Power distribution network three-phase imbalance optimization method based on power flow tracking partition

Abstract

The invention provides a power distribution network three-phase imbalance optimization method based on power flow tracking partition, and aims to solve the problem of three-phase voltage imbalance caused by large-scale distributed photovoltaic access in a low-voltage power distribution network. The method comprises the following steps: firstly, performing power analysis on a photovoltaic generator of each phase through load flow calculation, establishing a mathematical model of photovoltaic output and load distribution by using a downstream tracking method, and determining active power injection of each node; secondly, on the basis of defining the power flow distribution of the power grid, partitioning the low-voltage power distribution network according to the output radiation range of the photovoltaic generator; thirdly, aiming at the partitioning result of each phase, combining the power spring model and the power spring capacity, considering voltage deviation, network loss and operation cost in a power grid, respectively establishing an addressing model of each phase of power spring, and converting a three-phase voltage imbalance problem into a single-phase site selection optimization problem; and finally, solving the addressing model of the power spring by adopting a grey wolf algorithm, determining the access position of the power spring of each phase, and realizing optimization of voltage balance and load distribution.

Description

Technical Field

[0001] This invention relates to the field of low-voltage distribution network technology, and in particular to a method for optimizing three-phase imbalance in distribution networks based on power flow tracing and partitioning. Background Technology

[0002] With the rapid development of distributed photovoltaic (PV) power generation systems, especially their large-scale integration into low-voltage distribution networks, effectively addressing the power system imbalance caused by PV grid connection has become a significant technical challenge in the power sector. PV power generation, influenced by weather and seasonal factors, often exhibits strong volatility and intermittency, potentially leading to three-phase voltage imbalances in the power grid, thereby affecting grid stability, reliability, and power quality. Therefore, accurately identifying power flow distribution within the grid and effectively managing three-phase voltage imbalances has become a key technology for improving the operational efficiency and stability of low-voltage distribution networks.

[0003] Traditional methods for analyzing three-phase voltage imbalance often rely on static load balancing analysis, failing to adequately consider the dynamic changes in power flow after distributed generation (such as photovoltaic power generation) is connected to the grid. With more and more distributed photovoltaic systems being connected to the grid, their impact on three-phase voltage balance is becoming increasingly prominent. Therefore, there is an urgent need to develop a method that can accurately track power flow distribution and specifically optimize and control three-phase voltage imbalance. Summary of the Invention

[0004] To address the potential three-phase voltage imbalance problem in low-voltage distribution networks after the integration of distributed photovoltaic (PV) systems, a distribution network three-phase imbalance optimization method based on power flow tracking and zoning is developed, which has significant practical and technical value. This method can achieve more precise voltage control and load optimization control, thereby effectively suppressing imbalance and improving grid stability and power quality.

[0005] The objective of this invention can be achieved through the following technical solutions: Step S1: For low-voltage distribution network systems with large-scale distributed photovoltaic (three-phase and single-phase) access, perform power flow calculations for each phase to determine the active power injected into each node's DG and the sum of all injected active power at each node. Step S2: Based on the clear distribution of all power flows in the network, establish a model of all DG outputs to load distribution for each item using the downstream tracing method; Step S3: After obtaining the radiation range of photovoltaic output for each phase, perform overall partitioning of the three sets of data to obtain the final comprehensive partitioning result; Step S4: Based on the partitioning, considering the mathematical model of the electric spring and various constraints, determine the capacity of the electric spring for the three phases A, B, and C, and establish the addressing model of the electric spring respectively, simplifying the three-phase voltage imbalance optimization problem into a single-phase addressing problem. Step S5: Solve the electric spring addressing model using the Grey Wolf algorithm.

[0006] 2. Step S1 specifically includes the following steps: Step S11: Perform power flow calculations on phase A of the low-voltage distribution network system to obtain the active power injection of the DG at each node of phase A. And the injection of all active power at each node. , where n is the total number of nodes in the entire system; Step S12: Construct the matrix and , respectively A diagonal matrix of nodal photovoltaic active power output and total nodal injected active power; Step S13: Calculate the downstream assignment matrix ,in: Let the downstream assignment matrix be a lossless network with n nodes. For the line The active power transmitted from node i to node j. Total active power injected into node j: .

[0007] 3. Step S2 specifically includes the following steps: Step S21: Construct the A-phase allocation coefficient matrix : In the matrix In this system, the load of each node is entirely supplied by its own power sources, so the sum of the elements in each column of the matrix is ​​1. However, the sum of the elements in each row is not necessarily 1. Suppose a node g has a power source but receives no external power; in this case, the load of node g can only be supplied by its own power source. This power source not only supplies power to node g but also transmits power to other nodes; therefore, the sum of the elements in the row containing node g will be greater than 1. For accurate calculations, line losses typically need to be converted to equivalent loads at both ends of the line. To simplify the calculation process, this patent assumes a lossless network and ignores the impact of line losses during power flow tracing.

[0008] 4. Step S3 specifically includes the following steps: Step S31: Through the matrix Establish a model of the load distribution of all photovoltaic outputs in phase A, obtain the output radiation range of the node photovoltaic generators, and divide phase A into zones according to the output radiation range of the photovoltaic power generation system; Step S32: Repeat steps S11-S21 for phases BC to obtain... Obtain the output radiation range of the node photovoltaic generator, and divide the BC phases into two zones according to the output radiation range of the photovoltaic power generation system. Step S33: For the power flow tracking matrix of phases A, B, and C, analyze the radiation range of the photovoltaic system. After completing the partitioning, perform overall partitioning of the three sets of data to obtain the final comprehensive partitioning result.

[0009] 5. Step S4 specifically includes the following steps: Step S41: Verify the feasibility of the electric spring ES in addressing voltage imbalance in a three-phase low-voltage distribution network by analyzing its circuit model. Step S42: Connect the electric spring ES and non-critical loads in series to form a smart load. Take the intersection of the active power, reactive power, voltage, and current ranges on the smart load to obtain the adjustment range of the smart load, thereby constructing a mathematical model of the electric spring; Step S43: On phase A of each partition, construct the objective function considering the average voltage deviation of each node, minimum network loss, and minimum operating cost. Considering constraints such as system power flow, node voltage upper and lower limits, and dynamic changes in photovoltaic output, determine the power spring capacity and establish the addressing model for the power spring.

[0010] 6. Step S5 specifically includes the following steps: The Grey Wolf algorithm is used to solve the A-phase electric spring addressing model: Step S51: Stratify the wolf pack into social hierarchies, determine the rank within the pack, and select the three gray wolves with the best fitness to be labeled as follows: and The rest are The optimization process mainly consists of and These three optimal solutions have been completed; Step S52: When searching, the gray wolf will approach and surround its prey, through... and Gray wolves guide the search for potential prey (optimal solution). In each iteration, the top 3 gray wolves in the current population are retained. and Based on their location information, the positions of other gray wolves are updated, allowing the gray wolves to gradually approach their prey; Step S53: Adjust coefficient A so that it is in the range of [-1,1], so that the gray wolf's search position is anywhere between the current position and the prey; Step S54: Candidate Gray Wolf Relies and Information, when A distributed global search is performed, while the c-phasor provides random weights to avoid getting trapped in local optima; Step S55: Considering phases B and C, repeat steps S43-S54 to determine the connection position of the power spring. Attached Figure Description

[0011] Figure 1 The flowchart is based on the present invention; Figure 2 This is the equivalent model of an electric spring in the s-domain. Detailed Implementation

[0012] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0013] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0014] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0015] like Figure 1 As shown, this paper proposes a three-phase voltage imbalance optimization method based on power flow tracking and zoning for distribution networks, which addresses the potential three-phase voltage imbalance problem in low-voltage distribution networks after distributed photovoltaic (PV) integration. This method has significant practical and technical value. It enables more precise current distribution and load optimization control, effectively suppressing imbalance and improving grid stability and power quality.

[0016] The objective of this invention can be achieved through the following technical solutions: Step S1: For low-voltage distribution network systems with large-scale distributed photovoltaic (three-phase and single-phase) access, perform power flow calculations for each phase to determine the active power injected into each node's DG and the sum of all injected active power at each node. Step S2: Based on the clear distribution of all power flows in the network, establish a model of all DG outputs to load distribution for each item using the downstream tracing method; Step S3: After obtaining the radiation range of photovoltaic output for each phase, perform overall partitioning of the three sets of data to obtain the final comprehensive partitioning result; Step S4: Based on the partitioning, considering the mathematical model of the electric spring and various constraints, determine the capacity of the electric spring for the three phases A, B, and C, establish the addressing model of the electric spring, and simplify the three-phase voltage imbalance optimization problem into a single-phase addressing problem. Step S5: Solve the electric spring addressing model using the Grey Wolf algorithm.

[0017] 2. Step S1 specifically includes the following steps: Step S11: Perform power flow calculations on phase A of the low-voltage distribution network system to obtain the active power injection of the DG at each node of phase A. And the injection of all active power at each node. , where n is the total number of nodes in the entire system; Step S12: Construct the matrix and , respectively A diagonal matrix of nodal photovoltaic active power output and total nodal injected active power; Step S13: Calculate the downstream assignment matrix ,in: Let the downstream assignment matrix be a lossless network with n nodes. For the line The active power transmitted from node i to node j. Total active power injected into node j: .

[0018] 3. Step S2 specifically includes the following steps: Step S21: Construct the A-phase allocation coefficient matrix : In the matrix In this system, the load of each node is entirely supplied by its own power sources, so the sum of the elements in each column of the matrix is ​​1. However, the sum of the elements in each row is not necessarily 1. Suppose a node g has a power source but receives no external power; in this case, the load of node g can only be supplied by its own power source. This power source not only supplies power to node g but also transmits power to other nodes; therefore, the sum of the elements in the row containing node g will be greater than 1. For accurate calculations, line losses typically need to be converted to equivalent loads at both ends of the line. To simplify the calculation process, this patent assumes a lossless network and ignores the impact of line losses during power flow tracing.

[0019] 4. Step S3 specifically includes the following steps: Step S31: Through the matrix Establish a model of the load distribution of all photovoltaic outputs in phase A, obtain the output radiation range of the node photovoltaic generators, and divide phase A into zones according to the output radiation range of the photovoltaic power generation system; Step S32: Repeat steps S11-S21 for phases BC to obtain... Obtain the output radiation range of the node photovoltaic generator, and divide the BC phases into two zones according to the output radiation range of the photovoltaic power generation system. Step S33: For the power flow tracking matrix of phases A, B, and C, analyze the radiation range of the photovoltaic system. After completing the partitioning, perform overall partitioning of the three sets of data to obtain the final comprehensive partitioning result.

[0020] 5. Step S4 specifically includes the following steps: Step S41: Establish a mathematical model of the electric spring, and equate the smart load to a load combination consisting of the electric spring ES connected in series with a non-critical load. Figure 2 This is the equivalent model of an electric spring in the s-domain. The voltage on the grid side. This indicates the impedance value on the line. The current flowing through the power grid side. For the voltage on the critical load, The current flowing through the critical load. This represents the impedance value of the critical load. The impedance value is for a non-critical load, and the voltage across it is... , It is connected in series with an electric spring to form a smart load, and the current flowing through the smart load is The voltage across the electric spring is The inductor and capacitor of the low-pass filter are respectively With 1 / The currents flowing through the inductor and capacitor of the low-pass filter are respectively and The output square wave voltage value of the electric spring inverter circuit is... .

[0021] According to Norton's equivalence theorem: in This is the equivalent reactance under Norton's theorem.

[0022] According to KVL: Therefore: In the equivalent circuit, we have: in Equivalent reactance The current flowing through it.

[0023] Solving the above equations simultaneously, we get: in: The above analysis shows that the voltage across the critical load is... Subject to grid voltage The output square wave voltage value of the power spring inverter circuit is Therefore, the voltage across the load can be controlled by an electric spring, thereby suppressing the three-phase voltage imbalance.

[0024] Step S42: Establish corresponding models and corresponding model constraints based on the active and reactive feasible domains of the intelligent load.

[0025] The internal voltage and power of the intelligent load must meet the following conditions: in, This refers to the voltage amplitude at the connection bus of the intelligent load. This refers to the current amplitude at the connection bus of the intelligent load. The amplitude of the non-critical load voltage. The phase angle of the non-critical load voltage. This is a non-critical load impedance angle; , , and These represent the amplitude and phase angle of the voltage and current of the electric spring, respectively.

[0026]

[0027] Step S43: On phase A of each partition, construct the objective function considering the average voltage deviation of each node, minimum network loss, and minimum operating cost. Considering constraints such as system power flow, node voltage upper and lower limits, and dynamic changes in photovoltaic output, determine the power spring capacity and establish the addressing model for the power spring.

[0028] The overall optimization objective includes three sub-optimization objectives: average voltage deviation at each node, minimum network loss, and minimum operating cost. In the formula: , and The weighting coefficients for the three optimization objectives correspond to the minimum average node voltage deviation, minimum active power loss, and minimum operating cost, respectively. The average rated voltage of the node is 1 per unit. and These represent the minimum active power network loss and the minimum operating cost of the distribution network obtained by single-objective optimization without considering the adjustment of the power spring and distributed photovoltaic (DG).

[0029] The average voltage deviation at each node satisfies: Where N is the number of network nodes. This represents the actual voltage at the node.

[0030] For network active power loss, the following conditions must be met: in, For each line, the effective value of the current. This represents the resistance value for each line.

[0031] E represents the operating cost of the power distribution network, which satisfies: in, , , and These are the costs of electricity purchased by the distribution network from the upper-level grid per unit time, the operating costs of distributed generation, subsidies to users after using ES, and the cost of curtailment. The power of each node satisfies the following constraints: in, and For injection nodes Active and reactive power, and For nodes The active and reactive power of the connected load. For nodes Admittance between; For nodes The voltage phase angle difference between them and For nodes voltage.

[0032] The voltage at each node satisfies the voltage constraint condition: in This is the lower limit of the node voltage. This represents the upper limit of the node voltage.

[0033] 7) A distributed photovoltaic system at a certain node is connected to the grid via a photovoltaic inverter. The output of the low-voltage photovoltaic inverter varies within a certain range, satisfying the following constraints: in, They are respectively The active and reactive power output of the inverter at all times; The power factor angle; Power factor The sine and cosine values, The maximum value is 0.33; This refers to the rated power of the inverter.

[0034] 6. Step S5 specifically includes the following steps: The Grey Wolf algorithm is used to solve the A-phase electric spring addressing model: Step S51: Stratify the wolf pack into social hierarchies, determine the rank within the pack, and select the three gray wolves with the best fitness to be labeled as follows: and The rest are The optimization process mainly consists of and These three optimal solutions have been completed; Step S52: When searching, the gray wolf will approach and surround its prey, through... and Gray wolves guide the search for potential prey (optimal solution). In each iteration, the top 3 gray wolves in the current population are retained. and Based on their location information, the positions of other gray wolves are updated, allowing the gray wolves to gradually approach their prey; Step S53: Adjust coefficient A so that it is in the range of [-1,1], so that the gray wolf's search position is anywhere between the current position and the prey; Step S54: Candidate Gray Wolf Relies and Information, when A distributed global search is performed, while the c-phasor provides random weights to avoid getting trapped in local optima; Step S55: Considering phases B and C, repeat steps S43-S54 to determine the connection position of the power spring.

[0035] In summary, this invention addresses the potential three-phase voltage imbalance problem that may arise after distributed photovoltaic (PV) power is integrated into a low-voltage distribution network. It proposes a power flow tracking-based zoning optimization method for distribution network three-phase imbalance. This method calculates the power flow distribution of each phase and, combined with the output radiation range of the PV generators, optimizes the distribution network by zoning, thereby effectively suppressing voltage imbalance. The optimization process incorporates power spring regulation, achieving a more reasonable management of three-phase voltage imbalance and improving grid stability and power quality.

[0036] The above description is merely a preferred embodiment of the present invention and does not limit the implementation and protection scope of the present invention. Those skilled in the art should realize that any equivalent substitutions and obvious changes made based on the description and illustrations of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing three-phase imbalance in a distribution network based on power flow tracing and partitioning, characterized in that: Includes the following steps Step S1: For low-voltage distribution network systems with large-scale distributed photovoltaic (three-phase and single-phase) access, perform power flow calculations for each phase to determine the active power injected into each node's DG and the sum of all injected active power at each node. Step S2: Based on the clear distribution of all power flows in the network, establish a model of all DG outputs to load distribution for each item using the downstream tracing method; Step S3: After obtaining the radiation range of photovoltaic output for each phase, perform overall partitioning of the three sets of data to obtain the final comprehensive partitioning result; Step S4: Based on the partitioning, considering the mathematical model of the electric spring and various constraints, determine the capacity of the electric spring for the three phases A, B, and C, and establish the addressing model of the electric spring respectively, simplifying the three-phase voltage imbalance optimization problem into a single-phase addressing problem. Step S5: Solve the electric spring addressing model using the Grey Wolf algorithm.

2. The method for optimizing three-phase imbalance in a distribution network based on power flow tracing partitioning as described in claim 1, characterized in that: Step S1 specifically includes the following steps: Step S11: Perform power flow calculations on phase A of the low-voltage distribution network system to obtain the active power injection of the DG at each node of phase A. And the injection of all active power at each node. , where n is the total number of nodes in the entire system; Step S12: Construct the matrix and , respectively A diagonal matrix of nodal photovoltaic active power output and total nodal injected active power; Step S13: Calculate the downstream assignment matrix ,in: Let the downstream assignment matrix be a lossless network with n nodes. For the line The active power transmitted from node i to node j. Total active power injected into node j: 。 3. The method for optimizing three-phase imbalance in a distribution network based on power flow tracing partitioning as described in claim 1, characterized in that: Step S2 specifically includes the following steps: Step S21: Construct the A-phase allocation coefficient matrix : In the matrix In this system, the load of each node is entirely supplied by the various power sources within the system, so the sum of the elements in each column of the matrix is ​​1; however, the sum of the elements in each row is not necessarily 1; assuming a node g has a power source but does not receive any external power, then the load of node g can only be supplied by its own power source; this power source not only supplies power to node g but also transmits power to other nodes, therefore, the sum of the elements in the row containing node g will be greater than 1; when performing accurate calculations, line losses usually need to be converted into equivalent loads at both ends of the line; to simplify the calculation process, this patent assumes the network is lossless and ignores the impact of line losses when performing power flow tracing.

4. The method for optimizing three-phase imbalance in a distribution network based on power flow tracing partitioning according to claim 1, characterized in that: Step S3 specifically includes the following steps: Step S31: Through the matrix Establish a model of the load distribution of all photovoltaic outputs in phase A, obtain the output radiation range of the node photovoltaic generators, and divide phase A into zones according to the output radiation range of the photovoltaic power generation system; Step S32: Repeat steps S11-S21 for phases BC to obtain... Obtain the output radiation range of the node photovoltaic generator, and divide the BC phases into two zones according to the output radiation range of the photovoltaic power generation system. Step S33: For the power flow tracking matrix of phases A, B, and C, analyze the radiation range of the photovoltaic system. After completing the partitioning, perform overall partitioning of the three sets of data to obtain the final comprehensive partitioning result.

5. The method for optimizing three-phase imbalance in a distribution network based on power flow tracing partitioning according to claim 1, characterized in that: Step S4 specifically includes the following steps: Step S41: Verify the feasibility of the electric spring ES in addressing voltage imbalance in a three-phase low-voltage distribution network by analyzing its circuit model. Step S42: Connect the electric spring ES and non-critical loads in series to form a smart load; take the intersection of the active power, reactive power, voltage and current ranges on the smart load to obtain the adjustment range of the smart load, thereby constructing a mathematical model of the electric spring; Step S43: On phase A of each partition, construct the objective function by considering the average voltage deviation of each node, the minimum network loss, and the minimum operating cost; determine the power spring capacity by considering constraints such as system power flow, node voltage upper and lower limits, and dynamic changes in photovoltaic output, and establish the addressing model of the power spring.

6. The method for optimizing three-phase imbalance in a distribution network based on power flow tracing partitioning according to claim 1, characterized in that: Step S5 specifically includes the following steps: The Grey Wolf algorithm is used to solve the A-phase electric spring addressing model: Step S51: Stratify the wolf pack into social hierarchies, determine the rank within the pack, and select the three gray wolves with the best fitness to be labeled as follows: and The rest are The optimization process mainly consists of and These three optimal solutions have been completed; Step S52: When searching, the gray wolf will approach and surround its prey, through... and Gray wolves guide the search for potential prey (optimal solution); in each iteration, the best 3 gray wolves in the current population are retained. and Based on their location information, the positions of other gray wolves are updated, allowing the gray wolves to gradually approach their prey; Step S53: Adjust coefficient A so that it is in the range of [-1,1], so that the gray wolf's search position is anywhere between the current position and the prey; Step S54: Candidate Gray Wolf Relies and Information, when A distributed global search is performed, while the c-phasor provides random weights to avoid getting trapped in local optima; Step S55: Considering phases B and C, repeat steps S43-S54 to determine the connection position of the power spring.

Images