A Method for Emergency Voltage Regulation of Distributed Resources in Distribution Networks Based on Hierarchical Interactive Multidimensional Dynamic Aggregation

Abstract

The present invention discloses a method for emergency voltage regulation of distributed resources in a distribution network based on hierarchical interactive multi-dimensional dynamic aggregation, comprising the following steps: S1. Obtain the values of the priority indicators of various types of distributed resources according to the voltage regulation costs and tracking response characteristics of the distributed resources in the distribution network; S2. Sort various types of distributed resources according to the values of the priority indicators of various types of distributed resources; S3. Implement the lower-layer aggregation strategy; self-organize and aggregate various types of distributed resources according to the priority sorting results, and the formed aggregation unit responds to the voltage regulation demand after receiving the scheduling command and continuously adjusts its distributed resource scheduling scheme; S4. Implement the upper-layer adjustment strategy; optimize and control the voltage of the overall power grid according to the real-time data adjusted by the lower-layer aggregation strategy; S5. Feed back the real-time data of the overall power grid obtained after implementing the upper-layer adjustment strategy to S3 to perform the next round of lower-layer aggregation strategy, forming a dynamic closed-loop nested control mechanism.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and particularly to a method for emergency voltage regulation of distributed resources in a distribution network based on hierarchical interactive multi-dimensional dynamic aggregation. Background Art

[0002] With the continuous increase in the access scale of multi-source heterogeneous resources, the distribution network has changed from a traditional single-source and radial structure to a modern source-network-load-storage deeply coupled system. Due to the large number, uneven capacity, and scattered layout of distributed resources, the direct dispatching cost by the power grid dispatching center is too high, and it is difficult to be directly dispatched by the power grid. At the same time, the massive distributed resources have high randomness, intermittency, and uncertainty, bringing great operational challenges to the traditional distribution network mainly based on centralized control, resulting in problems such as two-way power flow and reverse overload in the distribution network, which are prone to cause sharp fluctuations in feeder power and voltage over-limit, and even seriously affect the operational stability of the distribution network in severe cases.

[0003] Under this background, aggregation regulation can solve the problem that a large number of distributed resources are difficult to be directly dispatched. At present, the main aggregation methods for flexible resources include the Minkowski summation method, the Chino polyhedron method, the virtual battery model method, the vertex enumeration method, etc. However, due to problems such as uneven scheduling response performance of distributed resources, it is easy to cause disordered scheduling. Therefore, it is necessary to consider the voltage regulation cost and tracking response characteristics of distributed resources in the distribution network, and propose a multi-dimensional self-organizing aggregation method for active-reactive power of distributed resources based on comprehensive priority sorting. On the other hand, the power grid needs to have fast response capabilities, while the existing centralized control methods often have a lag in response and are difficult to meet emergency requirements, and the current emergency voltage regulation research lacks hierarchical interactive collaboration and efficient scheduling of resources within the substation area. Summary of the Invention

[0004] In view of this, the present invention provides a method for emergency voltage regulation of distributed resources in a distribution network based on hierarchical interactive multi-dimensional dynamic aggregation, so as to at least solve the problems of lack of hierarchical interactive collaboration and efficient scheduling of resources within the substation area in the prior art emergency voltage regulation research.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] A method for emergency voltage regulation of distributed resources in a distribution network based on hierarchical interactive multi-dimensional dynamic aggregation includes the following steps:

[0007] S1. According to the voltage regulation cost and tracking response characteristics of distributed resources in the distribution network, obtain the values of the priority indicators of various types of distributed resources;

[0008] S2. Sort various types of distributed resources according to the values of the priority indicators of various types of distributed resources;

[0009] S3. Implement the lower-layer aggregation strategy; self-organize and aggregate various types of distributed resources according to the sorted priority results. The formed aggregation unit responds to the voltage regulation demand after receiving the scheduling command and continuously adjusts its distributed resource scheduling plan;

[0010] S4. Implement the upper-layer adjustment strategy; optimize and regulate the voltage of the overall power grid based on the real-time data adjusted by the lower-layer aggregation strategy, including: performing long-time scale scheduling on on-load tap changers and capacitor banks, and performing short-time scale scheduling on aggregation units based on the long-time scale scheduling;

[0011] S5. Feed back the real-time data of the overall power grid obtained after implementing the upper-layer adjustment strategy to S3, perform the next round of lower-layer aggregation strategy, and form a dynamic closed-loop nested regulation mechanism.

[0012] Preferably, the priority indicators of each type of distributed resource in S1 respectively include voltage sensitivity, voltage regulation cost, and tracking response coefficient;

[0013] (1) Voltage sensitivity:

[0014] ,

[0015] In the formula, and are the active and reactive voltage sensitivities of node to node respectively; is the voltage at the beginning of the line; and are the resistance and reactance between node and node respectively;

[0016] (2) The voltage regulation costs for each type include the voltage regulation cost of distributed photovoltaic equipment, the voltage regulation cost of energy storage equipment, the voltage regulation cost of electric vehicles, and the voltage regulation cost of SVG equipment;

[0017] 1) Voltage regulation cost of distributed photovoltaic equipment:

[0018] The total voltage regulation cost of photovoltaic discharge is:

[0019] ,

[0020] In the formula, is the depreciation cost of photovoltaic equipment; is the unit power maintenance cost of photovoltaic equipment; is the active power loss cost generated per unit power of photovoltaic equipment;

[0021] Then node The photovoltaic device for the node Unit reactive power regulation cost That is, the voltage regulation cost of the distributed photovoltaic device is:

[0022] ,

[0023] In the formula, is the voltage change amount before and after voltage regulation of the node ; is the reactive power output by the photovoltaic device;

[0024] 2) Energy storage device voltage regulation cost:

[0025] The total cost of the reactive power regulation power of the energy storage device is:

[0026] ,

[0027] In the formula, is the loss degree of the energy storage battery life under different depths of charge and discharge; is the total effective discharge capacity of the energy storage battery; is the initial investment cost of the energy storage battery; is the unit power maintenance cost of the energy storage; is the active power loss cost generated by unit power;

[0028] Then the energy storage device of the node for the node Unit reactive power regulation cost That is, the voltage regulation cost of the energy storage device is:

[0029] ,

[0030] In the formula, is the reactive power output by the energy storage device;

[0031] 3) Electric vehicle voltage regulation cost:

[0032] The total voltage regulation cost of the electric vehicle is:

[0033] ,

[0034] In the formula, is the charging unit cost of the electric vehicle; is the charging efficiency of the electric vehicle; is the charging power of the electric vehicle; is the charging duration of the electric vehicle at the node ; ; is the energy loss of the electric vehicle in each charging cycle

[0035] Then the node Electric vehicles at the node Unit active voltage regulation cost That is, the voltage regulation cost of electric vehicles is:

[0036] ,

[0037] In the formula, It is the active power adjusted by the electric vehicle due to voltage regulation requirements;

[0038] 4) SVG equipment voltage regulation cost:

[0039] Total cost of SVG equipment participating in grid voltage regulation for:

[0040] ,

[0041] In the formula, is the initial investment cost of the SVG equipment; For operating costs; For maintenance costs; For energy consumption cost;

[0042] Then the node SVG device pair node Unit reactive voltage regulation cost That is, the voltage regulation cost of SVG equipment is:

[0043] ,

[0044] In the formula, Reactive power output by SVG equipment;

[0045] (3) Tracking response coefficient:

[0046] During the scheduling period, the output of distributed resources is expressed in per-unit form as:

[0047] ,

[0048] In the formula, Individuals of distributed resources In the period The power per unit value; For individuals In the period The power value; For individuals The maximum power value; is the number of time periods in the scheduling period;

[0049] Power signal and load signal During the consistency index of the time period is:

[0050] ,

[0051] wherein, is the per-unit value of the power of the load signal during the time period ;

[0052] The consistency between the power signal of the individual and the load signal within time periods is measured by tracking the response coefficient :

[0053] ,

[0054] wherein, The smaller it is, the more consistent the power signal of the individual and the load signal .

[0055] Preferably, the specific content of S2 includes:

[0056] Normalize all priority indicators, and construct radar charts with voltage sensitivity, voltage regulation cost, and response speed as vertices for different types of distributed resources according to the normalized indicator values;

[0057] Select two eigenvectors, the area and perimeter of each radar chart, to construct an evaluation function H , and calculate the evaluation values H of each radar chart in turn;

[0058] According to the magnitudes of the respective evaluation values H , perform priority ranking.

[0059] Preferably, the specific content of the normalization includes:

[0060] Normalize the values of each priority indicator to the interval [0,1]:

[0061] ,

[0062] wherein, represents the minimum value of the priority indicator; represents the maximum value of the priority indicator;

[0063] The area and perimeter of each radar chart are respectively:

[0064] ,​​​​​​​​​

[0065] In the formula, and are the area and perimeter of the radar chart, is the side length formed by the th normalized index value; is the th included angle between the indexes;

[0066] Calculate the evaluation value of each radar chart in sequence H The specific content of is:

[0067] .

[0068] Preferably, the specific content of S3 includes:

[0069] Characterize the total aggregated power based on the current aggregation unit feasible region, is the active power of the aggregation unit in the time period , is the reactive power of the aggregation unit in the time period , to obtain the elliptical feasible region , and adopt the time-decoupled elliptical feasible region :

[0070] ,

[0071] In the formula, is the center of the ellipse; is a two-dimensional positive semi-definite matrix, representing the rotation and stretching transformation of the ellipse; is the variable constrained by the uncertainty set U:

[0072] ,

[0073] In the formula, is the set of all time periods ;

[0074] Take the feasible region as the uncertainty set, and take as the uncertainty set under the uncertainty variable, the distributed resource scheduling scheme is an adaptive variable that changes dynamically with the uncertainty variable. Establish an active-reactive multi-dimensional dynamic aggregation model, aiming at maximizing the total aggregation flexibility of the active-reactive power in different time periods, and perform optimal aggregation on the active-reactive flexibility. Among them, the active-reactive multi-dimensional dynamic aggregation model is:

[0075] , ​

[0076] ,

[0077] ,

[0078] ,

[0079] ,

[0080] In the formula, the elliptical feasible region is dynamically adjusted as changes, and a distributed resource power dispatch scheme is correspondingly obtained ; represents the determinant of matrix , which is equal to the area of the elliptical feasible region in time period ; and both represent the power constraint parameters of the elliptical feasible region; and respectively represent the apparent power capacity constraint parameters of the photovoltaic device and the energy storage device; and respectively represent other power constraints and network constraint parameters of the distributed resources;

[0081] Solve the active-reactive multi-dimensional dynamic aggregation model to obtain the optimal flexibility space of the optimal distributed resources, so as to maximize the elliptical feasible region.

[0082] Preferably, the specific content of the upper-layer long-time scale hourly scheduling model in S4 includes:

[0083] The distribution network layer comprehensively considers the action times of the on-load tap changer OLTC and the capacitor bank CB, constructs an OLTC and CB collaborative voltage regulation action strategy with the minimum system network loss as the goal, and completes the hourly optimization. The objective function is:

[0084] ,

[0085] In the formula, and are respectively the resistance and current of branch in time period ; is the number of time periods in a scheduling cycle; is the duration of each time period;

[0086] The constraint conditions include: OLTC constraint and CB constraint; among them,

[0087] The OLTC constraint is:

[0088] ,

[0089] ,

[0090] ,

[0091] ,

[0092] In the formula, is the voltage of node in time period ; is the tap ratio of the on-load tap-changer in time period ; is the number of tap positions of the on-load tap-changer; is the square of the reference voltage of node when the on-load tap-changer is in tap position in time period ; and are the active power, reactive power and reactance of branch in time period respectively; is an arbitrary positive number; represents a binary variable indicating whether the on-load tap-changer is in tap position in time period ; is the maximum number of transformer switching operations within a scheduling period;

[0093] The CB constraint is:

[0094] ,

[0095] ,

[0096] ,

[0097] ,

[0098] In the formula, is the output value of the switched capacitor bank in time period ; is the number of switched capacitor banks in time period ; is the reactive power compensation of a single capacitor; and are the upper and lower limits of the output of the switched capacitor bank; is the specific number of installed capacitor banks; is the upper limit of the number of switched capacitor banks in time period ; is the maximum number of capacitor switching operations; is the sign function, when a > 0, its value is 1, when a = 0, its value is 0.

[0099] Preferably, the specific content of the short-term scale minute-level control model in S4 includes:

[0100] After the on-load tap changer and capacitor bank switching operations are determined by the upper-layer long-term scale hour-level scheduling model, deterministic optimization of the active and reactive power of various distributed resources at the substation area layer is performed using a preset scheduling interval;

[0101] Construct a short-term scale optimization objective function with the minimum system node voltage deviation:

[0102] ,

[0103] In the formula, is the number of distribution network nodes; represents node at time period voltage; represents the rated voltage of the node at time period

[0104] As can be seen from the above technical solutions, compared with the prior art, the present invention discloses a method for emergency voltage regulation of distributed resources in a distribution network based on hierarchical interactive multi-dimensional dynamic aggregation, which has the following beneficial effects:

[0105] By considering the comprehensive regulation characteristics of distributed resources, the present invention proposes a method for multi-dimensional self-organization aggregation of active-reactive power of distributed resources based on comprehensive priority sorting. Compared with existing aggregation methods, it solves the problems of low scheduling efficiency, inaccurate voltage regulation effect, and lack of interactive coordination mechanism of distributed resources in the existing distribution network. The proposed emergency voltage regulation strategy for distributed resources based on hierarchical interactive dynamic aggregation nesting, compared with traditional voltage regulation technologies, decomposes the emergency voltage regulation problem into multiple levels, and through scheduling control at different time scales, has a faster response speed and higher accuracy for fault voltage recovery, ensuring precise control and efficient regulation of distributed resources in large-scale applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0106] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0107] Figure 1 is a flowchart of a method for emergency voltage regulation of distributed resources in a distribution network based on hierarchical interactive multi-dimensional dynamic aggregation disclosed by the present invention;

[0108] ​Figure 2 It is the architecture diagram of the distributed resource emergency voltage regulation based on hierarchical interactive dynamic aggregation and nesting provided by the embodiment of the present invention;

[0109] Figure 3 It is the radar chart of the comprehensive priority index of different resources at a certain moment provided by the embodiment of the present invention, Figure 3 (a) is the radar chart of the comprehensive priority index of the energy storage device, Figure 3 (b) is the radar chart of the comprehensive priority index of the photovoltaic device, Figure 3 (c) is the radar chart of the comprehensive priority index of the electric vehicle, Figure 3 (d) is the radar chart of the comprehensive priority index of the SVG. Specific embodiments

[0110] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0111] The present invention provides a method for emergency voltage regulation of distributed resources in a distribution network based on hierarchical interactive multi-dimensional dynamic aggregation, as Figure 1 shown, including the following steps:

[0112] S1. According to the voltage regulation cost and tracking response characteristics of distributed resources in the distribution network, obtain the values of the priority indicators of various types of distributed resources;

[0113] S2. Sort various types of distributed resources according to the values of the priority indicators of various types of distributed resources;

[0114] S3. Perform the lower-layer aggregation strategy; self-organize and aggregate various types of distributed resources according to the priority sorting results. The formed aggregation unit responds to the voltage regulation demand after receiving the dispatching command and continuously adjusts its distributed resource dispatching scheme;

[0115] S4. Perform the upper-layer adjustment strategy; optimize and control the voltage of the overall power grid according to the real-time data adjusted by the lower-layer aggregation strategy, including: performing long-time scale dispatching on on-load tap changers and capacitor banks, and performing short-time scale dispatching on the aggregation unit based on the long-time scale dispatching;

[0116] S5. Feed back the real-time data of the overall power grid obtained after performing the upper-layer adjustment strategy to S3 to perform the next round of lower-layer aggregation strategy, forming a dynamic closed-loop nested control mechanism.

[0117] It should be noted that:

[0118] In this embodiment, considering the voltage regulation costs and tracking response characteristics of distributed resources such as photovoltaic devices, energy storage devices, electric vehicles, and static var generators (SVG) in the distribution network, accurate indicators for the adjustable resource ranking and optimized flexibility space are obtained.

[0119] Based on the hierarchical interactive dynamic aggregation nested distributed resource emergency voltage regulation strategy, the core idea is to fully utilize the flexibility and adjustability of various distributed resources in the grid emergency voltage regulation through the hierarchical interaction structure and dynamic aggregation nested mechanism, optimize the voltage regulation effect, and the architecture diagram is as Figure 2 shown.

[0120] Through the bottom-up order, the lower layer self-organizes and aggregates multiple distributed resources according to the priority ranking results. After the formed aggregation unit receives the dispatching command, it responds to the voltage regulation demand based on local information and continuously adjusts the dispatching strategy and power distribution of its internal resources. These local regulation results and feedback will be transmitted back to the upper layer in real time, providing a basis for the upper layer decision-making. The upper layer optimally regulates the voltage of the overall power grid by comprehensively considering the feedback information from different levels and guides the adjustment of the lower layer, forming a dynamic closed-loop nested regulation mechanism. Through this interactive hierarchical aggregation and nesting, it can ensure that in the event of an emergency in the power grid, the resources at each level can be flexibly allocated according to real-time demands, quickly respond to the power grid voltage fluctuations, ensure the stability and security of the power grid, and at the same time improve the overall utilization efficiency and dispatching flexibility of the resources.

[0121] To further implement the above technical solution, the priority indicators of each type of distributed resource in S1 respectively include voltage sensitivity, voltage regulation cost, and tracking response coefficient;

[0122] (1) Voltage sensitivity:

[0123] Voltage sensitivity is an important parameter to measure the voltage regulation effect of active and reactive power of each node. The traditional calculation method depends on the real-time power of the whole network and is usually calculated by a centralized control server. However, to realize the application of voltage sensitivity in the distributed control strategy, the present invention approximately calculates the voltage sensitivity by using the line impedance of the distribution network:

[0124] ,

[0125] In the formula, and are respectively the active and reactive voltage sensitivities of node to node ; is the voltage at the beginning of the line; and are respectively node and node The resistance and reactance between;

[0126] (2) For each type of voltage regulation cost, including the voltage regulation cost of distributed photovoltaic equipment, energy storage equipment, electric vehicle voltage regulation cost, and SVG equipment voltage regulation cost;

[0127] 1) Voltage regulation cost of distributed photovoltaic equipment:

[0128] Photovoltaic equipment is generally operated by operators. To ensure power generation revenue, the output of its active power is usually not reduced. At the same time, the capacity of photovoltaic equipment is usually greater than the rated capacity. Therefore, reactive power regulation can be carried out without affecting the output of photovoltaic active power. Then the total cost of photovoltaic discharge voltage regulation is:

[0129] ,

[0130] In the formula, is the depreciation cost of photovoltaic equipment; is the unit power maintenance cost of photovoltaic equipment; is the active power loss cost generated per unit power of photovoltaic equipment;

[0131] Then for node the photovoltaic equipment of node unit reactive power voltage regulation cost That is, the voltage regulation cost of distributed photovoltaic equipment is:

[0132] ,

[0133] In the formula, is the voltage change amount before and after voltage regulation of node ; is the reactive power output by photovoltaic equipment;

[0134] 2) Voltage regulation cost of energy storage equipment:

[0135] During the voltage regulation process, energy storage equipment usually adjusts reactive power first, and then adjusts active power as needed. This invention mainly considers the reactive power regulation of energy storage equipment. Frequent charge and discharge operations, high-power operation, and deep discharge of energy storage systems will cause greater losses to the equipment and reduce its service life, thus increasing the system operation cost during voltage regulation. The total cost of reactive power voltage regulation of energy storage equipment is:

[0136] ,

[0137] In the formula, is the loss degree of energy storage battery life under different depths of charge and discharge; is the total effective discharge amount of energy storage battery; is the initial investment cost of the energy storage battery; is the maintenance cost per unit power of the energy storage; is the active power loss cost generated per unit power;

[0138] Then for node the energy storage device of the cost of unit reactive power voltage regulation i.e., the voltage regulation cost of the energy storage device is:

[0139] ,

[0140] In the formula, is the reactive power output by the energy storage device;

[0141] 3) Voltage regulation cost of electric vehicles:

[0142] The voltage regulation cost of electric vehicles can generally be expressed as the economic cost required for each unit power regulation generated by the electric vehicles when participating in voltage regulation. The total voltage regulation cost of electric vehicles is:

[0143] ,

[0144] In the formula, is the charging unit cost of the electric vehicle; is the charging efficiency of the electric vehicle; is the charging power of the electric vehicle; is the charging duration of the electric vehicle at node ; ; is the energy loss of the electric vehicle in each charging cycle;

[0145] Then for node the electric vehicle at the unit active power voltage regulation cost i.e., the voltage regulation cost of the electric vehicle is:

[0146] ,

[0147] In the formula, is the active power regulated by the electric vehicle due to voltage regulation requirements;

[0148] 4) Voltage regulation cost of SVG device:

[0149] The SVG device issues a compensation drive signal through the control chip, and the power electronic inverter circuit issues a reactive power compensation current, so as to achieve the purpose of dynamic reactive power compensation and realize the dynamic regulation of voltage. The total cost generated when the SVG device participates in power grid voltage regulation is:

[0150] ,

[0151] wherein, is the initial investment cost of the SVG device; is the operating cost, including the monitoring, control, and operation costs during the operation of the SVG device; is the maintenance cost, referring to the costs of regular maintenance and replacement of parts of the device; is the energy consumption cost, referring to the power consumption cost generated by internal losses during the operation of the SVG device;

[0152] Then, for the SVG device at node to node the unit reactive power voltage regulation cost i.e., the voltage regulation cost of the SVG device is:

[0153] ,

[0154] wherein, is the reactive power output by the SVG device;

[0155] (3) Tracking response coefficient:

[0156] To evaluate the consistency between the output level of distributed resources and the load change, the load tracking coefficient is defined. During the scheduling period, the output of distributed resources is represented in per-unit value as:

[0157] ,

[0158] wherein, is the per-unit power value of the individual of distributed resources at time period ; is the power value of the individual at time period ; is the maximum power value of the individual , which can generally be taken as the installed capacity; is the number of time periods during the scheduling period;

[0159] The consistency index of the power signal and the load signal at time period is:

[0160] ,

[0161] wherein, is the difference between the per-unit power value of the load signal and the per-unit power value of the individual , representing the consistency between the two at time period ; For a time period Load signal Per-unit value of power;

[0162] By tracking the response coefficient To measure Individuals within a time period Power signal and load signal Consistency between:

[0163] ,

[0164] Wherein, The smaller the value, the more consistent the power signal of the individual Power signal and load signal The more consistent.

[0165] To further implement the above technical solution, the specific content of S2 includes:

[0166] Normalize all priority indicators. According to the normalized indicator values, construct radar charts with voltage sensitivity, voltage regulation cost, and response speed as vertices for different types of distributed resources;

[0167] Select two eigenvectors, the area and perimeter of each radar chart, to construct an evaluation function H , and calculate the evaluation values H Of each radar chart in turn;

[0168] According to the magnitudes of the respective evaluation values H Perform priority ranking.

[0169] To further implement the above technical solution, the specific content of normalization includes:

[0170] Normalize the values of each priority indicator To the interval [0, 1] to ensure comparability between indicators:

[0171] ,

[0172] Wherein, Represents the minimum value of the priority indicator; Represents the maximum value of the priority indicator; after indicator normalization, the closer its value is to 1, the better the indicator effect of the resource. Construct a comprehensive priority indicator radar chart for different resources based on multiple indicators.

[0173] Select two eigenvectors, the area and perimeter of the radar chart, to construct a priority evaluation system for different resources. The present invention calculates the function value using the average area and average perimeter. The area and perimeter of each radar chart are respectively:

[0174] ,

[0175] In the formula, and are the area and perimeter of the radar chart, is the side length formed by the th normalized index value; is the th included angle between the indexes;

[0176] Calculate the evaluation value of each radar chart H The specific content is:

[0177] .

[0178] To further implement the above technical solution, the specific content of S3 includes:

[0179] Aggregate total power The exact feasible region is complex and difficult to obtain or use due to the existence of a large number of distributed resources and power flow network relationships, so internal approximation is usually performed to effectively characterize the feasible region of the total power. The elliptical feasible region is an effective method to characterize the active-reactive flexibility region for each aggregation. Therefore, use the time-decoupled elliptical parametric feasible region

[0180] . .

[0181] Based on the current aggregation unit, characterize the feasible region of the aggregate total power , is the active power of the aggregation unit at time period , is the reactive power of the aggregation unit at time period , and obtain the elliptical feasible region , and adopt the time-decoupled elliptical feasible region :

[0182] ,

[0183] In the formula, is the center of the ellipse; is a two-dimensional positive semi-definite matrix, representing the rotation and stretching transformation of the ellipse; is the variable constrained by the uncertainty set U:

[0184] ,

[0185] In the formula, is the set of all time periods ;

[0186] In two-stage adaptive robust optimization aggregation, the feasible region is used as the uncertainty set, and is used as the uncertainty set for the uncertainty variables below. The distributed resource scheduling scheme is an adaptive variable that can be determined after the uncertainty variables are revealed. Compared with static variables, introducing adaptive variables can significantly enhance the optimality of the robust solution.

[0187] The approximate feasible region has two desirable properties: 1) Aggregate optimality: is the optimal inner approximation of the exact feasible region with the largest volume; 2) Disaggregation feasibility: Without violating the operating constraints, any one aggregate power value within can be achieved through distributed scheduling.

[0188] To achieve these two important characteristics, an active-reactive multi-dimensional dynamic aggregation model is established, aiming to maximize the total aggregation flexibility of the active-reactive power in different time periods, and the active-reactive multi-dimensional dynamic aggregation model is:

[0189] ,

[0190] ,

[0191] ,

[0192] ,

[0193] ,

[0194] wherein, the elliptical feasible region dynamically adjusts with the change of and the corresponding distributed resource power scheduling scheme is obtained; represents the determinant of the matrix and is equal to the area of the elliptical feasible region in the time period ; and both represent the power constraint parameters of the elliptical feasible region. The power constraint of the elliptical feasible region ensures the feasibility of decomposition, indicating that for any total power within , there must exist a corresponding distributed resource scheduling scheme to achieve it; and respectively represent the apparent power capacity constraint parameters of the photovoltaic device and the energy storage device; and respectively represent other power constraints and network constraint parameters of distributed resources;

[0195] Solve the active-reactive multi-dimensional dynamic aggregation model to obtain the optimal flexibility space of distributed resources, so as to maximize the elliptical feasible region.

[0196] It should be noted that:

[0197] By drawing and calculating the radar chart, the evaluation values of different distributed resources H are obtained, and then priority sorting is carried out. Resources with high priority can be the first to intervene for aggregation. Such priority sorting ensures that the power grid can select the most suitable resources for regulation under different operating conditions, thereby reducing the overall cost of system operation and improving the stability and flexibility of the power grid. The self-organizing aggregation strategy allows distributed resources to automatically adjust their participation degrees and scheduling behaviors according to preset rules and real-time feedback without central control, enabling the resources to maximize their effectiveness on the premise of ensuring system stability. In addition, this strategy also optimizes the utilization of resources and realizes the intelligent operation of the power grid by coordinating the cooperation among various resources and avoiding over-reliance on a single resource.

[0198] Power aggregation essentially projects the high-dimensional feasible region onto a low-dimensional aggregated power space where the apparent power capacity constraints of photovoltaic devices and energy storage devices, other power constraints of distributed resources, and network constraints constitute the high-dimensional feasible region , the elliptical feasible region power constraint describes the mapping relationship from to and the constraints of ensure that the projection of the feasible set U in the aggregated power space is accurate, and this relationship is reflected in the model through adaptive variables.

[0199] This model establishes two specific two-stage optimization aggregation models: one calculates the time-decoupled optimal feasible interval of aggregated active-reactive power, and the other solves the optimal elliptical feasible region of the active-reactive aggregation domain. The feasible region Z can be selected as any convex set that can be appropriately parameterized, such as a time-coupled ellipsoid or polyhedron.

[0200] Since the active-reactive multi-dimensional aggregation model involves a two-stage adaptive robust optimization problem, the present invention adopts the widely used column constraint generation algorithm to solve it. According to the differences in two-stage decisions, the original model is decomposed into a master problem and a sub-problem, and then a master-sub iteration process is carried out to obtain the optimal solution. The Gurobi optimizer can effectively solve the master problem and the sub-problem.

[0201] To further implement the above technical solution, the specific content of the upper-layer long-term hourly scheduling model in S4 includes:

[0202] The distribution network layer comprehensively considers the action times of on-load tap changers (OLTCs) and capacitor banks (CBs), constructs a coordinated voltage regulation action strategy for OLTCs and CBs with the goal of minimizing system network loss, and completes hourly optimization. The objective function in the long-term scale stage is:

[0203] ,

[0204] In the formula, and are respectively the resistance and current of branch in time period ; is the number of time periods in a scheduling cycle; is the duration of each time period;

[0205] The constraint conditions include: OLTC constraints and CB constraints; among them,

[0206] Since OLTC devices belong to discrete control devices, convex transformation of the constraints is required. The OLTC constraints are:

[0207] ,

[0208] ,

[0209] ,

[0210] ,

[0211] In the formula, is the voltage of node in time period ; is the on-load tap changer ratio in time period ; is the number of taps of the on-load tap changer; is the square of the reference voltage of the on-load tap changer at node in time period when the on-load tap changer is in tap position ; and are respectively the active power, reactive power and reactance of branch in time period ; is an arbitrary positive number; represents a binary variable indicating whether the on-load tap changer is in tap position in time period ; is the maximum number of transformer switching operations within a scheduling period;

[0212] CB is a reactive power compensation device that is switched in groups and also belongs to a discrete control device. The constraints for CB are:

[0213] ,

[0214] ,

[0215] ,

[0216] ,

[0217] In the formula, is the output value of the switched capacitor bank during time period ; is the number of switched capacitor banks during time period ; is the reactive power compensation of a single capacitor; and are the upper and lower limits of the output of the switched capacitor bank; is the specific installed number of capacitor banks; is the upper limit of the number of switched capacitor banks during time period ; is the maximum number of switching operations of the capacitor; is the sign function. When a is greater than 0, its value is 1. When a is equal to 0, its value is 0.

[0218] To further implement the above technical solution, the specific content of the lower-layer short-time scale minute-level control model in S4 includes:

[0219] After the upper-layer long-time scale hourly scheduling model determines the switching operations of the on-load tap changer and the capacitor bank, deterministic optimization of the active and reactive powers of various distributed resources at the substation area layer is performed using a preset scheduling interval;

[0220] Construct a short-time scale optimization objective function with the minimum system node voltage deviation:

[0221] ,

[0222] In the formula, is the number of nodes in the distribution network; represents the voltage of node during time period ; represents the rated voltage of the node during time period ;

[0223] It should be noted that:

[0224] Based on the predicted value one hour in advance, in this embodiment, ten minutes is used as a scheduling interval to deterministically optimize the active and reactive power of photovoltaic devices, energy storage devices, electric vehicles, and SVG devices at the substation area level.

[0225] The present invention will be further described through simulation experiments as follows:

[0226] The model and method of the present invention are used for emergency voltage regulation of distributed resources in a hierarchical interactive multi-dimensional dynamic aggregated distribution network. It is programmed and implemented in Python according to the Figure 1 process shown. The radar chart of the comprehensive priority index of distributed resources at a certain moment is as shown in Figure 3 . It can be seen from the simulation results that the priority order of different resources is SVG (H = 262), photovoltaic device (H = 246), energy storage device (H = 243), and electric vehicle (H = 236).

[0227] To highlight the advantages of the method proposed by the present invention, the following simulation comparison schemes are adopted: 1) Scheme 1 is a voltage regulation method that does not consider priority sorting; 2) Scheme 2 is a voltage regulation method that does not consider hierarchical aggregation nesting; 3) Scheme 3 is the method proposed by the present invention. The comparison results of network loss and voltage under different methods are shown in Table 1. It can be analyzed that compared with Scheme 1 and Scheme 2, the system network loss of the method proposed by the present invention is reduced by 7.03% and 3.98% respectively, and the voltage level is also significantly improved, effectively improving the power quality.

[0228] Table 1 Comparison results of network loss and voltage under different methods .

[0229] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for emergency voltage regulation of distributed resources in a distribution network based on hierarchical interactive multi-dimensional dynamic aggregation, characterized in that: The following steps are involved: S1. Obtain the values ​​of priority indicators of various types of distributed resources according to the voltage regulation cost and tracking response characteristics of distributed resources in the distribution network; S2. Sort various distributed resources according to the values ​​of the priority indicators of various distributed resources; S3. Carry out the lower-layer aggregation strategy; self-organize and aggregate various distributed resources according to the priority sorting results. The formed aggregation unit responds to the voltage regulation demand after receiving the scheduling command and continuously adjusts its distributed resource scheduling plan; The specific contents of S3 include: Characterize the total aggregate power based on the current aggregate unit The feasible domain of For the period Active power of the aggregation unit, For the period Aggregate the reactive power of the unit and obtain the elliptical feasible region , using the time-decoupled elliptic feasible region : ， In the formula, is the center of the ellipse; is a two-dimensional positive semidefinite matrix, representing the rotation and stretching transformation of the ellipse; is a variable constrained by the uncertainty set U: ， In the formula, For all time periods A collection of; The feasible domain As an uncertainty set, As uncertainty set Distributed resource scheduling scheme under uncertain variables It is an adaptive variable that changes dynamically with the uncertainty variable. An active-reactive multi-dimensional dynamic aggregation model is established. The goal is to maximize the total aggregation flexibility of active-reactive power in different time periods, and the active-reactive flexibility is optimally aggregated. The active-reactive multi-dimensional dynamic aggregation model is: ， ， ， ， ， In the formula, the elliptical feasible region changes as Dynamically adjust according to changes, and obtain the corresponding distributed resource power dispatch plan ; Representation Matrix The determinant is equal to the feasible domain of the ellipse in the period area; and All represent the power constraint parameters of the elliptical feasible region; and represent the apparent power capacity constraint parameters of photovoltaic equipment and energy storage equipment respectively; and denote other power constraints and network constraint parameters of distributed resources respectively; Solve the active-reactive multi-dimensional dynamic aggregation model to obtain the optimal distributed resource flexibility space to maximize the elliptical feasible region; S4. Carry out upper-level adjustment strategies; optimize and regulate the voltage of the entire power grid according to the real-time data adjusted by the lower-level aggregation strategy, including: long-time-scale scheduling of on-load voltage regulators and capacitor banks, and short-time-scale scheduling of aggregation units based on long-time-scale scheduling; S5. Feedback the real-time data of the overall power grid obtained after the upper-level adjustment strategy to S3, and carry out the next round of lower-level aggregation strategy to form a dynamic closed-loop nested control mechanism.

2. According to claim 1, a method for emergency voltage regulation of distributed resources in a distribution network based on hierarchical interactive multi-dimensional dynamic aggregation is characterized in that: The priority indicators of each type of distributed resources in S1 include voltage sensitivity, voltage regulation cost and tracking response coefficient; (1) Voltage sensitivity: ， In the formula, and Node For Node Active and reactive voltage sensitivity; is the line starting voltage; and Node With Node resistance and reactance between (2) The voltage regulation costs for each type include distributed photovoltaic equipment, energy storage equipment, electric vehicle, and SVG equipment; 1) Voltage regulation cost of distributed photovoltaic equipment: Total cost of photovoltaic discharge voltage regulation for: ， In the formula, is the depreciation cost of the photovoltaic equipment; is the unit power maintenance cost of the photovoltaic equipment; is the active power loss cost per unit power generated by the photovoltaic device; Then the node PV equipment for nodes Unit reactive voltage regulation cost That is, the voltage regulation cost of distributed photovoltaic equipment is: ， In the formula, For Node The voltage change before and after voltage regulation; Reactive power output by photovoltaic equipment; 2) Voltage regulation cost of energy storage equipment: Total cost of reactive voltage regulation power of energy storage equipment for: ， In the formula, The loss degree of energy storage battery life under different depths of charge and discharge; is the total effective discharge capacity of the energy storage battery; The initial investment cost of the energy storage battery; The maintenance cost per unit power of energy storage; is the active power loss cost per unit power; Then the node Energy storage devices for nodes Unit reactive voltage regulation cost That is, the voltage regulation cost of energy storage equipment is: ， In the formula, Reactive power output by the energy storage device; 3) Electric vehicle voltage regulation cost: Total cost of voltage regulation for electric vehicles for: ， In the formula, the unit cost of charging an electric vehicle; Charging efficiency for electric vehicles; Charging power for electric vehicles; For electric vehicles at the node Charging time ; The energy loss in each charging cycle of electric vehicles; Then the node Electric vehicles at the node Unit active voltage regulation cost That is, the voltage regulation cost of electric vehicles is: ， In the formula, It is the active power adjusted by the electric vehicle due to voltage regulation requirements; 4) SVG equipment voltage regulation cost: Total cost of SVG equipment participating in grid voltage regulation for: ， In the formula, is the initial investment cost of the SVG equipment; For operating costs; For maintenance costs; For energy consumption cost; Then the node SVG device pair node Unit reactive voltage regulation cost That is, the voltage regulation cost of SVG equipment is: ， In the formula, Reactive power output by SVG equipment; (3) Tracking response coefficient: During the scheduling period, the output of distributed resources is expressed in per-unit form as: ， In the formula, Individuals of distributed resources In the period The power per unit value; For individuals In the period The power value; For individuals The maximum power value; is the number of time periods in the scheduling period; Power signal and load signal exist Consistency index for time periods for: ， In the formula, For the period Load signal The power per unit value; By tracking the response coefficient To measure Individuals in a period Power signal and load signal Consistency between: ， In the formula, The smaller the individual Power signal and load signal The more consistent.

3. The method for emergency voltage regulation of distributed resources in distribution network based on hierarchical interactive multi-dimensional dynamic aggregation according to claim 1 is characterized in that: The specific contents of S2 include: All priority indicators are normalized, and according to the normalized indicator values, radar charts are constructed for different types of distributed resources with voltage sensitivity, voltage regulation cost, and response speed as vertices. Select the area and perimeter of each radar chart as two feature vectors to construct the evaluation function H , and calculate the evaluation value of each radar chart in turn H ; According to each evaluation value H The size of the .

4. The method for emergency voltage regulation of distributed resources in a distribution network based on hierarchical interactive multi-dimensional dynamic aggregation according to claim 3 is characterized in that: The specific contents of normalization include: The value of each priority indicator Normalized to the interval [0,1]: ， In the formula, Indicates the minimum value of the priority indicator; Indicates the maximum value of the priority indicator; The area and perimeter of each radar chart are: ， In the formula, and is the area and perimeter of the radar chart, For the The length of the side formed by the normalized index values; For the The angle between the indicators; Calculate the evaluation value of each radar chart in turn H The specific content is: 。 5. The method for emergency voltage regulation of distributed resources in distribution network based on hierarchical interactive multi-dimensional dynamic aggregation according to claim 1 is characterized in that: The specific contents of the long-term hourly scheduling model in the upper layer of S4 include: The distribution network layer comprehensively considers the operation times of the on-load voltage regulator OLTC and the capacitor bank CB, constructs the OLTC and CB coordinated voltage regulation action strategy with the goal of minimizing the system network loss, and completes the hourly optimization. The objective function is: ， In the formula, and The time periods Branch Road resistance and current; is the number of time periods in a scheduling cycle; is the duration of each time period; The constraints include: OLTC constraints and CB constraints; among them, The OLTC constraints are: ， ， ， ， In the formula, For Node In the period Voltage; For the period The on-load voltage regulator ratio; It is the number of gears of the on-load voltage regulator; For Node In the period The on-load voltage regulator is in gear The square of the reference voltage at ; and The time periods Branch Road Active power, reactive power and reactance; is any positive number; Indicates time period Is the on-load voltage regulator Binary variable for gear position; The maximum number of transformer switching times in a dispatching cycle; The CB constraints are: ， ， ， ， In the formula, The capacitor bank switched in the time period Output value; For the period The number of groups of switched capacitors; is the reactive power compensation of a single capacitor; and It is the upper and lower limits of the output of the switched capacitor bank; The specific number of capacitor groups to be installed; For the period The upper limit of the number of capacitor switching groups; is the maximum switching times of the capacitor; is a sign function. When a is greater than 0, its value is 1, and when a is equal to 0, its value is 0.

6. The method for emergency voltage regulation of distributed resources in a distribution network based on hierarchical interactive multi-dimensional dynamic aggregation according to claim 1, characterized in that: The specific contents of the short-time scale minute-level control model in the lower layer of S4 include: After the upper-layer long-time-scale hourly dispatch model determines the switching action of the on-load voltage regulator and capacitor bank, the preset dispatch interval is used to perform deterministic optimization on the active and reactive power of various distributed resources at the substation level; Construct a short-time optimization objective function with the minimum system node voltage deviation: ， In the formula, is the number of distribution network nodes; Representative Node In the period Voltage; Representative node in the period Rated voltage.

Images