Evaluation method for flexibility injection domain of distribution network considering distributed energy integration

Abstract

The present invention relates to a distribution network flexibility injection safety domain evaluation method considering the access of distributed energy. The method consists of three parts. First, based on the historical data of distributed energy, a distributed energy scenario generation method considering nonlinear correlation is used to obtain scenario samples of distributed energy output; secondly, a non-iterative power flow calculation method is used to quickly screen unsafe scenarios; finally, an optimization model of distributed energy injection power margin is constructed for unsafe scenarios, and the distributed energy injection power safety domain is characterized by comprehensive safety scenarios. On the basis of this distribution network operation safety domain, the flexibility is evaluated by the minimum boundary distance between the online operation point and the safety domain, and it is used as a flexibility indicator.

Description

Technical Field

[0001] The present invention belongs to the technical field of evaluation of the flexibility injection domain of a distribution network, and in particular to an evaluation method of the flexibility injection domain of a distribution network to which distributed energy is connected. Background Art

[0002] The construction and development of distribution networks have always played a vital role in meeting social and economic development and people's quality of life. With the transformation of global energy and the diversification of user needs, distribution networks are tending to be built in an advanced form of intelligence and low carbon. In order to adapt to the economic development of the user side and cope with the challenges brought by the energy transformation on the power supply side, flexible loads such as multiple distributed power sources and electric vehicles are gradually connected to the distribution network, gradually realizing the function of interactive supply and demand between users and power grids. At the same time, the structure of the distribution network has gradually changed into a multi-power closed-loop network, and the operation of the distribution network has also shown strong randomness and volatility, which reduces the accuracy of traditional distribution network operation analysis methods. For example, the traditional dispatching method is to calculate the acceptable capacity offline based on the predicted output results and perform corresponding dispatching control. However, for distribution networks containing distributed energy, since the offline calculation results depend on the prediction accuracy of power and load, it is impossible to reflect the acceptance capacity of distributed power sources at the current moment in real time, resulting in a deviation between the predicted value and the actual value, affecting the safety assessment and resource utilization efficiency of system operation. In the strong uncertainty environment of distributed energy access, the system operation mode is diversified, and a comprehensive evaluation of the system operation mode is required to grasp the safety, economy and applicability of the power grid planning scheme.

[0003] Therefore, the flexibility of distribution networks has been taken into consideration in the current research on distribution network operation optimization, planning, dispatching, new energy consumption strategy and energy storage optimization, and methods and strategies that are more suitable for the development of modern distribution networks have been proposed. [2-6] The research results also reflect the flexibility requirements of distribution networks at different stages of planning and operation. Therefore, effective flexibility assessment technology can help provide an important information basis for the development and research of distribution networks.

[0004] At present, the definition of distribution network flexibility has not been completely unified. The international definition of power system flexibility emphasizes the response and change ability of the power system. For example, the International Energy Agency (IEA) believes that the flexibility of the power system is reflected in the system's ability to adjust power generation or load to maintain reliability when facing large disturbances. The North American Electric Reliability Council (NERC) defines power system flexibility as the system's ability to respond to changes in supply and load. Existing studies believe that the definition of power system flexibility should include four elements: time scale, flexibility resource set, system uncertainty, and cost constraints. Furthermore, existing studies have designed flexibility evaluation indicators for the power supply side, distribution network side, and load side. Or considering the equipment climbing capability, transmission capability, and regulation capability of the distribution network, the grid flexibility is divided into equipment level, network level, and system level. In the existing research on the evaluation method of distribution network flexibility, the focus has been on considering the problem of strong randomness and volatility in the distribution network, and the evaluation method of flexibility has been proposed from the perspective of multiple scenarios. However, distributed energy scenarios have certain correlations in practice, such as the complex correlations between different wind farm outputs. Therefore, directly using scenario reduction or typical scenario extraction may lead to inaccurate and incomplete distribution network flexibility evaluation results. Currently, few studies on distribution network flexibility evaluation consider distributed energy output scenario generation methods. Summary of the invention

[0005] The purpose of the present invention is to supplement the existing distribution network flexibility assessment technology, and propose an assessment method for the distribution network flexibility injection domain considering the access of distributed energy resources, which can achieve the scene sample of distributed energy output based on the historical data of distributed energy resources and the distributed energy scenario generation method considering nonlinear correlation; secondly, the non-iterative power flow calculation method is used to quickly screen unsafe scenes; finally, an optimization model of the distributed energy injection power margin is constructed for unsafe scenes, and the distributed energy injection power safety domain is characterized by comprehensive safety scenarios. On the basis of this distribution network operation safety domain, the flexibility is evaluated by the minimum boundary distance of the online operation point from the safety domain, and it is used as a flexibility indicator.

[0006] The present invention solves the technical problem by adopting the following technical solutions:

[0007] The evaluation method of distribution network flexibility injection security domain considering distributed energy access includes the following steps:

[0008] Step 1: Obtain historical measurement data sets of wind turbines and photovoltaics as input data;

[0009] Step 2: Model the distribution and correlation of the historical measurement data of the output of distributed energy resources by using appropriate Copula;

[0010] Step 3: Based on the Copula model obtained in step 2, LHSD is applied to generate output scenario samples of distributed energy;

[0011] Step 4: Use a power flow calculation method without iteration to quickly analyze the operating status of the distribution system under the distributed energy output scenario obtained in step 3, and judge whether the operating status of the distribution system under different distributed energy output scenarios is safe according to the distribution network safety verification index, and select safe scenarios and unsafe scenarios according to the judgment results;

[0012] Step 5: Optimize and calculate the maximum access power value of the unsafe scenario selected in step 4 to obtain the safety boundary of the distribution network operation, and on this basis, obtain the index of the flexibility of the distributed energy access capability;

[0013] Step 6: Combine the safe scenarios selected by the power flow calculation method without iteration in step 4 with the safety boundaries of the unsafe scenarios calculated in step 5 to characterize the safe operation domain of the distribution network to which distributed energy is connected.

[0014] Furthermore, the step 2 comprises the following steps:

[0015] Step 2.1, select the appropriate R-vine structure based on the Kendall-τ rank correlation coefficient and the maximum spanning tree MST algorithm;

[0016] Step 2.2: According to the correlation of the distributed energy data in step 2.1, the model constructed by calculating the alternative Copula function is compared with the empirical Copula function to select a suitable Copula function and estimate the corresponding model parameters;

[0017] Step 2.3: According to step 2.2, build Copula layer by layer for the R-vine structure selected in step 2.1, and finally form the best correlation model.

[0018] Moreover, the application of LHSD in step 3 to generate distributed energy output scenario samples includes stratified sampling, order quantity statistics and selection arrangement. For uniform random distribution U~U(0,1), the d-dimensional uniform distribution connected by any Copula i=1,…,n represents n are samples drawn independently from . It represents the order statistic of the i-th sample in the j-th dimension, which is calculated as follows: sort these n samples, expressed as x (1) <…<x (n) , then Ui The order statistics of is:

[0019]

[0020] Where n represents the number of samples and I is an indicator function, that is, if the condition in the brackets is met, I is 1, otherwise I is 0.

[0021] Then a sample of LHSD sampling can be expressed as:

[0022]

[0023] Among them, W i j Represents a sample. When considering obtaining joint distribution characteristics, The value of is usually 0.5 or 1.

[0024] 3. Moreover, the power flow calculation model without iteration adopted in step 4 is:

[0025] Consider the three-phase relationship between the three-phase voltage and current of the three-phase branch:

[0026]

[0027] Among them, I i and V i Represent the current and voltage vectors at the transmitting end, I j and V j represents the current and voltage vectors at the output, and the admittance matrix is ​​composed of Y ii , Y jj , Y ji , Y ij For a line without parallel admittance to ground, Y ii =Y ij =Z -1 , Y ij =Y ji =-Z -1 , Z is the impedance matrix of the corresponding phase;

[0028] For single-phase and two-phase lines, remove the elements corresponding to the non-existent phases. For lines with parallel admittance to ground, use the three-phase π model, Y ii =Y ij =Z -1 +Y / 2,Y ij =Y ji =-Z -1 ,The admittance model of the three-phase transformer adopts the traditional three-phase transformer model;

[0029] Current model of the source and output ends:

[0030] I i =Y ii V i +Y ij V j

[0031] I j =Y ji V i +Y jj V j

[0032] The output current model considering the output voltage is:

[0033]

[0034] in, Yes jj The pseudo-inverse of

[0035] The source current model is suitable for any type of line:

[0036] I i =S i I j +s i

[0037] in, When there is no transformer or parallel ground admittance in the line, s i = 0, and S i is the identity matrix corresponding to the phase; otherwise, and ignore s i ;

[0038] Starting from the end node k, the relationship model between the currents of each node in the entire network is:

[0039] I k-1 =S k I k

[0040] I k-2 =S k-1 (S k I k +I k-1 )

[0041] I k-3 =S k-2 (S k-1 S k I k +S k-1 I k-1 +I k-2 )

[0042] Among them, Ik =I load,k +I cap,k , which means that the equivalent injection current of the end node k of a line can be injected by the load current I load,k and the capacitor injection current I cap,k Indicates that node k-1 is the parent node of node k, node k-2 is the parent node of node k-1, and so on;

[0043] The current model at any node l is:

[0044]

[0045] Where ξ is the set of child nodes i including node l and the downstream branches directly connected to node l, H l,i =S l S l-1 …S i-1 S i It means that the current of the load and capacitor on node i is converted into the equivalent current passing through node l. The S matrix is ​​calculated from the admittance matrix of the branch from node l to node i. l represents the sum of the current injections of the child nodes of node l;

[0046] Voltage model of the downstream node j of the branch:

[0047] V j =V i -(R ij +jX ij )I i

[0048] Among them, V i is the upstream node voltage, I i is the upstream branch current;

[0049] For a branch with a transformer, the voltage model of the downstream node j is:

[0050]

[0051] in,[·] + Represents the pseudoinverse of a matrix.

[0052] The above calculation formula can replace Newton iteration, and the power flow solution in each scenario can be calculated by analytical expression to obtain node voltage and line power. This avoids the process of inverting the Jacobian matrix in the traditional Newton calculation method, reduces the calculation burden, and realizes the rapid verification of the safety of the generated scenario.

[0053] The distribution network operation safety verification index in step 4 can be defined as: voltage over-limit safety verification index: whether voltage over-limit occurs in a given distributed energy generation scenario; and thermal limit over-limit safety verification index: whether branch thermal limit over-limit occurs in a given distributed energy generation scenario.

[0054] Moreover, the model for calculating the maximum access power value of the unsafe scenario screened in step 4 in step 5 is:

[0055] maxλ

[0056] stf(x,μ,λ)=0

[0057]

[0058] Among them, f(x,μ,λ)=f(x,μ)+λb=0 is the parameterized power flow equation, x is the voltage amplitude and phase angle of the unbalanced node, that is, the state variable, μ is the active and reactive injection power of the generator node, that is, the control vector, and b is the change vector of the injection power, that is, the direction vector of the injection power change in the unsafe scenario. When λ=0, the parameterized power flow equation is the ground state power flow equation. The constraint vector equation f(x,μ,λ)=0 means that the solution to the problem must satisfy the parameterized power flow equation; the constraint vector equation The voltage amplitude of all nodes is required to be within a given per-unit range, typically 0.9 to 1.1 or 0.95 to 1.05 per-unit; the constraint vector equation Indicates that the current of all distribution lines and transformer branches must be less than the given current limit; constraint vector equation It is required that the reactive output of all distributed power sources must be within the given upper and lower limits.

[0059] Moreover, the specific implementation method of step 6 includes the following steps:

[0060] Step 6.1, combining the safe output scenario obtained by performing safety verification on the scenario generated in step 3 according to the power flow calculation method without iteration in step 4 and the safety domain boundary of the unsafe scenario calculated by the optimization model in step 5, that is, the calculated corresponding maximum injection power point, to achieve the characterization of the safe operation domain of the distribution network to which distributed energy is connected;

[0061] Step 6.2: According to the safe operation domain of the distribution network obtained in step 6.1, select an online operating point, calculate the minimum distance between the operating point and the safety boundary, obtain the flexibility index of the operating point, and realize the flexibility evaluation of the operating point.

[0062] The advantages and positive effects of the present invention are:

[0063] The scenario generation method of the present invention taking into account the correlation of wind power output scenarios can generate samples of the relationship between actual scenarios, and the non-iterative power flow calculation method based on this data can maintain the accuracy of scenario screening and improve the evaluation speed. Under different operating constraints, by characterizing the safe injection domain of the distribution network connected to distributed energy, the flexibility of the online operating point can be effectively evaluated. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 Generate a flow chart for the output scenario of the present invention;

[0065] Figure 2 A schematic diagram of the safety domain for injecting flexibility into distributed energy in the present invention;

[0066] Figure 3 It is the 123 node structure diagram of the present invention and the access location of distributed energy;

[0067] Figure 4 The actual output scenario diagram of the wind turbines located at nodes 88 and 95;

[0068] Figure 5 The wind turbine output scenario diagram at nodes 88 and 95 is generated based on the Copula function and the LHSD method;

[0069] Figure 6 The wind turbine output scenario diagram at nodes 88 and 95 generated based on the LHS method;

[0070] Figure 7 It is the difference between the calculation results of the power flow calculation method without iteration and the Newton power flow algorithm;

[0071] Figure 8 The power safety domain is injected for the flexibility of nodes 88 and 95 under different constraints. DETAILED DESCRIPTION

[0072] The present invention is further described in detail below with reference to the accompanying drawings.

[0073] The evaluation method of injecting flexibility into the security domain considering distributed energy access, such as Figure 1 and Figure 2 As shown, the following steps are included:

[0074] Step 1: Obtain historical measurement data sets of wind turbines and photovoltaics as input data.

[0075] Step 2: Model the distribution and correlation of historical measurement data of distributed energy output using appropriate Copula.

[0076] This step includes the following steps:

[0077] Step 2.1, select the appropriate R-vine structure based on the Kendall-τ rank correlation coefficient and the maximum spanning tree MST algorithm;

[0078] Step 2.2: According to the correlation of the distributed energy data in step 2.1, the model constructed by calculating the alternative Copula function is compared with the empirical Copula function to select a suitable Copula function and estimate the corresponding model parameters;

[0079] Step 2.3: According to step 2.2, build Copula layer by layer for the R-vine structure selected in step 2.1, and finally form the best correlation model.

[0080] Step 3: Based on the Copula model obtained in step 2, LHSD is used to generate output scenario samples of distributed energy.

[0081] The specific implementation methods of this step are divided into stratified sampling, order quantity statistics and selection arrangement. For uniform random distribution U~U(0,1), the d-dimensional uniform distribution connected by any Copula Represented as n from are samples drawn independently from . It represents the order statistic of the i-th sample in the j-th dimension, which is calculated as follows: sort these n samples, expressed as x (1) <…<x (n) , then U i The order statistics of is:

[0082]

[0083] Where n represents the number of samples and I is an indicator function, that is, if the condition in the brackets is met, I is 1, otherwise I is 0.

[0084] Then a sample of LHSD sampling can be expressed as:

[0085]

[0086] Among them, W i j Represents a sample. When considering obtaining joint distribution characteristics, The value of is usually 0.5 or 1.

[0087] Step 4: Use a power flow calculation method without iteration to quickly analyze the operating status of the distribution system under the distributed energy output scenario obtained in step 3, and judge whether the operating status of the distribution system under different distributed energy output scenarios is safe based on the distribution network safety verification indicators, and screen out safe and unsafe scenarios based on the judgment results.

[0088] The power flow calculation model without iteration adopted in step 4 is:

[0089] Consider the three-phase relationship between the three-phase voltage and current of the three-phase branch:

[0090]

[0091] Among them, I i and V i Represent the current and voltage vectors at the transmitting end, I j and V j represents the current and voltage vectors at the output, and the admittance matrix is ​​composed of Y ii , Y jj , Y ji , Y ij For a line without parallel admittance to ground, Y ii =Y ij =Z -1 , Y ij =Y ji =-Z -1 , Z is the impedance matrix of the corresponding phase. For single-phase and two-phase lines, remove the elements corresponding to the non-existent phases. For lines with parallel admittance to ground, the three-phase π model is used, Y ii =Y ij =Z -1 +Y / 2,Y ij =Y ji =-Z -1 The admittance model of the three-phase transformer adopts the traditional three-phase transformer model.

[0092] Current model of the source and output ends:

[0093] I i =Y ii V i +Y ij V j

[0094] I j =Y ji V i +Y jj V j

[0095] The output current model considering the output voltage is:

[0096]

[0097] in, Yes jj The pseudo-inverse of Considering the transformers with different wiring modes in the line, the matrix Y jjThere will be strange situations, for example, if the primary side of the transformer is grounded star connection, and the secondary side is ungrounded delta connection, then Y jj The admittance matrix is ​​a singular matrix.

[0098] Output current model suitable for any type of line:

[0099] I i =S i I j +s i

[0100] in, When there is no transformer or parallel admittance to ground in the line, it is not a transformer or there is no line charging, then s i = 0, and S i is the identity matrix corresponding to the phase; otherwise, and ignore s i .

[0101] Starting from the end node k, the relationship model between the currents of each node in the entire network is:

[0102] I k-1 =S k I k

[0103] I k-2 =S k-1 (S k I k +I k-1 )

[0104] I k-3 =S k-2 (S k-1 S k I k +S k-1 I k-1 +I k-2 )

[0105] …

[0106] Among them, I k =I load,k +I cap,k , which means that the equivalent injection current of the end node k of a line can be injected by the load current I load,k and the capacitor injection current I cap,k It means that node k-1 is the parent node of node k, node k-2 is the parent node of node k-1, and so on.

[0107] The current model at any node l is:

[0108]

[0109] Where ξ is the set of child nodes i including node l and the downstream branches directly connected to node l. l,i =S l S l-1 …S i-1 S i It means that the current of the load and capacitor on node i is converted into the equivalent current passing through node l, and the S matrix is ​​calculated from the admittance matrix of the branch from node l to node i. That is, it can represent the sum of the current injection of the sub-nodes of node l.

[0110] Voltage model of the downstream node j of the branch:

[0111] V j =V i -(R ij +jX ij )I j

[0112] Among them, V is the current information of each node calculated by the equivalent injection current of the end node, i is the upstream node voltage, I i is the upstream branch current.

[0113] For a branch with a transformer, the voltage model of the downstream node j is:

[0114]

[0115] in,[·] + Represents the pseudoinverse of a matrix.

[0116] The above calculation formula can replace Newton iteration, and the power flow solution in each scenario can be calculated by analytical expression to obtain node voltage and line power. This avoids the process of inverting the Jacobian matrix in the traditional Newton calculation method, reduces the calculation burden, and realizes the rapid verification of the safety of the generated scenario.

[0117] The distribution network operation safety verification index adopted in step 4 can be defined as: voltage over-limit safety verification index: voltage over-limit safety verification index: whether voltage over-limit occurs in a given distributed energy generation scenario; and thermal limit over-limit safety verification index: whether branch thermal limit over-limit occurs in a given distributed energy generation scenario.

[0118] Step 5: Optimize and calculate the maximum access power value of the unsafe scenario selected in step 4 to obtain the safety boundary of the distribution network operation.

[0119] The model for calculating the maximum access power value of the screened unsafe scenarios is:

[0120] maxλ

[0121] stf(x,μ,λ)=0

[0122]

[0123] Among them, f(x,μ,λ)=f(x,μ)+λb=0 is the parameterized power flow equation, x is the voltage amplitude and phase angle of the unbalanced node, that is, the state variable, μ is the active and reactive injection power of the generator node, that is, the control vector, and b is the change vector of the injection power, that is, the direction vector of the injection power change in the unsafe scenario. When λ=0, the parameterized power flow equation is the ground state power flow equation. The constraint vector equation f(x,μ,λ)=0 means that the solution to the problem must satisfy the parameterized power flow equation; the constraint vector equation The voltage amplitude of all nodes is required to be within a given per-unit range, typically 0.9 to 1.1 or 0.95 to 1.05 per-unit; the constraint vector equation Indicates that the current of all distribution lines and transformer branches must be less than the given current limit; constraint vector equation It is required that the reactive output of all distributed power sources must be within the given upper and lower limits.

[0124] Step 6: Combine the safety scenarios selected by the power flow calculation method without iteration in step 4 with the safety boundaries of the unsafe scenarios calculated in step 5 to characterize the safe operation domain of the distribution network to which distributed energy is connected, and on this basis, obtain the index of the flexibility of the distributed energy access capability.

[0125] This step includes the following steps:

[0126] Step 6.1, combining the safe output scenario obtained by performing safety verification on the scenario generated in step 3 according to the power flow calculation method without iteration in step 4 and calculating the safety domain boundary of the unsafe scenario using the optimization model in step 5, that is, calculating the corresponding maximum injection power point, to achieve the characterization of the safe operation domain of the distribution network to which distributed energy is connected;

[0127] Step 6.2: According to the safe operation domain of the distribution network obtained in step 6.1, select an online operating point, calculate the minimum distance between the operating point and the safety boundary, obtain the flexibility index of the operating point, and realize the flexibility evaluation of the operating point.

[0128] According to the above evaluation of the distribution network flexibility injection domain of distributed energy access, Figure 3 A test is carried out in the power distribution network shown in FIG. 1 to illustrate the effect of the present invention.

[0129] The IEEE 123-node distribution network is used as an example to test the proposed flexibility evaluation method. Its network structure is as follows: Figure 3 As shown. The network has 123 nodes and 122 branches. The initial voltage safety interval is set to [0.95, 1.05] pu, where node 1 is a balancing node. Distributed energy is connected to this network, that is, nodes 88 and 95 are connected to wind turbines, nodes 15, 25, 35, 46, 68, 80, 114 and 121 are connected to photovoltaic generators, and a constant reactive power output model is considered.

[0130] First, refer to the two wind farms provided by the National Renewable Energy Laboratory of the United States (respectively

[0131] #07806 and #07811) as historical measurement data, and the two wind farms have a strong correlation. The output scenario is modeled and sampled based on the Copula function and the LHSD method, and 1,000 scenario samples are generated. Figure 4 , Figure 5 and Figure 6 As shown in the figure, by comparing the wind turbine output scenario and the actual wind turbine output scenario generated by the Copula function and LHSD method with the LHS method without considering the output correlation, the scenario generated by the LHS method cannot capture the correlation between the actual wind turbine output scenarios, while the Copula function and LHSD method can generate scenarios showing the correlation between wind turbines through modeling and sampling considering the correlation, which can more accurately describe the uncertainty and correlation of the actual output of distributed energy.

[0132] Based on the generation of the output scenario sample set based on the Copula function and the LHSD method, the scenario can be quickly screened using the non-iterative power flow calculation method. Below, the accuracy of the power flow calculation method (dpf) without iteration in this paper and the traditional Newton method (nt) are compared. The results are as follows: Figure 7 As shown, the difference between the two methods can be as low as 10 -5. The sample set contains 1000 scenarios. When the rated power of the distributed power source connected to the network is 3MW, the Newton method screens out 184 unsafe voltage scenarios and 784 unsafe branch thermal limit scenarios. The non-iterative power flow calculation method of the present invention also screens out 184 unsafe voltage scenarios and 784 unsafe branch thermal limit scenarios. When the rated power of the distributed power source connected to the network is 2.5MW, 2MW, and 1MW, respectively, the unsafe scenarios screened out by the Newton method and the non-iterative power flow calculation method of the present invention are exactly the same, which are 450, 211, and 552 unsafe voltage scenarios, and 633, 488, and 0 unsafe branch power flow scenarios, respectively. The average calculation time of the Newton method is 5.2 seconds, and the average calculation time of the non-iterative power flow calculation method of the present invention is 3.1 seconds. Compared with the Newton method, the non-iterative power flow calculation method of this article has exactly the same results in screening unsafe scenarios, and the calculation time is less. The comparison results show that the non-iterative power flow calculation method can quickly and accurately verify the safety of the generated scenarios.

[0133] The screened safe and unsafe scenario information is integrated, and the maximum injected power optimization model is used to characterize the flexibility of the 123-node system under different operating constraints to inject into the safety domain. The safety scenario is a component of the safety domain of distributed energy injection power, and the safety boundary of the operation is obtained by calculating the accurate margin of the unsafe scenario. Figure 8 The wind power output at nodes 88 and 95 is taken as the dimension of the observed operation domain, where areas of different colors represent the flexibility injection safety domain characterized under different safety constraints.

[0134] It should be emphasized that the embodiments described in the present invention are illustrative rather than restrictive. Therefore, the present invention includes but is not limited to the embodiments described in the specific implementation manner. Any other implementation manners derived by those skilled in the art based on the technical solution of the present invention also fall within the scope of protection of the present invention.

Claims

1. A distribution network flexibility injection security domain assessment method considering distributed energy access, characterized by: The following steps are involved: Step 1: Obtain historical measurement data sets of wind turbines and photovoltaics as input data; Step 2: Use Copula function to model the distribution and correlation of historical measurement data of distributed energy output; Step 3: Based on the Copula model obtained in step 2, the Latin hypercube method LHSD considering correlation is used to generate output scenario samples of distributed energy resources; Step 4: Use a power flow calculation method without iteration to quickly analyze the operating status of the distribution system under the distributed energy output scenario obtained in step 3, and judge whether the operating status of the distribution system under different distributed energy output scenarios is safe according to the distribution network safety verification index, and select safe scenarios and unsafe scenarios according to the judgment results; Step 5: Optimize and calculate the maximum access power value of the unsafe scenario selected in step 4 to obtain the safety boundary of the distribution network operation; Step 6: Combining the safe scenarios selected by the power flow calculation method without iteration in step 4 with the safety boundaries of the unsafe scenarios calculated in step 5, the safe operation domain of the distribution network to which distributed energy is connected is characterized, and on this basis, the index of the flexibility of the distributed energy access capability is obtained; The specific implementation method of step 2 includes the following steps: Step 2.1, select the appropriate R-vine structure based on the Kendall-τ rank correlation coefficient and the maximum spanning tree MST algorithm; Step 2.2: According to the correlation of distributed energy data, the model constructed by calculating the alternative Copula function is compared with the empirical Copula function to select a suitable Copula function and estimate the corresponding model parameters; Step 2.3: According to step 2.2, build Copula layer by layer for the R-vine structure selected in step 2.1, and finally form the best correlation model; The specific implementation method of step 3 is: Step 3.1: Stratified sampling, order statistics and selection permutation; Step 3.2: For uniform random distribution U~U(0,1), d-dimensional uniform distribution connected by any copula x i Represented as n from The samples are independently drawn from , i = 1,…,n; Step 3.3: Set It represents the order statistic of the i-th sample in the j-th dimension, which is calculated as follows: for these n samples x i Sorting, expressed as x1<…<x n , then the order statistic is: Where n represents the number of samples, and I is an indicator function, that is, if the condition x in the brackets k ≤x i If true, I is 1, otherwise I is 0; Then a sample of LHSD sampling can be expressed as: Among them, W i j Represents a sample. When considering obtaining joint distribution characteristics, The value of is 0.5 or 1.

2. The method for evaluating the security domain of distribution network flexibility injection considering the access of distributed energy resources according to claim 1 is characterized in that: The power flow calculation model without iteration adopted in step 4 is: Consider the three-phase relationship between the three-phase voltage and current of the three-phase branch: Among them, I i and V i Represent the current and voltage vectors at the transmitting end, I j and V j It represents the current and voltage at the output The pressure vector, the admittance matrix is ​​composed of Y ii , Y jj , Y ji , Y ij For a line without parallel admittance to ground, Y ii =Y jj =Z -1 , Y ij =Y ji =-Z -1 , Z is the impedance matrix of the corresponding phase; For single-phase and two-phase lines, remove the elements corresponding to the non-existent phases. For lines with parallel admittance to ground, use the three-phase π model, Y ii =Y jj =Z -1 +Y / 2,Y ij =Y ji =-Z -1 ,The admittance model of the three-phase transformer adopts the traditional three-phase transformer model; Current model of the source and output ends: I i =Y ii V i +Y ij V j I j =Y ji V i +Y jj V j The output current model considering the output voltage is: in, Yes jj The pseudo-inverse of The source current model is suitable for any type of line: I i =S i I j +s i in, When there is no transformer or parallel ground admittance in the line, s i = 0, and S i is the identity matrix corresponding to the phase; otherwise, and ignore s i ; Starting from the end node k, the relationship model between the currents of each node in the entire network is: I k-1 =S k I k I k-2 =S k-1 (S k I k +I k-1 ) I k-3 =S k-2 (S k-1 S k I k +S k-1 I k-1 +I k-2 ) Among them, I k =I load,k +I cap,k , indicating that the equivalent injection current of the end node k of a line can be injected by the load current I load,k and the capacitor injection current I cap,k Indicates that node k-1 is the parent node of node k, node k-2 is the parent node of node k-1, and so on; The current model at any node l is: Where ξ is the set of child nodes i including node l and the downstream branches directly connected to node l, H l,i =S l S l-1 …S i-1 S i It means that the current of the load and capacitor on node i is converted into the equivalent current passing through node l. The S matrix is ​​calculated from the admittance matrix of the branch from node l to node i. l represents the sum of the current injections of the child nodes of node l; Voltage model of the downstream node j of the branch: V j =V i -(R ij +jX ij )I i Among them, V i is the upstream node voltage, I i is the upstream branch current; For a branch with a transformer, the voltage model of the downstream node j is: in,[·] + Represents the pseudoinverse of a matrix.

3. The method for evaluating the security domain of distribution network flexibility injection considering the access of distributed energy resources according to claim 1 is characterized in that: The distribution network operation safety check index in step 4 is defined as: voltage over-limit safety check index: whether voltage over-limit occurs in a given distributed energy generation scenario; And the thermal limit exceeding safety verification index: whether the branch thermal limit exceeding occurs in a given distributed energy generation scenario.

4. The method for evaluating the security domain of distribution network flexibility injection considering distributed energy access according to claim 1 is characterized in that: The model for calculating the maximum access power value of the unsafe scenario screened in step 4 in step 5 is: maxλ stf(x,μ,λ)=0 Among them, f(x,μ,λ)=f(x,μ)+λb=0 is the parameterized power flow equation, x is the voltage amplitude and phase angle of the unbalanced node, that is, the state variable, μ is the active and reactive injection power of the generator node, that is, the control vector, and b is the change vector of the injection power, that is, the direction vector of the injection power change in the unsafe scenario. When λ=0, the parameterized power flow equation is the ground state power flow equation, and the constraint vector equation f(x,μ,λ)=0 means that the solution to the problem must satisfy the parameterized power flow equation; the constraint vector equation The voltage amplitude of all nodes is required to be within a given per-unit range, typically 0.9 to 1.1 or 0.95 to 1.05 per-unit; the constraint vector equation Indicates that the current of all distribution lines and transformer branches must be less than the given current limit; constraint vector equation It is required that the reactive output of all distributed power sources must be within the given upper and lower limits.

5. The method for evaluating the security domain of distribution network flexibility injection considering distributed energy access according to claim 1 is characterized in that: The specific implementation method of step 6 comprises the following steps: Step 6.1, combining the safe output scenario obtained by performing safety verification on the scenario generated in step 3 according to the power flow calculation method without iteration in step 4 and the safety domain boundary of the unsafe scenario calculated by the optimization model in step 5, that is, the calculated corresponding maximum injection power point, to achieve the characterization of the safe operation domain of the distribution network to which distributed energy is connected; Step 6.2: According to the safe operation domain of the distribution network obtained in step 6.1, select an online operating point, calculate the minimum distance between the operating point and the safety boundary, obtain the flexibility index of the operating point, and realize the flexibility evaluation of the operating point.

Images