Method and device for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network

Abstract

The present invention discloses a method and device for generating a probabilistic optimization configuration scheme for measurement points in a direct current (DC) distribution network. The method comprises: respectively obtaining measurement data on lines and nodes of the DC distribution network, grid parameters of the DC distribution network grid, and a node pseudo-measurement vector group related to the load and photovoltaic injection power of the DC distribution network; constructing a probabilistic optimization configuration model for the measurement points of the DC distribution network using the measurement data, grid parameters, and the node pseudo-measurement vector group; and solving the probabilistic optimization configuration model for the measurement points according to a genetic algorithm to obtain an optimal configuration scheme for the measurement points of the DC distribution network. The present invention fully considers the influence of the randomness and correlation of the DC distribution network load and distributed photovoltaic injection power on system state estimation, and effectively reduces the configuration cost of the measurement points of the DC distribution network while ensuring the accuracy of the DC distribution network state estimation.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid configuration schemes, and in particular to a method and device for generating a probabilistic optimization configuration scheme for measurement points in a direct current distribution network. Background Art

[0002] With the development of science and technology, the traditional power system is transforming into a new power system. A large number of distributed new energy sources are connected to the distribution network, which has had a huge impact on the safety, stability and reliability of the distribution network operation. It is urgently needed to carry out accurate perception and optimization control of the distribution network operation status to ensure that the distribution network is in a stable working state.

[0003] Due to the lack of real-time measurement points and insufficient redundancy in traditional distribution networks, it is difficult to support distribution network state estimation. To better estimate the state of the distribution network, one common method is to use pseudo-measurement data to increase the redundancy of distribution network measurement data and improve the accuracy of distribution network state estimation. However, the distributed power sources and loads in the distribution network fluctuate greatly and are subject to significant randomness, which affects the accuracy of pseudo-measurement modeling. In addition, the geographical proximity of the nodes in the distribution network and similar meteorological conditions lead to correlations in the power fluctuations of the distributed power sources and loads, further affecting the accuracy of pseudo-measurement modeling and, in turn, the accuracy of distribution network state estimation. It is urgent to consider the impact of pseudo-measurements, rationally configure real-time measurement points, and ensure the accuracy of distribution network state estimation.

[0004] Existing commonly used methods have the following technical issues: When conducting research on DC distribution network measurement point configuration, they fail to consider the randomness and correlation of pseudo-measurement data from distributed power sources and loads. This can easily lead to optimistic configuration solutions and make it difficult to ensure the accuracy of distribution network state estimation. Therefore, there is an urgent need to conduct research on fast distribution network state estimation methods that consider the randomness and correlation of distributed power sources and loads, and to guide the optimal configuration of distribution network measurement points. Summary of the Invention

[0005] The present invention proposes a method and device for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network. After obtaining measurement data of DC distribution network lines and nodes, grid parameters, and photovoltaic injection power parameters, the method uses these parameters to construct a configuration model for the measurement points and solves the configuration model to obtain the optimal configuration scheme for the measurement points. This method effectively reduces the configuration cost of the measurement points in the DC distribution network while ensuring the accuracy of DC distribution network state estimation.

[0006] A first aspect of an embodiment of the present invention provides a method for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network, the method comprising:

[0007] respectively obtaining measurement data on lines and nodes of the DC distribution network, grid parameters of the DC distribution network, and a node pseudo-measurement vector group of loads and photovoltaic injection power of the DC distribution network;

[0008] Constructing a measurement point probabilistic optimization configuration model of a DC distribution network using the measurement data, the grid parameters and the node pseudo-measurement vector group;

[0009] The measurement point probability optimization configuration model is solved by genetic algorithm to obtain the measurement point probability optimization configuration scheme of the DC distribution network.

[0010] In a possible implementation manner of the first aspect, the operation of obtaining the node pseudo measurement vector group includes:

[0011] Calculating a weight vector corresponding to a first pseudo measurement vector group of the load of the DC distribution network and the distributed photovoltaic injection power according to a three-point estimation algorithm, and constructing a vector cumulative distribution function according to the weight vector;

[0012] After obtaining the second pseudo measurement vector group, a node pseudo measurement vector group of the load and photovoltaic injection power of the DC distribution network is calculated according to the vector cumulative distribution function and the second pseudo measurement vector group using a Nataf transform algorithm.

[0013] In a possible implementation manner of the first aspect, the operation of obtaining the second pseudo measurement vector group includes:

[0014] Construct a correlation coefficient matrix about the correlation between the load of each node in the DC distribution network and the distributed photovoltaic injection power;

[0015] Decomposing the correlation coefficient matrix according to the Cholesky decomposition algorithm to obtain an upper triangular matrix and a lower triangular matrix;

[0016] A second pseudo measurement vector group regarding the DC distribution network load and the distributed photovoltaic injection power is calculated using the lower triangular matrix, wherein the second pseudo measurement vector group satisfies the product sum of the lower triangular matrix and the first pseudo measurement vector group.

[0017] In a possible implementation of the first aspect, the vector cumulative distribution function includes: a load cumulative distribution function of a random variable of load injection power of the DC distribution network, and a photovoltaic cumulative distribution function of a random variable of distributed photovoltaic injection power of the DC distribution network;

[0018] The load cumulative distribution function is shown in the following formula:

[0019]

[0020] In the above formula, is the load cumulative distribution function, is the probability density function;

[0021] The probability density function is shown in the following formula:

[0022]

[0023] In the above formula, μ L and σ L are the expected value and standard deviation of the load injection power, respectively;

[0024] The photovoltaic cumulative distribution function is shown in the following formula:

[0025]

[0026] In the above formula, is the photovoltaic cumulative distribution function, is the probability density function;

[0027] The probability density function is shown in the following formula:

[0028]

[0029] In the above formula, P pv.max is the rated capacity of distributed photovoltaics; Γ(·) is the Gamma function; α and β are the first and second shape parameters of the Beta distribution, respectively;

[0030] The first shape parameter and the second shape parameter are expressed as follows:

[0031]

[0032] In the above formula, P pv Inject actual power value into distributed photovoltaics; and are the expected value and standard deviation of the ratio of the actual value of distributed photovoltaic injection power to the rated capacity, respectively.

[0033] In a possible implementation of the first aspect, constructing a measurement point probabilistic optimization configuration model for a DC distribution network using the measurement data, the grid parameters, and the node pseudo-measurement vector group includes:

[0034] Constructing a probabilistic state estimation model of a DC distribution network according to the measurement data, the grid parameters and the node pseudo-measurement vector group;

[0035] The probabilistic state estimation model is calculated using a quadratic constrained programming algorithm and a Gram-Charlie series expansion to obtain a cumulative distribution function of the node voltage estimation deviation;

[0036] After determining the measurement point objective function of the DC distribution network, the deviation cumulative distribution function and the preset line measurement point configuration decision are used to constrain the objective function to construct a measurement point probability optimization configuration model, wherein the measurement point objective function is a function that corresponds to the minimum number of measurement point configurations in the DC distribution network.

[0037] In a possible implementation of the first aspect, the construction of the probabilistic state estimation model includes:

[0038] Calculating a measurement value using the measurement data and the node pseudo-measurement vector group, and constructing an objective function of an estimation model by minimizing the absolute value sum of the difference between the measurement value and the measurement equation value of the node voltage state estimation quantity corresponding to the measurement value and a preset first equivalent variable to obtain an estimated objective function;

[0039] A probabilistic state estimation model is constructed based on the estimation objective function and preset estimation constraints, wherein the estimation constraints include measurement equation constraints of line power with respect to node voltage state estimation, measurement equation constraints of line current with respect to node voltage state estimation, measurement equation constraints of load and distributed photovoltaic injection power with respect to node voltage state estimation, and preset upper and lower limit constraints of the first equivalent variable.

[0040] In a possible implementation of the first aspect, the probabilistic state estimation model is shown in the following formula:

[0041]

[0042]

[0043] In the above formula, K is the total number of dimensions of the real-time measurement data of the line and the node pseudo-measurement vector of the load and distributed photovoltaic injection power; are the kth measurement value and state estimation value of the jth node pseudo-measurement vector respectively; is the measurement equation value containing the node voltage state estimation quantity involved in the pseudo measurement vector of the jth node; is the kth first equivalent variable involved in the pseudo-measurement vector of the jth node; the constraints c1, c2, c3, and c4 respectively represent the measurement equation constraints of the line power on the node voltage state estimate, the measurement equation constraints of the line current on the node voltage state estimate, the measurement equation constraints of the load and distributed photovoltaic injection power on the node voltage state estimate, and the upper and lower limit constraints of the preset first equivalent variable; are the line pq power measurement equation value, line pq current measurement equation value, and node p injected power measurement equation value of the j-th node pseudo measurement vector respectively; are the voltage state estimates of node p and node q involved in the jth node pseudo-measurement vector; b k is the kth binary configuration variable; M is a preset positive number; g pq is the conductance parameter of line pq; Ω is the node set.

[0044] In a possible implementation of the first aspect, the operation of calculating the deviation cumulative distribution function includes:

[0045] Solving the probabilistic state estimation model according to the quadratic constrained programming algorithm to obtain a state estimation vector group about the voltage state of the DC distribution network node, and using the state estimation vector group to construct a voltage estimation deviation vector group;

[0046] Calculating deviation origin moment information corresponding to the voltage estimation deviation vector group according to the weight vector;

[0047] The deviation origin moment information is calculated using a Gram-Charlie series expansion to obtain a deviation cumulative distribution function of the node voltage estimation deviation.

[0048] In a possible implementation of the first aspect, the grid parameters include topological connection relationships of node lines, line conductance parameters, and distributed photovoltaic rated capacity;

[0049] The measurement data includes line real-time measurement data and node non-real-time data;

[0050] The line real-time measurement data refers to line power and line current data that can be uploaded in real time; the node non-real-time data refers to node injection power data that cannot be uploaded in real time.

[0051] A second aspect of an embodiment of the present invention provides a device for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network, the device comprising:

[0052] an acquisition module, configured to respectively acquire measurement data on lines and nodes of a DC distribution network, grid parameters of a DC distribution network grid, and a node pseudo-measurement vector group of loads and photovoltaic injection power of the DC distribution network;

[0053] A construction module, configured to construct a probability optimization configuration model of measurement points of a DC distribution network using the measurement data, the grid parameters and the node pseudo-measurement vector group;

[0054] The generating module is used to solve the measurement point probability optimization configuration model according to the genetic algorithm to obtain the measurement point probability optimization configuration scheme of the DC distribution network.

[0055] Compared to the prior art, the method and apparatus for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network provided by an embodiment of the present invention have the following beneficial effects: the present invention can first obtain measurement data of DC distribution network lines and nodes, grid parameters, and photovoltaic injection power parameters, use the measurement data, grid parameters, and photovoltaic injection power parameters to construct a configuration model for the measurement points, and then use an algorithm to solve the configuration model to obtain the optimal probabilistic configuration scheme for the measurement points. By combining parameters such as load and photovoltaic injection power, the estimation deviation can be reduced to improve the accuracy of the estimation analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a flow chart of a method for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network provided by one embodiment of the present invention;

[0057] Figure 2 The present invention is a schematic structural diagram of a device for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network according to an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0059] In order to solve the above problems, a method for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network provided by an embodiment of the present application will be introduced and explained in detail through the following specific embodiments.

[0060] Reference Figure 1 , which shows a flow chart of a method for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network provided by an embodiment of the present invention.

[0061] As an example, the method for generating a probabilistic optimization configuration scheme for DC distribution network measurement points may include:

[0062] S11. Obtain measurement data on lines and nodes of the DC distribution network, grid parameters of the DC distribution network, and a node pseudo-measurement vector group of loads and photovoltaic injection power of the DC distribution network.

[0063] In one embodiment, the grid parameters include topological connection relationships of node lines, line conductance parameters, and distributed photovoltaic rated capacity;

[0064] The measurement data includes real-time measurement data of the lines of the distribution network and non-real-time data of the nodes;

[0065] The line real-time measurement data refers to line power and line current data that can be uploaded in real time; the node non-real-time data refers to node injection power data that cannot be uploaded in real time.

[0066] The node pseudo-measurement vector group may be a vector group of DC distribution network load and distributed photovoltaic injection power that takes randomness and correlation into account, calculated based on a three-point estimation algorithm and a Nataf transformation algorithm.

[0067] As an example, the operation of obtaining the node pseudo measurement vector group may include the following sub-steps:

[0068] S111 . Calculate a weight vector corresponding to a first pseudo measurement vector group of the load of the DC distribution network and the distributed photovoltaic injection power according to a three-point estimation algorithm, and construct a vector cumulative distribution function according to the weight vector.

[0069] A first pseudo measurement vector group Z of the DC distribution network load and the distributed photovoltaic injection power, and a weight vector W corresponding to the first pseudo measurement vector group are calculated according to the three-point estimation algorithm.

[0070] The first pseudo measurement vector group is shown in the following formula:

[0071]

[0072] The weight vector is shown below:

[0073]

[0074] In the above formula, They represent the first estimation point vector and the second estimation point vector of the i-th load or distributed photovoltaic injection power respectively.

[0075] And the first estimated point vector and the second estimated point vector need to satisfy And there is The values ​​of other elements are all 0; N is the total number of DC distribution network loads and distributed photovoltaics; Z 2N+1 is an N-dimensional zero vector.

[0076] Then, a vector cumulative distribution function may be constructed according to the weight vector. The vector cumulative distribution function is a cumulative distribution function of random variables of the DC distribution network load and the distributed photovoltaic injection power that takes randomness into account.

[0077] In one embodiment, the vector cumulative distribution function includes: a load cumulative distribution function of a random variable of load injection power of the DC distribution network, and a photovoltaic cumulative distribution function of a random variable of distributed photovoltaic injection power of the DC distribution network;

[0078] The load cumulative distribution function is shown in the following formula:

[0079]

[0080] In the above formula, is the load cumulative distribution function, is the probability density function;

[0081] The probability density function is shown in the following formula:

[0082]

[0083] In the above formula, μ L and σ L are the expected value and standard deviation of the load injection power, respectively;

[0084] The photovoltaic cumulative distribution function is shown in the following formula:

[0085]

[0086] In the above formula, is the photovoltaic cumulative distribution function, is the probability density function;

[0087] The probability density function is shown in the following formula:

[0088]

[0089] In the above formula, P pv.max is the rated capacity of distributed photovoltaics; Γ(·) is the Gamma function; α and β are the first and second shape parameters of the Beta distribution, respectively;

[0090] The first shape parameter and the second shape parameter are expressed as follows:

[0091]

[0092] In the above formula, P pv Inject actual power value into distributed photovoltaics; and are the expected value and standard deviation of the ratio of the actual value of distributed photovoltaic injection power to the rated capacity, respectively.

[0093] S112. After obtaining the second pseudo measurement vector group, use a Nataf transform algorithm to calculate a node pseudo measurement vector group regarding the load of the DC distribution network and the photovoltaic injection power according to the vector cumulative distribution function and the second pseudo measurement vector group.

[0094] In one embodiment, the second pseudo measurement vector group is a matrix vector of DC distribution network load and distributed photovoltaic injection power constructed according to the correlation between the load of each node in the DC distribution network and the distributed photovoltaic injection power.

[0095] After obtaining the second pseudo measurement vector group, the second pseudo measurement vector group can be calculated according to the vector cumulative distribution function using the Nataf transform algorithm, thereby obtaining the node pseudo measurement vector group about the load of the DC distribution network and the photovoltaic injection power.

[0096] Specifically, the second pseudo measurement vector group is Y. Based on the second pseudo measurement vector group and the vector cumulative distribution function of the random variables of the DC distribution network load and the distributed photovoltaic injection power considering randomness, the Nataf transformation algorithm is used to calculate the node pseudo measurement vector group X of the DC distribution network load and the distributed photovoltaic injection power considering randomness and correlation. The node pseudo measurement vector group X is shown in the following formula:

[0097] X=[X1,X2,…,X j ,…,X 2N+1 ]

[0098]

[0099] Where, X j is the pseudo measurement vector of the jth node of DC distribution network load and distributed photovoltaic injection power considering randomness and correlation; ij The element in the i-th row and j-th column of the node pseudo measurement vector group X of the DC distribution network load and distributed photovoltaic injection power; is the vector cumulative distribution function of the random variable of the i-th load or distributed photovoltaic injection power considering randomness, Its corresponding inverse function; Φ(y ij ) is the element y in the i-th row and j-th column of the second pseudo measurement vector group Y ij The corresponding cumulative probability value under the standard normal distribution.

[0100] As an example, the operation of obtaining the second pseudo measurement vector group may include the following sub-steps:

[0101] S1121. Construct a correlation coefficient matrix regarding the correlation between the loads at each node in the DC distribution network and the distributed photovoltaic injection power.

[0102] S1122. Decompose the correlation coefficient matrix according to a Cholesky decomposition algorithm to obtain an upper triangular matrix and a lower triangular matrix.

[0103] S1123. Calculate a second pseudo measurement vector group related to the DC distribution network load and the distributed photovoltaic injection power using the lower triangular matrix, wherein the second pseudo measurement vector group satisfies the product sum of the lower triangular matrix and the first pseudo measurement vector group.

[0104] Specifically, the correlation coefficient matrix ρ can be predefined X , the correlation coefficient matrix can express the correlation between the load of each node and the distributed photovoltaic injection power in the DC distribution network. The correlation coefficient matrix is ​​shown as follows:

[0105]

[0106] In the above formula, The Pearson correlation coefficient between the injected power of the first load or distributed photovoltaic and the second load or distributed photovoltaic;

[0107] Then, the correlation coefficient matrix ρ can be decomposed into X Decompose into an upper triangular matrix U and a lower triangular matrix D, where the upper triangular matrix U and the lower triangular matrix D need to satisfy the following conditions:

[0108] ρ X =DU=DD T ;

[0109] Finally, the second pseudo measurement vector group Y of DC distribution network load and distributed photovoltaic injection power can be calculated based on the lower triangular matrix D obtained by decomposing the correlation coefficient matrix.

[0110] It should be noted that the second pseudo measurement vector group meets the following conditions:

[0111] Y=DZ.

[0112] S12: Constructing a probabilistic optimization configuration model for measurement points of a DC distribution network using the measurement data, the grid parameters, and the node pseudo-measurement vector group.

[0113] In an optional embodiment, the measurement data, grid parameters and node pseudo-measurement vector groups may be used as constraints to constrain the preset objective function, thereby constructing a measurement point probability optimization configuration model.

[0114] As an example, step S12 may include the following sub-steps:

[0115] S121: Construct a probabilistic state estimation model of a DC distribution network according to the measurement data, the grid parameters, and the node pseudo-measurement vector group.

[0116] In one embodiment, a corresponding model objective function may be constructed using the numerical values ​​of the measurement data, grid parameters, and node pseudo-measurement vector groups, and a probabilistic state estimation model may be constructed after setting the constraints of the model objective function.

[0117] As an example, the operation of constructing the probabilistic state estimation model may include the following sub-steps:

[0118] S1211. Calculate the measurement value using the measurement data and the node pseudo-measurement vector group, and construct the objective function of the estimation model by minimizing the absolute value sum of the difference between the measurement value and the measurement equation value of the node voltage state estimation quantity corresponding to the measurement value and the preset first equivalent variable to obtain the estimated objective function.

[0119] In one embodiment, measurement values ​​are calculated using measurement data and a node pseudo-measurement vector group. The measurement values ​​may include real-time measurement data of distribution network lines and node pseudo-measurement vectors of DC distribution network load and distributed photovoltaic injection power that take into account randomness and correlation.

[0120] Specifically, a minimized value of the measurement value can be calculated. The sum of the absolute values ​​of the differences between the minimized value and the corresponding value of the measurement equation containing the node voltage state estimate is used as the objective function of the probabilistic state estimation model. Simultaneously, a preset first equivalent variable can be introduced into the objective function. The objective function of the probabilistic state estimation model can be equal to the absolute value of the difference between the measurement value and the corresponding value of the measurement equation containing the node voltage state estimate, thereby obtaining an estimated objective function.

[0121] S1212. Construct a probabilistic state estimation model based on the estimation objective function and preset estimation constraints, wherein the estimation constraints include measurement equation constraints of line power with respect to node voltage state estimation, measurement equation constraints of line current with respect to node voltage state estimation, measurement equation constraints of load and distributed photovoltaic injection power with respect to node voltage state estimation, and preset upper and lower limit constraints of the first equivalent variable.

[0122] In one embodiment, the probabilistic state estimation model is shown as follows:

[0123]

[0124]

[0125] In the above formula, K is the total number of dimensions of the real-time measurement data of the line and the node pseudo-measurement vector of the load and distributed photovoltaic injection power; are the kth measurement value and state estimation value of the jth node pseudo-measurement vector respectively; is the measurement equation value containing the node voltage state estimation quantity involved in the pseudo measurement vector of the jth node; is the kth first equivalent variable involved in the pseudo-measurement vector of the jth node; the constraints c1, c2, c3, and c4 respectively represent the measurement equation constraints of the line power on the node voltage state estimate, the measurement equation constraints of the line current on the node voltage state estimate, the measurement equation constraints of the load and distributed photovoltaic injection power on the node voltage state estimate, and the upper and lower limit constraints of the preset first equivalent variable; are the line pq power measurement equation value, line pq current measurement equation value, and node p injected power measurement equation value of the j-th node pseudo measurement vector respectively; are the voltage state estimates of node p and node q involved in the jth node pseudo-measurement vector; b k is the kth binary configuration variable; M is a preset positive number; g pq is the conductance parameter of line pq; Ω is the node set.

[0126] S122 , using a quadratic constrained programming algorithm and a Gram-Charlie series expansion to calculate the probabilistic state estimation model, and obtaining a deviation cumulative distribution function of the node voltage estimation deviation.

[0127] In one embodiment, the probabilistic state estimation model can be solved using a quadratic constrained programming algorithm, and then the solved formula can be used to calculate the probabilistic state estimation model using a Gram-Charlie series expansion. Finally, the deviation cumulative distribution function of the node voltage estimation deviation can be obtained.

[0128] As an example, step S122 may include the following sub-steps:

[0129] S1221. Solve the probabilistic state estimation model according to the quadratic constrained programming algorithm to obtain a state estimation vector group about the voltage state of the DC distribution network node, and use the state estimation vector group to construct a voltage estimation deviation vector group.

[0130] S1222. Calculate, based on the weight vector, deviation origin moment information corresponding to the voltage estimation deviation vector group.

[0131] S1223 , using a Gram-Charlie series expansion to calculate the deviation origin moment information, and obtain a deviation cumulative distribution function of the node voltage estimation deviation.

[0132] Specifically, according to the quadratic constrained programming algorithm, the probabilistic state estimation model is solved to obtain a state estimation vector group about the voltage state of the DC distribution network node. The state estimation vector group can be expressed as follows:

[0133]

[0134] In the above formula, T is the total number of nodes in the DC distribution network;

[0135] The state estimation vector group is used to construct a voltage estimation deviation vector group. In one embodiment, the voltage estimation deviation vector group of the DC distribution network node is defined as follows:

[0136]

[0137] In the above formula, The voltage estimation deviation of node p involved in the j-th node pseudo-measurement vector needs to meet the following conditions:

[0138]

[0139] In the above formula, is the true value of the voltage flow at the node p in which the j-th node pseudo-measurement vector participates.

[0140] Next, the deviation origin moment information corresponding to the voltage estimation deviation vector group may be calculated based on the weight vector.

[0141] The specific calculation can be shown as follows:

[0142]

[0143] Where, is the l-order origin moment of the voltage estimation deviation at node p; w j is the jth weight coefficient, is the lth power of the voltage estimation deviation of the node p involved in the jth node pseudo measurement vector;

[0144] Finally, after determining the deviation origin moment information of each node voltage estimation deviation, the deviation origin moment information can be substituted into the Gram-Charlie series expansion, and the cumulative distribution function of the node voltage estimation deviation is calculated according to the Gram-Charlie series expansion to obtain the deviation cumulative distribution function.

[0145] S123. After determining the measurement point objective function of the DC distribution network, constrain the objective function using the cumulative deviation distribution function and a preset line measurement point configuration decision to construct a measurement point probabilistic optimization configuration model, wherein the measurement point objective function is a function that minimizes the number of measurement point configurations in the DC distribution network.

[0146] In one embodiment, the measurement point objective function may be to minimize the number of measurement points configured in the DC distribution network. Specifically, the objective function of the DC distribution network measurement point probability optimization configuration model may be defined as minimizing the number of measurement points configured in the DC distribution network, as shown in the following formula:

[0147] minf=minN m ;

[0148] The objective function is constrained by using the cumulative distribution function of deviation and the preset line measurement point configuration decision, so that a probability optimization configuration model of measurement points can be constructed.

[0149] Specifically, the constraints of the DC distribution network measurement point probabilistic optimization configuration model can be defined as line measurement point configuration decision constraints and node voltage estimation deviation opportunity constraints, wherein the node voltage estimation deviation opportunity constraint can be the above-mentioned deviation cumulative distribution function.

[0150] Specifically, the above two constraints can be expressed as:

[0151] in,

[0152] In the above formula, d pq The decision variable is configured for the line pq measurement point. If it is 1, it means that the measurement point is configured, otherwise it is not configured. Pr{·} is the probability of the event being established; ε0 is the node voltage deviation estimation accuracy threshold; ψ0 is the chance constraint confidence threshold.

[0153] S13. Solve the measurement point probability optimization configuration model according to the genetic algorithm to obtain a measurement point probability configuration scheme for the DC distribution network.

[0154] Finally, the probability optimization configuration model of measurement points can be solved according to the genetic algorithm, so as to obtain the optimal probability configuration scheme of measurement points in the DC distribution network.

[0155] The above operation fully considers the impact of the randomness and correlation of the DC distribution network load and distributed photovoltaic injection power on the system state estimation. While ensuring the accuracy of the DC distribution network state estimation, it effectively reduces the configuration cost of the DC distribution network measurement points.

[0156] In this embodiment, an embodiment of the present invention provides a method for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network. The beneficial effect of the method is that: the method can first obtain measurement data of DC distribution network lines and nodes, grid parameters, and photovoltaic injection power parameters, use the measurement data, grid parameters, and photovoltaic injection power parameters to construct a configuration model of the measurement points, and then use an algorithm to solve the configuration model to obtain the optimal probabilistic configuration scheme for the measurement points. By combining parameters such as load and photovoltaic injection power, the estimation deviation can be reduced to improve the accuracy of the estimation analysis. In addition, the method also fully considers the impact of the randomness and correlation of the DC distribution network load and distributed photovoltaic injection power on the system state estimation. Under the premise of ensuring the accuracy of the DC distribution network state estimation, the configuration cost of the DC distribution network measurement points is effectively reduced.

[0157] The embodiment of the present invention also provides a device for generating a probability optimization configuration scheme for measurement points in a DC distribution network, see Figure 2 , shows a structural diagram of a device for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network provided by an embodiment of the present invention.

[0158] As an example, the device for generating a probabilistic optimization configuration scheme for DC distribution network measurement points may include:

[0159] An acquisition module 201 is configured to respectively acquire measurement data on lines and nodes of a DC distribution network, grid parameters of a DC distribution network grid, and a node pseudo-measurement vector group of loads and photovoltaic injection power of the DC distribution network;

[0160] A construction module 202 is configured to construct a probabilistic optimization configuration model of measurement points of a DC distribution network using the measurement data, the grid parameters, and the node pseudo-measurement vector group;

[0161] The generating module 203 is configured to solve the measurement point probability optimization configuration model according to a genetic algorithm to obtain a measurement point probability optimization configuration scheme for the DC distribution network.

[0162] Optionally, the operation of obtaining the node pseudo measurement vector group includes:

[0163] Calculating a weight vector corresponding to a first pseudo measurement vector group of the load of the DC distribution network and the distributed photovoltaic injection power according to a three-point estimation algorithm, and constructing a vector cumulative distribution function according to the weight vector;

[0164] After obtaining the second pseudo measurement vector group, a node pseudo measurement vector group of the load and photovoltaic injection power of the DC distribution network is calculated according to the vector cumulative distribution function and the second pseudo measurement vector group using a Nataf transform algorithm.

[0165] Optionally, the operation of obtaining the second pseudo measurement vector group includes:

[0166] Construct a correlation coefficient matrix about the correlation between the load of each node in the DC distribution network and the distributed photovoltaic injection power;

[0167] Decomposing the correlation coefficient matrix according to the Cholesky decomposition algorithm to obtain an upper triangular matrix and a lower triangular matrix;

[0168] A second pseudo measurement vector group regarding the DC distribution network load and the distributed photovoltaic injection power is calculated using the lower triangular matrix, wherein the second pseudo measurement vector group satisfies the product sum of the lower triangular matrix and the first pseudo measurement vector group.

[0169] Optionally, the vector cumulative distribution function includes: a load cumulative distribution function of a random variable of load injection power of the DC distribution network, and a photovoltaic cumulative distribution function of a random variable of distributed photovoltaic injection power of the DC distribution network;

[0170] The load cumulative distribution function is shown in the following formula:

[0171]

[0172] In the above formula, is the load cumulative distribution function, is the probability density function;

[0173] The probability density function is shown in the following formula:

[0174]

[0175] In the above formula, μ L and σ L are the expected value and standard deviation of the load injection power, respectively;

[0176] The photovoltaic cumulative distribution function is shown in the following formula:

[0177]

[0178] In the above formula, is the photovoltaic cumulative distribution function, is the probability density function;

[0179] The probability density function is shown in the following formula:

[0180]

[0181] In the above formula, P pv.maxis the rated capacity of distributed photovoltaics; Γ(·) is the Gamma function; α and β are the first and second shape parameters of the Beta distribution, respectively;

[0182] The first shape parameter and the second shape parameter are expressed as follows:

[0183]

[0184] In the above formula, P pv Inject actual power value into distributed photovoltaics; and are the expected value and standard deviation of the ratio of the actual value of distributed photovoltaic injection power to the rated capacity, respectively.

[0185] Optionally, the building block is further configured to:

[0186] Constructing a probabilistic state estimation model of a DC distribution network according to the measurement data, the grid parameters and the node pseudo-measurement vector group;

[0187] The probabilistic state estimation model is calculated using a quadratic constrained programming algorithm and a Gram-Charlie series expansion to obtain a cumulative distribution function of the node voltage estimation deviation;

[0188] After determining the measurement point objective function of the DC distribution network, the deviation cumulative distribution function and the preset line measurement point configuration decision are used to constrain the objective function to construct a measurement point probability optimization configuration model, wherein the measurement point objective function is a function that corresponds to the minimum number of measurement point configurations in the DC distribution network.

[0189] Optionally, the operation of constructing the probabilistic state estimation model includes:

[0190] Calculating a measurement value using the measurement data and the node pseudo-measurement vector group, and constructing an objective function of an estimation model by minimizing the absolute value sum of the difference between the measurement value and the measurement equation value of the node voltage state estimation quantity corresponding to the measurement value and a preset first equivalent variable to obtain an estimated objective function;

[0191] A probabilistic state estimation model is constructed based on the estimation objective function and preset estimation constraints, wherein the estimation constraints include measurement equation constraints of line power with respect to node voltage state estimation, measurement equation constraints of line current with respect to node voltage state estimation, measurement equation constraints of load and distributed photovoltaic injection power with respect to node voltage state estimation, and preset upper and lower limit constraints of the first equivalent variable.

[0192] Optionally, the probabilistic state estimation model is shown as follows:

[0193]

[0194] In the above formula, K is the total number of dimensions of the real-time measurement data of the line and the node pseudo-measurement vector of the load and distributed photovoltaic injection power; are the kth measurement value and state estimation value of the jth node pseudo-measurement vector respectively; is the measurement equation value containing the node voltage state estimation quantity involved in the pseudo measurement vector of the jth node; is the kth first equivalent variable involved in the pseudo-measurement vector of the jth node; the constraints c1, c2, c3, and c4 respectively represent the measurement equation constraints of the line power on the node voltage state estimate, the measurement equation constraints of the line current on the node voltage state estimate, the measurement equation constraints of the load and distributed photovoltaic injection power on the node voltage state estimate, and the upper and lower limit constraints of the preset first equivalent variable; are the line pq power measurement equation value, line pq current measurement equation value, and node p injected power measurement equation value of the j-th node pseudo measurement vector respectively; are the voltage state estimates of node p and node q involved in the jth node pseudo-measurement vector; b k is the kth binary configuration variable; M is a preset positive number; g pq is the conductance parameter of line pq; Ω is the node set.

[0195] Optionally, the operation of calculating the deviation cumulative distribution function includes:

[0196] Solving the probabilistic state estimation model according to the quadratic constrained programming algorithm to obtain a state estimation vector group about the voltage state of the DC distribution network node, and using the state estimation vector group to construct a voltage estimation deviation vector group;

[0197] Calculating deviation origin moment information corresponding to the voltage estimation deviation vector group according to the weight vector;

[0198] The deviation origin moment information is calculated using a Gram-Charlie series expansion to obtain a deviation cumulative distribution function of the node voltage estimation deviation.

[0199] Optionally, the grid parameters include topological connection relationships of node lines, line conductance parameters, and distributed photovoltaic rated capacity;

[0200] The measurement data includes line real-time measurement data and node non-real-time data;

[0201] The line real-time measurement data refers to line power and line current data that can be uploaded in real time; the node non-real-time data refers to node injection power data that cannot be uploaded in real time.

[0202] Those skilled in the art can clearly understand that, for the sake of convenience and brevity, the specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0203] Furthermore, an embodiment of the present application also provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network as described in the above embodiment is implemented.

[0204] Furthermore, an embodiment of the present application also provides a computer-readable storage medium, which stores a computer-executable program, and the computer-executable program is used to enable a computer to execute the method for generating a probabilistic optimization configuration scheme for DC distribution network measurement points as described in the above embodiment.

[0205] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network, characterized in that: The method comprises: respectively obtaining measurement data on lines and nodes of the DC distribution network, grid parameters of the DC distribution network, and a node pseudo-measurement vector group of loads and photovoltaic injection power of the DC distribution network; Constructing a measurement point probability optimization configuration model of a DC distribution network using the measurement data, the grid parameters and the node pseudo-measurement vector group; Solving the measurement point probability optimization configuration model according to the genetic algorithm to obtain the measurement point probability optimization configuration scheme of the DC distribution network; The operation of obtaining the node pseudo measurement vector group includes: Calculating a weight vector corresponding to a first pseudo measurement vector group of the load of the DC distribution network and the distributed photovoltaic injection power according to a three-point estimation algorithm, and constructing a vector cumulative distribution function according to the weight vector; After obtaining the second pseudo measurement vector group, a node pseudo measurement vector group of the load and photovoltaic injection power of the DC distribution network is calculated using the Nataf transform algorithm according to the vector cumulative distribution function and the second pseudo measurement vector group; The operation of obtaining the second pseudo measurement vector group includes: Construct a correlation coefficient matrix about the correlation between the load of each node in the DC distribution network and the distributed photovoltaic injection power; Decomposing the correlation coefficient matrix according to the Cholesky decomposition algorithm to obtain an upper triangular matrix and a lower triangular matrix; A second pseudo measurement vector group regarding the DC distribution network load and the distributed photovoltaic injection power is calculated using the lower triangular matrix, wherein the second pseudo measurement vector group satisfies the product sum of the lower triangular matrix and the first pseudo measurement vector group.

2. The method for generating a probabilistic optimization configuration scheme for DC distribution network measurement points according to claim 1, characterized in that: The vector cumulative distribution function includes: a load cumulative distribution function of a random variable of load injection power of a DC distribution network, and a photovoltaic cumulative distribution function of a random variable of distributed photovoltaic injection power of a DC distribution network; The load cumulative distribution function is shown in the following formula: In the above formula, is the load cumulative distribution function, is the probability density function; The probability density function is shown in the following formula: In the above formula, μ L and σ L are the expected value and standard deviation of the load injection power, respectively; The photovoltaic cumulative distribution function is shown in the following formula: In the above formula, is the photovoltaic cumulative distribution function, is the probability density function; The probability density function is shown in the following formula: In the above formula, P pv.max is the rated capacity of distributed photovoltaics; Γ(·) is the Gamma function; α and β are the first and second shape parameters of the Beta distribution, respectively; The first shape parameter and the second shape parameter are expressed as follows: In the above formula, P pv Inject actual power value into distributed photovoltaics; and are the expected value and standard deviation of the ratio of the actual value of distributed photovoltaic injection power to the rated capacity, respectively.

3. The method for generating a probabilistic optimization configuration scheme for DC distribution network measurement points according to claim 1, characterized in that: The method of constructing a measurement point probabilistic optimization configuration model for a DC distribution network using the measurement data, the grid parameters, and the node pseudo-measurement vector group includes: Constructing a probabilistic state estimation model of a DC distribution network according to the measurement data, the grid parameters and the node pseudo-measurement vector group; The probabilistic state estimation model is calculated using a quadratic constrained programming algorithm and a Gram-Charlie series expansion to obtain a cumulative distribution function of the node voltage estimation deviation; After determining the measurement point objective function of the DC distribution network, the deviation cumulative distribution function and the preset line measurement point configuration decision are used to constrain the objective function to construct a measurement point probability optimization configuration model, wherein the measurement point objective function is a function that corresponds to the minimum number of measurement point configurations in the DC distribution network.

4. The method for generating a probabilistic optimization configuration scheme for DC distribution network measurement points according to claim 3, characterized in that: The construction operation of the probabilistic state estimation model includes: Calculating a measurement value using the measurement data and the node pseudo-measurement vector group, and constructing an objective function of an estimation model by minimizing the absolute value sum of the difference between the measurement value and the measurement equation value of the node voltage state estimation quantity corresponding to the measurement value and a preset first equivalent variable to obtain an estimated objective function; A probabilistic state estimation model is constructed based on the estimation objective function and preset estimation constraints, wherein the estimation constraints include measurement equation constraints of line power with respect to node voltage state estimation, measurement equation constraints of line current with respect to node voltage state estimation, measurement equation constraints of load and distributed photovoltaic injection power with respect to node voltage state estimation, and preset upper and lower limit constraints of the first equivalent variable.

5. The method for generating a probabilistic optimization configuration scheme for DC distribution network measurement points according to claim 3, characterized in that: The probabilistic state estimation model is shown in the following formula: In the above formula, K is the total number of dimensions of the real-time measurement data of the line and the node pseudo-measurement vector of the load and distributed photovoltaic injection power; are the kth measurement value and state estimation value of the jth node pseudo-measurement vector respectively; is the measurement equation value containing the node voltage state estimation quantity involved in the pseudo measurement vector of the jth node; is the kth first equivalent variable involved in the pseudo-measurement vector of the jth node; the constraints c1, c2, c3, and c4 respectively represent the measurement equation constraints of the line power on the node voltage state estimate, the measurement equation constraints of the line current on the node voltage state estimate, the measurement equation constraints of the load and distributed photovoltaic injection power on the node voltage state estimate, and the upper and lower limit constraints of the preset first equivalent variable; are the line pq power measurement equation value, line pq current measurement equation value, and node p injected power measurement equation value of the j-th node pseudo measurement vector respectively; are the voltage state estimates of node p and node q involved in the jth node pseudo-measurement vector; b k is the kth binary configuration variable; M is a preset positive number; g pq is the conductance parameter of line pq; Ω is the node set.

6. The method for generating a probabilistic optimization configuration scheme for DC distribution network measurement points according to claim 3, characterized in that: The operation of calculating the deviation cumulative distribution function includes: Solving the probabilistic state estimation model according to the quadratic constrained programming algorithm to obtain a state estimation vector group about the voltage state of the DC distribution network node, and using the state estimation vector group to construct a voltage estimation deviation vector group; Calculating deviation origin moment information corresponding to the voltage estimation deviation vector group according to the weight vector; The deviation origin moment information is calculated using a Gram-Charlie series expansion to obtain a deviation cumulative distribution function of the node voltage estimation deviation.

7. The method for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network according to any one of claims 1 to 6, characterized in that: The grid parameters include topological connection relationship of node lines, line conductance parameters, and distributed photovoltaic rated capacity; The measurement data includes line real-time measurement data and node non-real-time data; The line real-time measurement data refers to line power and line current data that can be uploaded in real time; the node non-real-time data refers to node injection power data that cannot be uploaded in real time.

8. A device for generating a probabilistic optimization configuration scheme for measurement points in a DC distribution network, characterized in that: The device comprises: an acquisition module, configured to respectively acquire measurement data on lines and nodes of a DC distribution network, grid parameters of a DC distribution network grid, and a node pseudo-measurement vector group of loads and photovoltaic injection power of the DC distribution network; A construction module, configured to construct a probability optimization configuration model of measurement points of a DC distribution network using the measurement data, the grid parameters and the node pseudo-measurement vector group; A generation module, configured to solve the measurement point probability optimization configuration model according to a genetic algorithm to obtain a measurement point probability optimization configuration scheme for the DC distribution network; The operation of obtaining the node pseudo measurement vector group includes: Calculating a weight vector corresponding to a first pseudo measurement vector group of the load of the DC distribution network and the distributed photovoltaic injection power according to a three-point estimation algorithm, and constructing a vector cumulative distribution function according to the weight vector; After obtaining the second pseudo measurement vector group, a node pseudo measurement vector group of the load and photovoltaic injection power of the DC distribution network is calculated using the Nataf transform algorithm according to the vector cumulative distribution function and the second pseudo measurement vector group; The operation of obtaining the second pseudo measurement vector group includes: Construct a correlation coefficient matrix about the correlation between the load of each node in the DC distribution network and the distributed photovoltaic injection power; Decomposing the correlation coefficient matrix according to the Cholesky decomposition algorithm to obtain an upper triangular matrix and a lower triangular matrix; A second pseudo measurement vector group regarding the DC distribution network load and the distributed photovoltaic injection power is calculated using the lower triangular matrix, wherein the second pseudo measurement vector group satisfies the product sum of the lower triangular matrix and the first pseudo measurement vector group.

Images