A reactive power compensation optimization method and system based on grid-side energy storage

Abstract

The application provides a reactive power compensation optimization method and system based on grid side energy storage, and relates to the technical field of reactive power compensation, and comprises the following steps: acquiring all nodes in the power grid; establishing a basic reactive power demand model according to the power grid structure, the basic reactive power demand model being used to calculate the basic reactive power demand of each node according to real-time data; detecting the voltage value of each node in the last time window to form a time sequence voltage vector of each node respectively; based on the time sequence voltage vector, the residual reactive power resource capacity, the voltage disturbance resistance ability and the voltage out-of-limit degree of the node are fused to calculate the real-time reactive power distribution priority of each node; an optimization function is established and solved to obtain a reactive power compensation optimization scheme of the grid side energy storage in the next time window. The application has the advantage that a flexible and accurate reactive power compensation acquisition scheme is realized.

Description

Technical Field

[0001] This invention relates to the field of reactive power compensation technology, and more specifically, to a reactive power compensation optimization method and system based on grid-side energy storage. Background Technology

[0002] Reactive power compensation is an important technical measure in power systems used to regulate grid voltage, reduce network losses, and improve power supply quality.

[0003] While traditional reactive power compensation equipment can reduce reactive power deficits, it suffers from limitations such as slow response speed and lack of flexibility. With the development of energy storage technology, grid-side energy storage has demonstrated significant advantages in flexible and convenient reactive power regulation. However, in practical applications, grid-side energy storage generally faces the following challenges: the reactive power demand of each node in the grid varies, closely related to real-time node load status, line impedance, and network topology. Therefore, it is difficult to accurately trace reactive power demand solely from voltage fluctuations, while reactive power compensation based solely on the theoretical reactive power demand of nodes lacks flexibility, leading to difficulties in accurately obtaining the necessary reactive power compensation based on the reactive power demand of each node. Furthermore, node voltage is significantly affected by disturbances from neighboring nodes, and factors such as voltage disturbance resistance, remaining reactive power capacity, and voltage exceedance are interdependent. Reactive power allocation based solely on a single voltage deviation can easily lead to unreasonable allocation of compensation resources.

[0004] Therefore, it is necessary to optimize reactive power compensation based on grid-side energy storage to achieve a flexible and accurate reactive power compensation acquisition scheme. Summary of the Invention

[0005] The purpose of this invention is to provide a reactive power compensation optimization method and system based on grid-side energy storage, which can realize a flexible and accurate reactive power compensation acquisition scheme.

[0006] This invention is achieved through the following technical solution:

[0007] A reactive power compensation optimization method based on grid-side energy storage includes the following steps:

[0008] Obtain all nodes within the power grid;

[0009] A basic reactive power demand model is established based on the power grid structure. The basic reactive power demand model is used to calculate the basic reactive power demand of each node based on real-time data.

[0010] Within the previous time window, the voltage values ​​of each node are detected, forming the time-series voltage vector for each node. The length of one time window is... ;

[0011] Based on the time-series voltage vector, the remaining reactive power resource capacity, voltage disturbance resistance and voltage limit exceedance of the nodes are integrated to calculate the real-time reactive power allocation priority of each node.

[0012] Based on the basic reactive power demand of each node, the reactive power compensation optimization scheme of grid-side energy storage in the previous time window obtained by solving, and the real-time reactive power allocation priority of the previous time window calculated, an optimization function is established and solved to obtain the reactive power compensation optimization scheme of grid-side energy storage in the next time window.

[0013] Preferably, the method for establishing the basic reactive power demand model is as follows:

[0014] An initial model is established based on the power grid structure. The output of the initial model is the first... Initial value of reactive power demand at node at time t , It is a positive integer representing the node number. N is the total number of nodes;

[0015] According to the above Establish an optimal fitting function, and the output of the optimal fitting function is the first... The basic reactive power demand of the node at time point t ;

[0016] By selecting fitting samples from historical data to fit the optimized fitting function, the basic reactive power demand model is obtained.

[0017] Preferably, the method for establishing an initial model based on the power grid structure is as follows:

[0018] Establish the first Reactive power of the load at time t ;

[0019] ;

[0020] in, For the first The active power of the load at the node at time point t. For the first The power factor of a node at time t;

[0021] Get the The reactive power generated by a node at time t due to neighboring nodes :

[0022] ;

[0023] in, For the first The voltage at the node at time t, For the line susceptivity, Represents all connected to the first A set of lines for a node;

[0024] Obtain the value of the power generation unit or inverter at time t. reactive power output of nodes ;

[0025] Get the Initial value of reactive power demand at node at time t :

[0026] ;

[0027] According to the The method for establishing an optimal fitting function is as follows:

[0028] ;

[0029] in, , and All of these are parameters to be fitted.

[0030] Preferably, the method for selecting fitting samples from historical data is as follows:

[0031] The first The historical data of the node is arranged according to the second time window. Divide into, and get the first Nodes Historical data for each sub-time period, including data from multiple time points sampled periodically;

[0032] Based on the The node in the th Average voltage within the historical data of the sub-time period and voltage standard deviation , ;

[0033] Select the average voltage The voltage is greater than the preset voltage threshold and the voltage standard deviation Multiple sets of historical data within sub-time periods that are less than a preset standard deviation threshold are used as the first set of data. Fitted samples of nodes;

[0034] The fitting method is as follows:

[0035] Based on historical data from sub-time periods used as fitting samples, extract and calculate As input to the optimized fitting function, the actual reactive power compensation amount at the corresponding time point is obtained as the output of the optimized fitting function.

[0036] Preferably, the method for calculating the real-time reactive power allocation priority of each node is as follows:

[0037] According to the first Calculation of node's resistance to voltage disturbances and voltage exceedance. Node immunity to voltage disturbances and voltage limit exceedance evaluation parameters. It is a positive integer representing the node number. N is the total number of nodes;

[0038] Through the first Calculate the remaining reactive power resource capacity, voltage disturbance immunity evaluation parameters, and voltage over-limit evaluation parameters of the node. The priority of real-time reactive power allocation for nodes.

[0039] Preferably, calculate the first The voltage disturbance immunity evaluation parameters of the node The method is as follows:

[0040] ;

[0041] ;

[0042] ;

[0043] in, The set of all node numbers. Here, are intermediate parameters, and max and min are functions for finding the maximum and minimum values, respectively. To prevent constants with a denominator of 0, The first Node in the first time window The first of the time-series voltage vectors The element and the first One element, Representing the The set of numbers of the neighboring nodes of a node. For a node in the first time window The total number of elements in the time-series voltage vector. The weight is 'if', and 'if' is the truth function. The preset pressure difference threshold between adjacent time points, For the first Node and the Line impedance between nodes;

[0044] Calculate the first The voltage over-limit evaluation parameters of the node The method is as follows:

[0045] ;

[0046]

[0047] Where exp is the natural exponential function and B is an intermediate parameter. The standard voltage of the i-th node is... This is the preset voltage deviation threshold.

[0048] Preferably, the calculation of the first Real-time reactive power allocation priority of nodes The method is as follows:

[0049] ;

[0050] ;

[0051] Wherein, B is the fusion feature of the voltage disturbance immunity evaluation parameter and the voltage limit exceedance evaluation parameter. Furthermore, it is a parameter used to control the importance of remaining reactive power resource capacity. To control The coefficient of the rate at which the remaining reactive power resource capacity decreases. For the first The remaining reactive power capacity of the node. And for use in control The parameter of sensitivity to the fusion feature. And for use in control A parameter relating to the degree of nonlinearity of the fusion feature.

[0052] Preferably, the method for establishing the optimization function is as follows:

[0053] To prioritize reactive power compensation for nodes with higher real-time reactive power allocation priority and achieve lower consumption costs, an objective function and constraints are established.

[0054] Preferably, the objective function is:

[0055] ;

[0056] in, Representative solution The objective function is to minimize. For dimension The matrix and its first The element represents the first The change in reactive power compensation of the grid-side energy storage at the node in the next time window relative to the previous time window. For dimension The matrix and its first The element represents the first The reactive power compensation of the grid-side energy storage at the node in the previous time window. For dimension The matrix and its first The element represents the first The unit cost of reactive power compensation received by a node from grid-side energy storage; sum is a function for summing matrix elements. For dimension The matrix and its first The element represents the first The average voltage of the node in the previous time window, For dimension The matrix and its first The element represents the first Standard voltage of the node For dimension The matrix and its first The element represents the first The linear sensitivity of the node's voltage change relative to the reactive power compensation. , The total number of nodes;

[0057] The constraints include:

[0058] Deviation constraints of reactive power compensation at each node relative to the corresponding basic reactive power demand:

[0059] ;

[0060] in, For dimension The matrix, and its first... The element represents the first The average basic reactive power demand of the node in the previous time window, the first The average base reactive power demand of a node is obtained based on the aforementioned base reactive power demand model. The preset deviation threshold for basic reactive power demand;

[0061] Real-time reactive power allocation priority constraints:

[0062] ;

[0063] in, For dimension The matrix and its first The element represents the first The real-time reactive power allocation priority of the node, where a larger value indicates a higher priority for reactive power allocation. This is a preset constant threshold;

[0064] Single-node reactive power compensation total constraint:

[0065] ;

[0066] in, For the first The remaining reactive power resource capacity of the node.

[0067] This invention also provides a reactive power compensation optimization system based on grid-side energy storage, applied to the aforementioned reactive power compensation optimization method based on grid-side energy storage, comprising:

[0068] The node acquisition module is used to acquire all nodes within the power grid.

[0069] The basic reactive power demand model building module is used to build a basic reactive power demand model based on the power grid structure. The basic reactive power demand model is used to calculate the basic reactive power demand of each node based on real-time data.

[0070] The voltage acquisition module is used to detect the voltage values ​​of each node within the previous time window, forming a time-series voltage vector for each node. The length of one time window is... ;

[0071] The real-time reactive power allocation priority acquisition module is used to calculate the real-time reactive power allocation priority of each node based on the time-series voltage vector, by integrating the remaining reactive power resource capacity and voltage disturbance resistance of the node to the degree of voltage limit exceedance of the corresponding node.

[0072] The reactive power compensation optimization scheme acquisition module is used to establish and solve an optimization function based on the basic reactive power demand of each node, the reactive power compensation optimization scheme of grid-side energy storage in the previous time window obtained by solving, and the real-time reactive power allocation priority in the previous time window, so as to obtain the reactive power compensation optimization scheme of grid-side energy storage in the next time window.

[0073] The technical solution of the present invention has at least the following advantages and beneficial effects:

[0074] This invention establishes a basic reactive power demand model based on the structure of the power grid itself and real-time measurement data. It can estimate the basic reactive power demand of each node in real time based on the power grid operating status, thereby reducing the compensation deviation caused by relying solely on static experience values.

[0075] This invention utilizes the node voltage change trend and anti-voltage disturbance capability within the previous time window, and incorporates the node's remaining reactive power resource capacity to quantitatively evaluate the node's voltage over-limit risk. It also considers the reactive power compensation capability at the node to form a real-time reactive power allocation priority, so that nodes with large reactive power compensation gaps and sufficient reactive power compensation capabilities can receive priority compensation, which helps to improve voltage stability and reliability.

[0076] This invention enables energy storage devices to prioritize supporting nodes with the most urgent reactive power demand and the greatest compensation benefits by performing differentiated scheduling based on the priority of real-time reactive power allocation. This effectively reduces unnecessary reactive power output from energy storage, improves the utilization efficiency of reactive power compensation, and helps to enhance the cost-effectiveness of reactive power compensation.

[0077] This invention employs a sliding time window mechanism, which integrates the optimization results of the previous time window into the constraints of the next time window. This enables the reactive power regulation behavior of energy storage devices to be adjusted more flexibly and specifically according to voltage changes and load demands. Attached Figure Description

[0078] Figure 1 This is a flowchart illustrating the reactive power compensation optimization method based on grid-side energy storage provided in Embodiment 1 of the present invention.

[0079] Figure 2 This is a schematic diagram of the reactive power compensation optimization system based on grid-side energy storage provided in Embodiment 2 of the present invention. Detailed Implementation

[0080] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0081] Example 1

[0082] This embodiment provides a reactive power compensation optimization method based on grid-side energy storage. (See reference...) Figure 1 This includes the following steps:

[0083] Step S1: Obtain all nodes within the power grid. Nodes can be categorized based on the connection nodes possessing a reference voltage.

[0084] Step S2: Establish a basic reactive power demand model based on the power grid structure. The basic reactive power demand model is used to calculate the basic reactive power demand of each node based on real-time data.

[0085] As a preferred option, the method for establishing the basic reactive power demand model is as follows:

[0086] Step S201: Establish an initial model based on the power grid structure. The output of the initial model is the first... Initial value of reactive power demand at node at time t , It is a positive integer representing the node number. N is the total number of nodes, and the method is as follows:

[0087] Establish the first Reactive power of the load at time t ;

[0088] ;

[0089] in, For the first The active power of the load at the node at time point t. For the first The power factor of a node at time t;

[0090] Get the The reactive power generated by a node at time t due to neighboring nodes :

[0091] ;

[0092] in, For the first The voltage at the node at time t, For the line susceptivity, Represents all connected to the first A set of lines for a node;

[0093] Obtain the value of the power generation unit or inverter at time t. reactive power output of nodes ;

[0094] Get the Initial value of reactive power demand at node at time t :

[0095] .

[0096] The initial values ​​calculated above reflect the basic reactive power deficit of each node in the power grid under its current structure and operating conditions. The initial model starts from physical quantities reflecting circuit characteristics, such as node load, line susceptance, and reactive power output of generators / inverters, to obtain the initial reactive power demand reflecting the actual physical process of the power grid, ensuring the reliability of the calculation results.

[0097] Step S202: According to the above Establish an optimal fitting function, and the output of the optimal fitting function is the first... The basic reactive power demand of the node at time point t The method is as follows:

[0098] ;

[0099] in, , and All of these are parameters to be fitted.

[0100] Based on the initial values, a quadratic function is established for fitting, mapping the initial values ​​calculated based on the power grid physical model to a basic reactive power demand that better conforms to historical operating patterns. This also effectively characterizes the nonlinear characteristics of reactive power demand changes with operating conditions. In other words, this step, by fitting the initial values, compensates for the errors in the initial model under complex operating conditions, improving the accuracy of reactive power demand estimation.

[0101] Step S203: Select fitting samples from historical data to fit the optimized fitting function to obtain the basic reactive power demand model.

[0102] In this embodiment, the method for selecting fitting samples from historical data is as follows:

[0103] The first The historical data of the node is arranged according to the second time window. Divide into, and get the first Nodes Historical data for each sub-time period, including data from multiple time points sampled periodically;

[0104] Based on the The node in the th Average voltage within the historical data of the sub-time period and voltage standard deviation , ;

[0105] Select the average voltage The voltage is greater than the preset voltage threshold and the voltage standard deviation Multiple sets of historical data within sub-time periods that are less than a preset standard deviation threshold are used as the first set of data. Fitted samples for nodes.

[0106] Based on this, the fitting method is as follows:

[0107] Based on historical data from sub-time periods used as fitting samples, extract and calculate As input to the optimized fitting function, the actual reactive power compensation amount at the corresponding time point is obtained as the output of the optimized fitting function.

[0108] In this embodiment, historical samples are selected based on average voltage value and standard deviation, ensuring that the fitting only uses data under stable reactive power compensation conditions. In other words, the reactive power compensation amount of this data best matches the actual basic reactive power demand model, making the established model more representative of the real operating state and the parameters more reliable.

[0109] Step S3: Detect the voltage value of each node within the previous time window, and form the time-series voltage vector for each node. The length of one time window is... .

[0110] Step S4: Based on the time-series voltage vector, the remaining reactive power resource capacity, voltage disturbance resistance, and voltage limit exceedance of each node are integrated to calculate the real-time reactive power allocation priority of each node. The method is as follows:

[0111] Step S401: According to the first Calculation of node's resistance to voltage disturbances and voltage exceedance. Node immunity to voltage disturbances and voltage limit exceedance evaluation parameters. It is a positive integer representing the node number. N is the total number of nodes, and the method is as follows:

[0112] First, calculate the... The voltage disturbance immunity evaluation parameters of the node The method is as follows:

[0113] ;

[0114] ;

[0115] ;

[0116] in, The set of all node numbers. Here, are intermediate parameters, and max and min are functions for finding the maximum and minimum values, respectively. To prevent constants with a denominator of 0, The first Node in the first time window The first of the time-series voltage vectors The element and the first One element, Representing the The set of numbers of the neighboring nodes of a node. For a node in the first time window The total number of elements in the time-series voltage vector. The weight is 'if', and 'if' is the truth function. The preset pressure difference threshold between adjacent time points, For the first Node and the Line impedance between nodes.

[0117] This item can be found in adjacent nodes. When voltage fluctuations are too large, i.e., the difference from the standard voltage is too large, the first step is to determine... To determine whether nodes exhibit significant differences due to fluctuations, the average of multiple time points is extracted to avoid randomness. The IF function is used to filter out unimportant minor disturbances and normal data. Reflecting different adjacent nodes on the first The influence of nodes is determined by the fact that the greater the impedance, the easier it is for voltage to interact between two nodes. In other words, the likelihood of voltage changes being accidental is smaller, and therefore, the node is given a higher weight.

[0118] In summary, the voltage disturbance immunity evaluation parameters Aiming to quantify the first The sensitivity of a node to disturbances from adjacent nodes is used to evaluate whether a node can maintain voltage stability. The larger it is, the more likely it is to be the first The more easily a node voltage change is amplified by disturbances from adjacent nodes, the worse the disturbance immunity. The voltage disturbance immunity evaluation parameters obtained after normalization are... The closer it is to 1, the more likely it is to be the first. Nodes are more susceptible to power grid disturbances.

[0119] Secondly, calculate the first The voltage over-limit evaluation parameters of the node The method is as follows:

[0120] ;

[0121]

[0122] Where exp is the natural exponential function and B is an intermediate parameter. The standard voltage of the i-th node is... This is the preset voltage deviation threshold.

[0123] Voltage over-limit evaluation parameters Used to measure the degree to which a node voltage deviates from its standard voltage. When the overall node voltage deviates significantly from the standard voltage, the value of B will be greater than 1. The greater the deviation of the value from 1, the higher the corresponding value. The value will approach 1 more and more.

[0124] In other words, voltage disturbance immunity evaluation parameters Voltage over-limit evaluation parameters They have all been standardized to a numerical range of 0-1.

[0125] Step S402: Through the first Calculate the remaining reactive power resource capacity, voltage disturbance immunity evaluation parameters, and voltage over-limit evaluation parameters of the node. The priority of real-time reactive power allocation for nodes.

[0126] Here, the remaining reactive power resource capacity, voltage disturbance immunity, and voltage limit exceedance of a node are integrated to obtain the first... The real-time reactive power allocation priority of the nodes. This embodiment calculates the priority of the first node. Real-time reactive power allocation priority of nodes The method is as follows:

[0127] ;

[0128] ;

[0129] Wherein, B is the fusion feature of the voltage disturbance immunity evaluation parameter and the voltage limit exceedance evaluation parameter. Furthermore, it is a parameter used to control the importance of remaining reactive power resource capacity. To control The coefficient of the rate at which the remaining reactive power resource capacity decreases. For the first The remaining reactive power capacity of the node. And for use in control The parameter of sensitivity to the fusion feature. And for use in control A parameter relating to the degree of nonlinearity of the fusion feature.

[0130] exist Under the influence of one factor, when the remaining reactive resources are less, The greater the attenuation, the less likely tasks will be assigned to nodes without resources. This is especially important when a node has poor disturbance tolerance or is at risk of exceeding voltage limits. The contribution increases, raising the probability of priority compensation for the corresponding node. During the fusion of voltage disturbance immunity evaluation parameters and voltage limit exceedance evaluation parameters, It can enhance the capture of critical nodes that are both susceptible to disturbances and have voltage limits.

[0131] Step S5: Based on the basic reactive power demand of each node, the reactive power compensation optimization scheme of grid-side energy storage in the previous time window obtained by solving, and the real-time reactive power allocation priority in the previous time window obtained by calculation, establish an optimization function and solve it to obtain the reactive power compensation optimization scheme of grid-side energy storage in the next time window.

[0132] As a preferred embodiment, the method for establishing the optimization function is as follows:

[0133] To prioritize reactive power compensation for nodes with higher real-time reactive power allocation priority and achieve lower consumption costs, an objective function and constraints are established.

[0134] Based on this, the objective function is:

[0135] ;

[0136] in, Representative solution The objective function is to minimize. For dimension The matrix and its first The element represents the first The change in reactive power compensation of the grid-side energy storage at the node in the next time window relative to the previous time window. For dimension The matrix and its first The element represents the first The reactive power compensation of the grid-side energy storage at the node in the previous time window. For dimension The matrix and its first The element represents the first The unit cost of reactive power compensation received by a node from grid-side energy storage; sum is a function for summing matrix elements. For dimension The matrix and its first The element represents the first The average voltage of the node in the previous time window, For dimension The matrix and its first The element represents the first Standard voltage of the node For dimension The matrix and its first The element represents the first The linear sensitivity of the node's voltage change relative to the reactive power compensation. , This represents the total number of nodes.

[0137] Based on the above objective function, Indicates the total compensation cost. The addition of this item aims to reduce the degree to which the voltage deviates from the standard value after reactive power compensation adjustment.

[0138] The constraints include:

[0139] Deviation constraints of reactive power compensation at each node relative to the corresponding basic reactive power demand:

[0140] ;

[0141] in, For dimension The matrix, and its first... The element represents the first The average basic reactive power demand of the node in the previous time window, the first The average base reactive power demand of a node is obtained based on the aforementioned base reactive power demand model. This is the preset deviation threshold for basic reactive power demand.

[0142] This constraint makes the optimization results closer to the basic reactive power demand, improves the rationality of compensation, and can also avoid waste.

[0143] Real-time reactive power allocation priority constraints:

[0144] ;

[0145] in, For dimension The matrix and its first The element represents the first The real-time reactive power allocation priority of the node, where a larger value indicates a higher priority for reactive power allocation. This is a preset constant threshold.

[0146] Under the constraint of real-time reactive power allocation priority, the optimal solution must make the overall compensation direction consistent with the real-time reactive power allocation priority, so that high-priority nodes can obtain as much reactive power compensation resources as possible in the next time window.

[0147] Finally, there are the constraints on the total reactive power compensation of a single node to ensure the safety and feasibility of system operation:

[0148] ;

[0149] in, For the first The remaining reactive power resource capacity of the node.

[0150] It should be noted that, for a power grid, steps S1-S2 can be performed once if the basic characteristics such as the circuit topology remain unchanged, and then the remaining steps can be performed periodically.

[0151] Example 2

[0152] This embodiment provides a reactive power compensation optimization system based on grid-side energy storage, applied to the reactive power compensation optimization method based on grid-side energy storage described in the above embodiment. (See reference...) Figure 2 ,include:

[0153] The node acquisition module is used to acquire all nodes within the power grid.

[0154] The basic reactive power demand model building module is used to build a basic reactive power demand model based on the power grid structure. The basic reactive power demand model is used to calculate the basic reactive power demand of each node based on real-time data.

[0155] The voltage acquisition module is used to detect the voltage values ​​of each node within the previous time window, forming a time-series voltage vector for each node. The length of one time window is... ;

[0156] The real-time reactive power allocation priority acquisition module is used to calculate the real-time reactive power allocation priority of each node based on the time-series voltage vector, by integrating the remaining reactive power resource capacity and voltage disturbance resistance of the node to the degree of voltage limit exceedance of the corresponding node.

[0157] The reactive power compensation optimization scheme acquisition module is used to establish and solve an optimization function based on the basic reactive power demand of each node, the reactive power compensation optimization scheme of grid-side energy storage in the previous time window obtained by solving, and the real-time reactive power allocation priority in the previous time window, so as to obtain the reactive power compensation optimization scheme of grid-side energy storage in the next time window.

[0158] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing reactive power compensation based on grid-side energy storage, characterized in that, The method comprises the following steps: acquiring all nodes in a power grid; establishing a basic reactive power demand model according to a power grid structure, the basic reactive power demand model being used to calculate basic reactive power demands of the nodes according to real-time data; The voltage values of the nodes are detected in a previous time window, and time sequence voltage vectors of the nodes are formed respectively, and the length of one time window is ; calculating real-time reactive power distribution priorities of the nodes based on time-series voltage vectors, by fusing residual reactive power resource capacities, voltage disturbance resistance capabilities and voltage out-of-limit degrees of the nodes; establishing an optimization function according to the basic reactive power demands of the nodes, a reactive power compensation optimization scheme of a power grid-side energy storage obtained by solving in a previous time window and the real-time reactive power distribution priorities, and solving the optimization function to obtain a reactive power compensation optimization scheme of the power grid-side energy storage in a next time window.

2. The method of claim 1, wherein, The method for establishing the basic reactive power demand model comprises the following steps: According to the power grid structure, an initial model is established, and an output of the initial model is a first The initial value of the reactive power demand of the node at the time point t , is a positive integer and represents a node number, , and N is a total number of nodes. According to the described establishing an optimization fitting function, an output of the optimization fitting function is the first the base reactive demand quantity of the node at the time point t ; fitting an optimization fitting function by selecting fitting samples from historical data to obtain the basic reactive power demand model.

3. The method of claim 2, wherein, The method for establishing the initial model according to the power grid structure comprises the following steps: establishing a first the load reactive power of the node at time point t ; ； wherein, is the first active power of the node at the time point t, is the first power factor of the node at the time point t; acquiring a first reactive power generated by the node at the time point t due to the adjacent node : ； wherein, is the first node voltage at time point t, is the line susceptance, represents the set of all lines connected to the first node; The reactive output of the node at time point t is obtained by the reactive output of the node ; acquiring a first initial value of the reactive power demand of the node at the time point t : ； The method according to the The method for establishing the optimization fitting function is: ； wherein, , and are parameters to be fitted.

4. The method of claim 3, wherein, The method for selecting the fitting samples from the historical data comprises the following steps: The first The historical data of the node is arranged according to the second time window. Divide into, and get the first Nodes Historical data for each sub-time period, including data from multiple time points sampled periodically; based on the first the voltage average value of the node in the first group sub-period historical data and voltage standard deviation , ; selected voltage average value greater than a preset voltage threshold and voltage standard deviation a plurality of groups of historical data of sub-time periods, which are less than a preset standard deviation threshold, as the first fitting sample of the node; The fitting method comprises the following steps: Based on the historical data of the sub-time period as the fitting sample, extract and calculate As the input of the optimization fitting function, the actual reactive compensation quantity of the corresponding time point is obtained as the output of the optimization fitting function.

5. The method of claim 1, wherein, The method for calculating the real-time reactive power distribution priorities of the nodes comprises the following steps: According to the first The voltage disturbance resistance capability and the voltage overrun degree of the node are calculated to obtain the voltage disturbance resistance evaluation parameter and the voltage overrun evaluation parameter of the node The voltage disturbance resistance capability and the voltage overrun degree of the node are calculated to obtain the voltage disturbance resistance evaluation parameter and the voltage overrun evaluation parameter of the node is a positive integer and represents the node number, N is the total number of nodes. The real-time reactive power distribution priority of the node is calculated by the remaining reactive power resource capacity, the voltage disturbance resistance evaluation parameter and the voltage out-of-limit evaluation parameter of the node. The real-time reactive power distribution priority of the node is calculated by the remaining reactive power resource capacity, the voltage disturbance resistance evaluation parameter and the voltage out-of-limit evaluation parameter of the node. The real-time reactive power distribution priority of the node is calculated by the remaining reactive power resource capacity, the voltage disturbance resistance 6. The method of claim 5, wherein, The method comprises the following steps of: The anti-voltage disturbance evaluation parameter of the node The method comprises the following steps of: ； ； ； in, The set of all node numbers. Here, is an intermediate parameter, and max and min are functions for finding the maximum and minimum values, respectively. To prevent constants with a denominator of 0, The first Node in the first time window The first of the time-series voltage vectors The element and the first One element, Representing the The set of numbers of the neighboring nodes of a node. For a node in the first time window The total number of elements in the time-series voltage vector. The weight is 'if', and 'if' is the truth function. The preset pressure difference threshold between adjacent time points, For the first Node and the Line impedance between nodes; The method for calculating the The voltage out-of-limit evaluation parameter of the node The method is as follows: ； wherein exp is the natural exponential function, B is an intermediate parameter, is a standard voltage of the i-th node, is a preset voltage deviation threshold.

7. The method of claim 6, wherein, calculating the first real-time reactive power distribution priority of the node The method is as follows: ； ； B is a fusion feature of the voltage disturbance resistance evaluation parameter and the voltage excursion evaluation parameter, and is a parameter for controlling the importance of the remaining reactive power resource capacity, is a control coefficient of the speed of decay with the decrease of the remaining reactive power resource capacity, is the first the remaining reactive power resource capacity of the node, and is a parameter for controlling the sensitivity to the fusion feature, and is a parameter for controlling the nonlinearity degree to the fusion feature.

8. The method of claim 1, wherein, The method for establishing the optimization function comprises the following steps: A target function and a constraint are established for the purpose of preferentially allocating reactive power compensation to nodes with higher real-time reactive power distribution priorities and reducing consumption cost.

9. The method of claim 8, wherein, The target function is: ； in, Representative solution The objective function is to minimize. For dimension The matrix and its first The element represents the first The change in reactive power compensation of the grid-side energy storage at the node in the next time window relative to the previous time window. For dimension The matrix and its first The element represents the first The reactive power compensation of the grid-side energy storage at the node in the previous time window. For dimension The matrix and its first The element represents the first The unit cost of reactive power compensation received by a node from grid-side energy storage; sum is a function for summing matrix elements. For dimension The matrix and its first The element represents the first The average voltage of the node in the previous time window, For dimension The matrix and its first The element represents the first Standard voltage of the node For dimension The matrix and its first The element represents the first The linear sensitivity of the node's voltage change relative to the reactive power compensation. , The total number of nodes; The constraint comprises: a deviation constraint of the reactive power compensation of each node relative to the corresponding basic reactive power demand; ； wherein, is a matrix of dimensions , and its element represents the average base reactive demand amount of the node on an average basis of the previous time window, the average base reactive demand amount of the node being obtained based on the base reactive demand model, is a preset base reactive demand deviation threshold value; a real-time reactive power distribution priority constraint; ； wherein, is a matrix of dimension and its th element represents the real-time reactive power allocation priority of the th node, the larger the value of the real-time reactive power allocation priority represents the higher the priority of reactive power allocation, is a preset constant threshold; a total amount constraint of the reactive power compensation of each node; ； wherein, the first the remaining reactive resource capacity of the node.

10. A system for optimizing reactive power compensation based on grid-side energy storage, applied to the method for optimizing reactive power compensation based on grid-side energy storage according to any one of claims 1-9, characterized in that, The method comprises the following steps: a node acquisition module, configured to acquire all nodes in a power grid; a basic reactive power demand model establishment module, configured to establish a basic reactive power demand model according to a power grid structure, the basic reactive power demand model being used to calculate basic reactive power demands of the nodes according to real-time data; A voltage acquisition module is configured to detect voltage values of each node in a previous time window, and form a time sequence voltage vector of each node, wherein the length of one time window is ; a real-time reactive power distribution priority acquisition module, configured to calculate real-time reactive power distribution priorities of the nodes based on time-series voltage vectors, by fusing residual reactive power resource capacities, voltage disturbance resistance capabilities and voltage out-of-limit degrees of the nodes; a reactive power compensation optimization scheme acquisition module, configured to establish an optimization function according to the basic reactive power demands of the nodes, a reactive power compensation optimization scheme of a power grid-side energy storage obtained by solving in a previous time window and the real-time reactive power distribution priorities calculated in the previous time window, and solve the optimization function to obtain a reactive power compensation optimization scheme of the power grid-side energy storage in a next time window.

Images