access node i respectively; is the energy storage charging and discharging state 0-1 variable of access node i; E i,t is the capacity of the energy storage connected to node i at time t; is the charging and discharging efficiency of the energy storage connected to node i; is the maximum capacity of the energy storage connected to node i; Δt is the sampling time, which is defined as 1.

6. The method for characterizing a dynamic flexible domain of an active distribution network containing massive distributed flexible resources according to claim 2 is characterized in that: The substation power constraints are as follows: in, P i is the upper and lower limits of the active power of the substation connected to node i; Q i are the upper and lower limits of reactive power of the substation connected to node i.

Images