Power distribution network dynamic partitioning method and system considering network topology and resource time-varying characteristics

Abstract

The invention belongs to the technical field of power systems, and relates to a power distribution network dynamic partitioning method and system considering network topology and resource time-varying characteristics. The method comprises the following steps: establishing a power distribution network dynamic partition evaluation index system; constructing a power distribution network dynamic partition multi-objective optimization model and setting constraint conditions of power distribution network dynamic partition; by setting a threshold value of a comprehensive electrical distance difference value and a threshold value of a source-load correlation coefficient difference value, dynamic adjustment of the power distribution network partition structure is triggered; and optimizing and solving to obtain the optimal solution of the power distribution network partition. According to the invention, the strategy of triggering the updating of the partition structure based on the threshold value is adopted, so that the power distribution network can make a quick response according to the source-load correlation and the real-time change of the topological structure; according to the method, the partition structure is adjusted in time through a dynamic updating mechanism, dynamic characteristics and topological structure changes of distributed resources and loads are better adapted, power flow distribution is optimized, network loss is reduced, power supply quality is improved, and reliability and stability of the system are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of power systems, and more specifically, relates to a dynamic partitioning method and system for a distribution network considering network topology and resource time-variation. Background Art

[0002] With the in-depth promotion of the energy transition, the penetration rate of distributed renewable energy sources (such as photovoltaic and wind power) in the distribution network is continuously increasing. However, the output of these resources has significant spatio-temporal volatility, and combined with the dynamic changes in the network topology, it makes it difficult for traditional static partitioning methods to meet the actual operation requirements. Photovoltaic power generation depends on light intensity and duration, while wind power is affected by wind speed and direction. This uncertainty leads to drastic changes in its power output at different time and space scales.

[0003] Traditional static partitioning methods for distribution networks plan partitions based on fixed network structures and load prediction data, and problems become prominent when facing large-scale access of distributed energy sources and frequent changes in network topology. In terms of operation efficiency, it cannot adapt to changes in energy output and topology, easily resulting in unreasonable power distribution and increased line losses; in terms of reliability, it is difficult to cope with emergencies and affects power supply stability. For example, in a certain urban distribution network during the summer peak electricity consumption, the distributed photovoltaic output is large but there is still a power supply and demand imbalance in some areas. When the network topology changes, the traditional static partitioning method cannot be adjusted in time, resulting in power outages for some users.

[0004] Therefore, there is an urgent need for a dynamic partitioning method for distribution networks that considers network topology and resource time-variation, aiming to achieve a balance between refined partitioning and efficient operation, improve the operation efficiency and reliability of the distribution network, and promote the development of the power grid towards a more green, intelligent, and reliable direction. Summary of the Invention

[0005] To solve the above technical problems, the present invention provides a dynamic partitioning method and system for a distribution network considering network topology and resource time-variation.

[0006] In a first aspect, the present invention provides a dynamic partitioning method for a distribution network considering network topology and resource time-variation, including: Establish a dynamic partitioning evaluation index system for the distribution network, including a comprehensive electrical distance index, a voltage regulation ability index, and a source-load spatio-temporal correlation index; the comprehensive electrical distance index is defined according to voltage sensitivity information and is used to describe the tightness of the electrical connection between any two nodes in the system; the voltage regulation ability index is used to characterize the regulation ability of the adjustable active and reactive power of distributed resources in the region on the maximum voltage deviation; the source-load spatio-temporal correlation index is used to characterize the correlation between the output of distributed resources and the load demand; According to the dynamic partition evaluation index system of the distribution network, weights are assigned to the comprehensive electrical distance index, voltage regulation ability index, and source-load spatio-temporal correlation index, and the comprehensive evaluation index is used as the output to construct a multi-objective optimization model for dynamic partition of the distribution network and set the constraint conditions for dynamic partition of the distribution network; By setting the threshold of the difference in comprehensive electrical distance and the threshold of the difference in the source-load correlation coefficient between the previous and current moments, the dynamic adjustment of the distribution network partition structure is triggered; Taking the comprehensive evaluation index as the objective function, a partition optimization algorithm is used to optimize and solve the multi-objective optimization model for dynamic partition of the distribution network to obtain the optimal solution for the distribution network partition.

[0007] On the second aspect, the present invention provides a distribution network dynamic partition system considering network topology and resource time-variation, including an evaluation index system establishment unit, a multi-objective optimization model construction unit, a setting unit, and a solution unit; The evaluation index system establishment unit is used to establish a distribution network dynamic partition evaluation index system, including a comprehensive electrical distance index, a voltage regulation ability index, and a source-load spatio-temporal correlation index; the comprehensive electrical distance index is defined according to voltage sensitivity information and is used to describe the tightness of the electrical connection between any two nodes in the system; the voltage regulation ability index is used to characterize the regulation ability of the adjustable active and reactive power of distributed resources in the region on the maximum voltage deviation; the source-load spatio-temporal correlation index is used to characterize the correlation between the output of distributed resources and the load demand; The multi-objective optimization model construction unit is used to assign weights to the comprehensive electrical distance index, voltage regulation ability index, and source-load spatio-temporal correlation index according to the distribution network dynamic partition evaluation index system, use the comprehensive evaluation index as the output, construct a multi-objective optimization model for dynamic partition of the distribution network, and set the constraint conditions for dynamic partition of the distribution network; The setting unit is used to trigger the dynamic adjustment of the distribution network partition structure by setting the threshold of the difference in comprehensive electrical distance and the threshold of the difference in the source-load correlation coefficient between the previous and current moments; The solution unit is used to take the comprehensive evaluation index as the objective function, use a partition optimization algorithm to optimize and solve the multi-objective optimization model for dynamic partition of the distribution network, and obtain the optimal solution for the distribution network partition.

[0008] Based on the above technical solutions, the present invention can also be improved as follows.

[0009] Further, the calculation of voltage sensitivity information includes: assuming the voltage amplitude is , the change in voltage amplitude is , the phase angle is , the change in phase angle is , the active power is , the change in active power is , the reactive power is , the change in reactive power is , the phase angle sensitivity matrix related to active power is , the phase angle sensitivity matrix related to reactive power is , the voltage magnitude sensitivity matrix related to active power is , the voltage magnitude sensitivity matrix related to reactive power is , then: ; The calculation formula for voltage sensitivity information is: .

[0010] Furthermore, the calculation of the comprehensive electrical distance includes: Let be the voltage change of the th node under the influence of reactive power, be the voltage change of the th node under the influence of reactive power, be the reactive power variable of the th node, be the reactive voltage change sensitivity of the th node to the th node, be the voltage change of the th node under the influence of reactive power, be the voltage change of the th node under the influence of reactive power, be the reactive power variable of the th node, be the active voltage change sensitivity of the th node to the th node, be the sensitivity of the reactive power change of node to the voltage magnitude of node , be the sensitivity of the reactive power change of node to its own voltage magnitude; Under the influence of reactive power, the voltage change relationship between the th node and the th node defined by voltage sensitivity is expressed as: ; The active voltage change sensitivity of the th node to the th node is: ; Let the comprehensive electrical distance be , at time Cluster The comprehensive electrical distance of ; According to the reactive voltage sensitivity and the active voltage sensitivity , construct the expression of the comprehensive electrical distance between the th node and the th node: ; .

[0011] Furthermore, the calculation of the voltage regulation ability index includes: The voltage regulation ability of area at is , then: Let be the voltage deviation of the th node at moment, and be the maximum voltage deviation of the th node in area at moment; ; represents the voltage regulation ability of area at moment, then: ; represents the active voltage sensitivity matrix, represents the reactive voltage sensitivity matrix, is the regulation ability of the adjustable active power of node at moment on the maximum voltage deviation, is the regulation ability of the adjustable reactive power of node at moment on the maximum voltage deviation, is the adjustable active power of distributed photovoltaic of node at when the sum of the edge weights of all nodes in the network is moment, is the adjustable active power of energy storage of node at when the sum of the edge weights of all nodes in the network is moment, is the adjustable active power of node at when the sum of the edge weights of all nodes in the network is ​​​​​​​​​​​​​​​The adjustable active power of the flexible load at a certain moment, For the sum of the edge weights of all nodes in the network to be At this time, node At The adjustable reactive power of the distributed photovoltaic at a certain moment, For the sum of the edge weights of all nodes in the network to be At this time, node At The adjustable reactive power of the energy storage at a certain moment, For the sum of the edge weights of all nodes in the network to be At this time, node At The adjustable reactive power of the flexible load at a certain moment, For the sum of the edge weights of all nodes in the network to be At this time, node At The adjustable reactive power of the reactive power compensation device at a certain moment, then: ; ; Let the voltage regulation ability of the distribution network be , The number of partitions, and the voltage regulation ability of the distribution network is the average of the regulation abilities of each region, expressed as: .

[0012] Furthermore, the calculation of the source-load correlation index includes: Let Be The source-load correlation index at a certain moment, Be the photovoltaic output power, Be the load magnitude, Be The covariance of the photovoltaic output power and the load at a certain moment, Be The standard deviation of the photovoltaic output power at a certain moment, Be The standard deviation of the load at a certain moment, then The source-load correlation index at a certain moment is: .

[0013] Furthermore, the calculation of the comprehensive evaluation index includes: Let Be the comprehensive evaluation index, Be the voltage regulation ability of region At A certain moment, Be The comprehensive electrical distance index at a certain moment, Be The weight coefficient of the voltage regulation ability of the region at at the time, is ; The weight coefficient of the comprehensive electrical distance index at is ; The weight coefficient of the source-load correlation index at time, and .

[0014] Furthermore, let be the number of nodes of the adjustable resources within the cluster , be the number of nodes of the loads within the cluster . The constraint conditions include that a single node cannot form a partition and each partition contains at least one adjustable resource node. Then the constraint conditions are expressed as: ; .

[0015] Furthermore, by setting the threshold of the difference in the comprehensive electrical distance, the dynamic adjustment of the distribution network partition structure is triggered, including: by comparing the electrical distance index value at the current time after the topological change with the electrical distance index value at the previous time before the topological change. When the difference between the two exceeds the first threshold, the distribution network partition structure is changed; Let the electrical distance index value at the current time after the topological change be , the electrical distance index value at the previous time before the topological change be , and the first threshold be . Then the dynamic adjustment method is expressed as: .

[0016] Furthermore, by setting the threshold of the difference in the source-load correlation coefficient between the previous and current times, the dynamic adjustment of the distribution network partition structure is triggered, including: by comparing the source-load correlation index value at the current time with the source-load correlation index value at the previous time , calculating the difference between the two. When the difference exceeds the preset second threshold, the source-load relationship changes, triggering the dynamic adjustment of the distribution network partition structure; Let the source-load correlation index value at the current time be , the source-load correlation index value at the previous time be , and the second threshold be , the dynamic adjustment method is expressed as: .

[0017] The beneficial effects of the present invention are as follows: (1) The present invention establishes a dynamic partition evaluation index system for a distribution network considering the time-varying nature of network topology and resources, and adopts a strategy based on threshold-triggered partition structure update, enabling the distribution network to respond quickly to real-time changes in source-load correlation and topological structure; (2) The present invention timely adjusts the partition structure through a dynamic update mechanism, constructs a multi-objective optimization model for dynamic partitioning of a distribution network considering the time-varying nature of network topology and resources. The distribution network can better adapt to the dynamic characteristics of distributed resources and loads and topological structure changes, optimize the power flow distribution, reduce network losses, and improve power supply quality; (3) The dynamic partition method adopted by the present invention can enhance the ability of the distribution network to cope with faults and abnormal situations, quickly respond to network topology and resource changes, dynamically update the partition results, improve the reliability and stability of the system, and provide a strong guarantee for the safe and efficient operation of the distribution network. Description of the Drawings

[0018] Figure 1 is the schematic diagram of the dynamic partition method for a distribution network considering the time-varying nature of network topology and resources provided in Embodiment 1 of the present invention; Figure 2 is the simulation diagram of the normalized curves of typical daily photovoltaic power output and load; Figure 3 is the specific dynamic partition flow chart; Figure 4 is the principle block diagram of the dynamic partition system for a distribution network considering the time-varying nature of network topology and resources provided in Embodiment 2 of the present invention. Detailed Embodiments

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and illustrated herein can be arranged and designed in various different configurations.

[0020] Embodiment 1 As an embodiment, as shown in the appendix Figure 1 , to solve the above technical problems, this embodiment provides a dynamic partition method for a distribution network considering the time-varying nature of network topology and resources, including: Establish a dynamic partition evaluation index system for the distribution network, including the comprehensive electrical distance index, voltage regulation ability index, and source-load spatio-temporal correlation index; the comprehensive electrical distance index is defined based on voltage sensitivity information and is used to describe the tightness of the electrical connection between any two nodes in the system; the voltage regulation ability index is used to characterize the regulation ability of the adjustable active and reactive powers of distributed resources in the region on the maximum voltage deviation; the source-load spatio-temporal correlation index is used to characterize the correlation between the output of distributed resources and the load demand; According to the dynamic partition evaluation index system of the distribution network, weight the comprehensive electrical distance index, voltage regulation ability index, and source-load spatio-temporal correlation index, take the comprehensive evaluation index as the output, construct a multi-objective optimization model for the dynamic partition of the distribution network, and set the constraint conditions for the dynamic partition of the distribution network; Trigger the dynamic adjustment of the distribution network partition structure by setting the threshold of the comprehensive electrical distance difference and by setting the threshold of the difference in the source-load correlation coefficient between the previous and current moments; Take the comprehensive evaluation index as the objective function, and use the partition optimization algorithm to optimize and solve the multi-objective optimization model for the dynamic partition of the distribution network to obtain the optimal solution of the distribution network partition.

[0021] Optionally, the calculation of voltage sensitivity information includes: Let the voltage amplitude be , the change in voltage amplitude be , the phase angle be , the change in phase angle be , the active power be , the change in active power be , the reactive power be , the change in reactive power be , the phase angle sensitivity matrix related to active power be , the phase angle sensitivity matrix related to reactive power be , the voltage amplitude sensitivity matrix related to active power be , the voltage amplitude sensitivity matrix related to reactive power be , then: ; The calculation formula for voltage sensitivity information is: .

[0022] The change in the node voltage of the distribution network is related to both active and reactive powers. , , and and have the same weight.

[0023] Optionally, the calculation of the comprehensive electrical distance includes: Let be the The voltage change of the node under the influence of reactive power, is the reactive power variable of the node, is the sensitivity of the reactive power - voltage change of the node to the node; is the reactive power variable of the node, is the sensitivity of the active power - voltage change of the node to the node; is the sensitivity of the reactive power change of the node to the voltage amplitude of the node itself; Under the influence of reactive power, the voltage change relationship between the node and the node defined by voltage sensitivity is expressed as: Let the comprehensive electrical distance be at time ; According to the reactive power - voltage sensitivity and the active power - voltage sensitivity node and the node, construct the expression of the comprehensive electrical distance between them:

[0024] Optionally, the calculation of the voltage regulation ability index includes: The voltage regulation ability of area at time is Let​​​​​​​​​​​​​​​​​​ is the voltage deviation of the th node at moment, is the maximum voltage deviation of the th node in the area at moment; ; represents the voltage regulation ability of the area at moment, then: ; represents the active voltage sensitivity matrix, represents the reactive voltage sensitivity matrix, is the regulation ability of the adjustable active power of the th node at moment on the maximum voltage deviation, is the regulation ability of the adjustable reactive power of the th node at moment on the maximum voltage deviation, is the adjustable active power of the distributed PV of the th node at moment when the sum of the edge weights of all nodes in the network is , is the adjustable active power of the energy storage of the th node at moment when the sum of the edge weights of all nodes in the network is , is the adjustable active power of the flexible load of the th node at moment when the sum of the edge weights of all nodes in the network is , is the adjustable reactive power of the distributed PV of the th node at moment when the sum of the edge weights of all nodes in the network is , is the adjustable reactive power of the energy storage of the th node at moment when the sum of the edge weights of all nodes in the network is , is the adjustable reactive power of the flexible load of the th node at moment when the sum of the edge weights of all nodes in the network is , is the adjustable active power of the th node at moment when the sum of the edge weights of all nodes in the network is For the adjustable reactive power of the instantaneous reactive power compensation device, then: ; ; Let the voltage regulation ability of the distribution network be , be the number of partitions. The voltage regulation ability of the distribution network is the average of the regulation abilities of each area, expressed as: .

[0025] Each area should adjust the over-limit voltage by the active / reactive power of the internal adjustable resources as much as possible to reduce the power loss between areas.

[0026] Optionally, the calculation of the source-load correlation index includes: Let be the source-load correlation index at time be the photovoltaic output power, be the load size, be the covariance of the photovoltaic output power and the load at time be the standard deviation of the photovoltaic output power at time be the standard deviation of the load at time then the source-load correlation index at time .

[0027] The output of distributed resources has a strong correlation with the load demand, but its volatility is large and its predictability is poor, which brings challenges to the stable operation of the power system when a large amount of new energy is connected to the grid. Therefore, the correlation between new energies is considered when partitioning the flexible distribution network. The correlation coefficient ranges from -1 to 1. The closer the value is to 1, the stronger the positive correlation between the two. The closer the value is to -1, the stronger the negative correlation. 0 indicates no correlation.

[0028] Optionally, the calculation of the comprehensive evaluation index includes: Let be the comprehensive evaluation index, be the voltage regulation ability of area at time, be the comprehensive electrical distance index at time be the source-load correlation index at time be the weight coefficient of the voltage regulation ability of area at time, be The weight coefficient of the comprehensive electrical distance index at a moment is the weight coefficient of the source-load correlation index at a moment, and , then: .

[0029] Optionally, let be the number of nodes of adjustable resources inside the cluster be the number of nodes that are loads inside the cluster. The constraint conditions include that a single node cannot form a partition and each partition contains at least one adjustable resource node. Then the constraint conditions are expressed as: ; .

[0030] A single node cannot form a partition. Each partition contains at least one adjustable resource node, which can ensure the voltage regulation ability and power balance inside the partition.

[0031] In the actual application process, the all-day output of distributed resources and the network dynamic topology fluctuate with time, causing changes in the electrical characteristics, voltage regulation ability, and source-load correlation of each partition of the distribution network. The fixed static partition scheme cannot adapt to the dynamic operating conditions of the distribution network. Therefore, combined with the power flow prediction data, find the optimal partition scheme for each time period and adjust the partition in real-time dynamically. Optionally, the time scale is counted in 5 minutes; in actual operation, it can be adjusted according to three typical time periods: morning peak, midday peak load, and evening peak.

[0032] Voltage sensitivity characterizes the sensitive response degree of the node voltage to the change of injected power (such as reactive power or active power). This characteristic can usually be reflected by the Jacobian matrix in the power flow calculation.

[0033] The electrical distance is indirectly defined through voltage sensitivity. When the topological structure of the distribution network changes, the voltage sensitivity of the distribution network will change accordingly with the changes of active power and reactive power. Since the electrical distance is closely related to voltage sensitivity, when the voltage sensitivity changes, the electrical distance will also change accordingly; and the change of the electrical distance will further prompt the update of the partition result, so that the partition result can better adapt to the current control strategy.

[0034] Optionally, by setting the threshold of the comprehensive electrical distance difference, trigger the dynamic adjustment of the distribution network partition structure, including: by comparing the electrical distance index value at the current moment after the topological change with the electrical distance index value at the previous moment The difference between the electrical distance index values at a certain time. When the difference exceeds the first threshold, it triggers a change in the distribution network partition structure; assume that at the current moment after the topological change the electrical distance index value is , at the previous moment before the topological change the electrical distance index value is , the first threshold is , then the dynamic adjustment method is expressed as: .

[0035] By setting the threshold of the comprehensive electrical distance difference, triggering the dynamic adjustment of the distribution network partition structure can effectively improve the operation performance, adaptability and reliability of the distribution network.

[0036] Distributed resources exhibit the characteristics of uncertainty and randomness, and their power output is comprehensively affected by various factors such as time and weather. Taking photovoltaic power generation as an example, during the day, the photovoltaic panels can fully receive sunlight and thus generate more electrical energy; while at night, due to the lack of sunlight, the photovoltaic power generation is almost zero. At the same time, there is also a certain correlation in the time dimension of the power load. Within a specific time period, the correlation between the power load and the power source will change dynamically. The normalized curves of photovoltaic output and load on a typical day are as Figure 2 shown. The abscissa is time, unit: h, and the ordinate is the photovoltaic normalized value and the load normalized value. L1 represents the photovoltaic normalized curve, and L2 represents the load normalized curve.

[0037] As time goes by, the mutual relationship and influence between distributed resources and loads will also change. In order to accurately describe the dynamic impact of the time-varying characteristics of in-day resources on the distribution network partition structure, the present invention introduces a source-load correlation index for quantitative analysis. Optionally, the partition structure of the distribution network is updated every 15 minutes to ensure that the partition results can reflect the real-time changes of the source-load relationship in a timely and accurate manner.

[0038] Optionally, by setting the threshold of the difference between the source-load correlation coefficients at the previous and current moments, triggering the dynamic adjustment of the distribution network partition structure includes: by comparing the source-load correlation index value at the current moment with the source-load correlation index value at the previous moment , calculating the difference between the two. When the difference exceeds the preset second threshold, the source-load relationship changes, triggering the dynamic adjustment of the distribution network partition structure; assume that the source-load correlation index value at the current moment is , the source-load correlation index value at the previous moment is , the second threshold is , then the dynamic adjustment method is expressed as: 。

[0039] As shown in the specific dynamic partition flow chart attached Figure 3 The specific process includes: inputting the parameters of the distribution network and the information of distributed resources, calculating the electrical distance and the source-load correlation coefficient of the distribution network at the current moment; calculating the difference in the electrical distance index before and after the topological change; if the difference in the electrical distance index before and after the topological change is greater than the first threshold, then calculating the difference in the source-load correlation index value, otherwise retaining the original partition structure; if the difference in the source-load correlation index value before and after the topological change is greater than the second threshold, then taking the comprehensive evaluation index constructed based on the comprehensive electrical distance index, voltage regulation ability index, and source-load spatio-temporal correlation index within the partition as the objective function, and using the partition optimization algorithm to perform optimization solution on the multi-objective optimization model of the distribution network dynamic partition to obtain the optimal solution of the distribution network partition, outputting the distribution network partition result, otherwise retaining the original partition structure and outputting the distribution network partition result.

[0040] The present invention adopts a threshold-triggered partition structure update strategy, which enables the distribution network to make a rapid response according to the real-time changes of the source-load correlation and topological structure; through the dynamic update mechanism, the partition structure can be adjusted in a timely manner, so that the distribution network can better adapt to the dynamic characteristics of distributed resources and loads and topological structure changes, optimize the power flow distribution, reduce network losses, and improve the power supply quality; at the same time, the dynamic partition can enhance the ability of the distribution network to cope with faults and abnormal situations, and improve the reliability and stability of the system. Generally speaking, the present invention helps to effectively improve the operation performance, adaptability, and reliability of the distribution network, and provides a strong guarantee for the safe and efficient operation of the distribution network.

[0041] Embodiment 2 Based on the same principle as the method shown in Embodiment 1 of the present invention, as shown in the attachment Figure 4 The embodiment of the present invention also provides a distribution network dynamic partition system considering the time-varying nature of network topology and resources, including an evaluation index system establishment unit, a multi-objective optimization model construction unit, a setting unit, and a solution unit; The evaluation index system establishment unit is used to establish an evaluation index system for the distribution network dynamic partition, including a comprehensive electrical distance index, a voltage regulation ability index, and a source-load spatio-temporal correlation index; the comprehensive electrical distance index is defined according to the voltage sensitivity information and is used to describe the tightness of the electrical connection between any two nodes in the system; the voltage regulation ability index is used to characterize the regulation ability of the adjustable active and reactive power of distributed resources in the region on the maximum voltage deviation; the source-load spatio-temporal correlation index is used to characterize the correlation between the output of distributed resources and the load demand; The multi-objective optimization model construction unit is used to assign weights to the comprehensive electrical distance index, voltage regulation ability index, and source-load spatio-temporal correlation index according to the distribution network dynamic partition evaluation index system, take the comprehensive evaluation index as the output, construct a multi-objective optimization model for the distribution network dynamic partition, and set the constraint conditions for the distribution network dynamic partition; The setting unit is used to trigger the dynamic adjustment of the distribution network partition structure by setting the threshold of the comprehensive electrical distance difference and by setting the threshold of the difference in the source-load correlation coefficient between the front and back moments; The solving unit is used to take the comprehensive evaluation index as the objective function, and use the partition optimization algorithm to perform optimization solution on the multi-objective optimization model of the distribution network dynamic partition to obtain the optimal solution of the distribution network partition.

[0042] Optionally, the calculation of voltage sensitivity information includes: Let the voltage amplitude be , the change in voltage amplitude be , the phase angle be , the change in phase angle be , the active power be , the change in active power be , the reactive power be , the change in reactive power be , the phase angle sensitivity matrix related to active power be , the phase angle sensitivity matrix related to reactive power be , the voltage amplitude sensitivity matrix related to active power be , the voltage amplitude sensitivity matrix related to reactive power be , then: ; The calculation formula of voltage sensitivity information is: .

[0043] Optionally, the calculation of the comprehensive electrical distance includes: Let be the voltage change amount of the th node under the influence of reactive power, be the voltage change amount of the th node under the influence of reactive power, be the reactive power variable of the th node, be the reactive power voltage change sensitivity of the th node to the th node, be the voltage change amount of the th node under the influence of reactive power, be the voltage change amount of the th node under the influence of reactive power, be the The reactive power variable of the th node to the th node's active voltage change sensitivity is The sensitivity of the reactive power change of node to the voltage amplitude of node is The sensitivity of the reactive power change of node to its own voltage amplitude under reactive power influence; Under reactive power influence, the voltage change relationship between the th node and the th node defined by voltage sensitivity is expressed as: The active voltage change sensitivity of the th node to the th cluster is ; According to the reactive voltage sensitivity and the active voltage sensitivity , construct the expression of the comprehensive electrical distance between the th node and the th node;

[0044] Optionally, the calculation of the voltage regulation ability index includes: The voltage regulation ability of area at time is , then: Let be the voltage deviation of the th node at time, is the maximum voltage deviation of the th node in area ; represents the voltage regulation ability of area at time, then: ; represents the active voltage sensitivity matrix, represents the reactive voltage sensitivity matrix, is the node at the regulation ability of the adjustable active power of the node at time is the node at the regulation ability of the adjustable reactive power of the node at time is the sum of the edge weights of all nodes in the network is when the node at the adjustable active power of distributed photovoltaic at time is the sum of the edge weights of all nodes in the network is when the node at the adjustable active power of energy storage at time is the sum of the edge weights of all nodes in the network is when the node at the adjustable active power of flexible load at time is the sum of the edge weights of all nodes in the network is when the node at the adjustable reactive power of distributed photovoltaic at time is the sum of the edge weights of all nodes in the network is when the node at the adjustable reactive power of energy storage at time is the sum of the edge weights of all nodes in the network is when the node at the adjustable reactive power of flexible load at time is the sum of the edge weights of all nodes in the network is when the node at the adjustable reactive power of reactive power compensation equipment at time ; ; Let the voltage regulation ability of the distribution network be , is the number of partitions, and the voltage regulation ability of the distribution network is the average value of the regulation abilities of each region, expressed as: .

[0045] Optionally, the calculation of the source-load correlation index includes: Let be Moment source-load correlation index is the PV output power is the load magnitude is the covariance of the PV output power and the load at moment is the standard deviation of the PV output power at moment is the standard deviation of the load at moment, then the moment source-load correlation index is .

[0046] Optionally, the calculation of the comprehensive evaluation index includes: Let be the comprehensive evaluation index be the voltage regulation ability of area at moment is the comprehensive electrical distance index at moment is the moment source-load correlation index is the weight coefficient of the voltage regulation ability of area at moment is the weight coefficient of the comprehensive electrical distance index at moment is the weight coefficient of the moment source-load correlation index, and , then .

[0047] Optionally, let be the number of nodes of adjustable resources within cluster be the number of nodes with load within cluster ; .

[0048] Optionally, by setting the threshold of the comprehensive electrical distance difference, the dynamic adjustment of the distribution network partition structure is triggered, including: by comparing the electrical distance index value at the current moment after the topology change with the electrical distance index value at the previous moment before the topology change, when the difference between the two exceeds the first threshold, the distribution network partition structure is changed; let the current moment ​​The electrical distance index value under is ; the electrical distance index value at the previous moment before the topological change is ; the first threshold is . Then the dynamic adjustment method is expressed as: .

[0049] Optionally, by setting the threshold of the difference in the source-load correlation coefficient between the previous and current moments, the dynamic adjustment of the distribution network partition structure is triggered, including: by comparing the source-load correlation index value at the current moment with the source-load correlation index value at the previous moment , calculating the difference between the two. When the difference exceeds the preset second threshold, the source-load relationship changes, triggering the dynamic adjustment of the distribution network partition structure; let the source-load correlation index value at the current moment be , the source-load correlation index value at the previous moment be , and the second threshold be . Then the dynamic adjustment method is expressed as: .

[0050] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A distribution network dynamic partitioning method considering network topology and resource time-varying characteristics, characterized in that: include: Establish a dynamic zoning evaluation index system for distribution networks, including comprehensive electrical distance index, voltage regulation capability index and source-load spatiotemporal correlation index; the comprehensive electrical distance index is defined based on voltage sensitivity information and is used to describe the closeness of the electrical connection between any two nodes in the system; the voltage regulation capability index is used to characterize the ability of the adjustable active and reactive power of distributed resources in the region to regulate the maximum voltage deviation; the source-load spatiotemporal correlation index is used to characterize the correlation between the output of distributed resources and load demand; According to the evaluation index system of dynamic partition of distribution network, weights are assigned to comprehensive electrical distance index, voltage regulation capability index and source-load time-space correlation index, and the comprehensive evaluation index is used as output to construct a multi-objective optimization model of dynamic partition of distribution network and set the constraint conditions of dynamic partition of distribution network. By setting the threshold of the comprehensive electrical distance difference and the threshold of the source-load correlation coefficient difference before and after, the dynamic adjustment of the distribution network partition structure is triggered; Taking the comprehensive evaluation index as the objective function, the partition optimization algorithm is used to solve the multi-objective optimization model of dynamic partitioning of distribution network and obtain the optimal solution of distribution network partitioning.

2. The method for dynamic partitioning of a distribution network considering network topology and time-varying resources according to claim 1 is characterized in that: The calculation of voltage sensitivity information includes: assuming that the voltage amplitude is , the change in voltage amplitude is , the phase angle is , the change in phase angle is , the active power is , the change in active power is , the reactive power is , the change in reactive power is , the phase angle sensitivity matrix related to active power is , the phase angle sensitivity matrix related to reactive power is , the voltage amplitude sensitivity matrix related to active power is , the voltage amplitude sensitivity matrix related to reactive power is ,but: ； The voltage sensitivity information is calculated as: 。 3. The method for dynamic partitioning of a distribution network considering network topology and time-varying resources according to claim 1 is characterized in that: The calculation of comprehensive electrical distance includes: For the The voltage change of each node under the influence of reactive power, For the The voltage change of each node under the influence of reactive power, For the The reactive variables of each node, For the Node pair The reactive voltage change sensitivity of each node, For the The voltage change of each node under the influence of reactive power, For the The voltage change of each node under the influence of reactive power, For the The reactive variables of each node, For the Node pair Active voltage change sensitivity of each node, For Node The reactive power change of the node Sensitivity of voltage amplitude, For Node The sensitivity of reactive power changes to its own voltage amplitude; Under the influence of reactive power, the voltage sensitivity is defined as Nodes and The voltage change relationship between the nodes is expressed as: ； No. Node pair The active voltage change sensitivity of each node is: ； Assume the comprehensive electrical distance is ,time Cluster The comprehensive electrical distance is ; According to reactive voltage sensitivity and active voltage sensitivity , construct Nodes and The expression of the comprehensive electrical distance between nodes is: ； 。 4. The method for dynamic partitioning of a distribution network considering network topology and time-varying resources according to claim 1 is characterized in that: The calculation of voltage regulation capability index includes: Region exist The voltage regulation capability at the moment is ,but: set up For the Nodes in The voltage deviation at the moment, For Region Neidi Nodes in The maximum voltage deviation at any moment; ； Indicates area exist The voltage regulation capability at the moment is: ； represents the active voltage sensitivity matrix, represents the reactive voltage sensitivity matrix, For Node exist The ability to adjust the active power to the maximum voltage deviation at any time. For Node exist The ability of the reactive power to adjust the maximum voltage deviation at any time. The sum of the edge weights of all nodes in the network is Time Node exist The adjustable active power of distributed photovoltaics at all times, is the sum of the edge weights of all nodes in the network. Time Node exist Adjustable active power with energy storage at all times, is the sum of the edge weights of all nodes in the network. Time Node exist Adjustable active power of flexible load at all times, The sum of the edge weights of all nodes in the network is Time Node exist The adjustable reactive power of distributed photovoltaic at all times, The sum of the edge weights of all nodes in the network is Time Node exist Adjustable reactive power for constant energy storage, The sum of the edge weights of all nodes in the network is Time Node exist Adjustable reactive power of flexible load at all times, The sum of the edge weights of all nodes in the network is Time Node exist The adjustable reactive power of the reactive power compensation device at any moment is: ； ； Assume that the voltage regulation capability of the distribution network is , is the number of zones, and the voltage regulation capacity of the distribution network is the average value of the regulation capacity of each zone, expressed as: 。 5. The method for dynamic partitioning of a distribution network considering network topology and time-varying resources according to claim 1 is characterized in that: The calculation of source-load correlation index includes: for Time source-load correlation index, is the photovoltaic output power, is the load size, for The covariance of photovoltaic output power and load at the moment, for The standard deviation of photovoltaic output power at each moment, for The standard deviation of the load at each moment is The moment source-load correlation index is: 。 6. The method for dynamic partitioning of a distribution network considering network topology and time-varying resources according to claim 1 is characterized in that: The calculation of comprehensive evaluation index includes: is a comprehensive evaluation index. For Region exist Voltage regulation capability at all times, for Comprehensive electrical distance indicators at all times, for Time source-load correlation index, For Region exist The weight coefficient of the voltage regulation capability at the moment, for The weight coefficient of the comprehensive electrical distance index at all times, for The weight coefficient of the source-load correlation index at the moment, and ,but: 。 7. The method for dynamic partitioning of a distribution network considering network topology and time-varying resources according to claim 1, characterized in that: set up For cluster The number of nodes of internal adjustable resources, For cluster The internal part is the number of nodes with load. The constraints include that a single node cannot form a partition and each partition contains at least one adjustable resource node. The constraints are expressed as: ； 。 8. The method for dynamic partitioning of a distribution network considering network topology and time-varying resources according to claim 1, characterized in that: By setting the threshold of the comprehensive electrical distance difference, the dynamic adjustment of the distribution network partition structure is triggered, including: by comparing the current time after the topology changes The electrical distance index value under the topology change is When the difference between the electrical distance index values ​​under the first threshold exceeds the first threshold, the distribution network partition structure is changed; suppose the current time after the topology change The electrical distance indicator value is , the moment before the topology change The electrical distance indicator value is , the first threshold is , then the dynamic adjustment method is expressed as: 。 9. The method for dynamic partitioning of a distribution network considering network topology and time-varying resources according to claim 1, characterized in that: By setting the threshold of the difference between the source-load correlation coefficient before and after, the dynamic adjustment of the distribution network partition structure is triggered, including: by comparing the current The source-load correlation index value is the same as the previous moment The source-load correlation index value is calculated, and the difference between the two is calculated. When the difference exceeds the preset second threshold, the source-load relationship changes, triggering the dynamic adjustment of the distribution network partition structure; suppose the current time The source-load correlation index is , the previous moment The source-load correlation index is , the second threshold is , then the dynamic adjustment method is expressed as: 。 10. A distribution network dynamic partitioning system considering network topology and resource time-varying characteristics, characterized in that: It includes evaluation index system establishment unit, multi-objective optimization model construction unit, setting unit and solution unit; An evaluation index system establishment unit is used to establish an evaluation index system for dynamic partitioning of distribution networks, including a comprehensive electrical distance index, a voltage regulation capability index, and a source-load time-space correlation index; the comprehensive electrical distance index is defined based on voltage sensitivity information and is used to describe the closeness of the electrical connection between any two nodes in the system; the voltage regulation capability index is used to characterize the ability of the adjustable active and reactive power of distributed resources in the region to regulate the maximum voltage deviation; the source-load time-space correlation index is used to characterize the correlation between the output of distributed resources and load demand; The multi-objective optimization model building unit is used to weight the comprehensive electrical distance index, voltage regulation capability index and source-load spatiotemporal correlation index according to the distribution network dynamic partition evaluation index system, take the comprehensive evaluation index as output, build the distribution network dynamic partition multi-objective optimization model and set the distribution network dynamic partition constraint conditions; A setting unit, used to trigger the dynamic adjustment of the distribution network partition structure by setting a threshold value of the comprehensive electrical distance difference and a threshold value of the source-load correlation coefficient difference between the previous and next moments; The solving unit is used to take the comprehensive evaluation index as the objective function, adopt the partition optimization algorithm to solve the multi-objective optimization model of dynamic partition of the distribution network, and obtain the optimal solution of the distribution network partition.

Images