Calculation method for semi-AC power flow in distribution networks considering reactive power distribution and network losses

By calculating reactive power distribution and network losses using a semi-AC power flow model, the problem of low computational efficiency in large-scale power distribution systems is solved, achieving fast and accurate power flow calculations, which are suitable for power distribution system optimization and control.

CN117439088BActive Publication Date: 2026-11-17TIANJIN UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202311499944.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-11-10
Publication Date
2026-11-17
Estimated Expiration
2043-11-10

AI Technical Summary

Technical Problem

In existing technologies, AC power flow calculation is inefficient and difficult to converge in large-scale power distribution systems, and DC power flow models cannot accurately calculate reactive power distribution and network losses.

Method used

A semi-AC power flow model is adopted. By establishing the node admittance matrix and voltage phase angle vector, reactive power distribution and network loss are calculated, branch resistance and voltage difference are simplified, and the calculation accuracy and speed are improved.

Benefits of technology

It enables rapid and accurate calculation of reactive power distribution and network losses in power distribution systems, improves calculation convergence and speed, and becomes an effective tool for optimizing operation and control.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117439088B_ABST
    Figure CN117439088B_ABST
Patent Text Reader

Abstract

This invention discloses a semi-AC power flow calculation method for distribution networks that considers reactive power distribution and network losses. The method mainly includes: establishing active and reactive power matrices and node admittance matrices based on the operating parameters of the distribution system; and calculating the following results of the semi-AC power flow model: voltage phase angle vector, voltage virtual phase angle vector, active and reactive power distribution of branches without considering network losses, active and reactive power losses of branches, and active and reactive power distribution of branches considering network losses. Compared to DC power flow models, this invention improves the accuracy of calculations and can calculate reactive power distribution and network losses. Simultaneously, compared to AC power flow models, it significantly shortens the power flow calculation time, achieving rapid calculation and exhibiting good convergence. It is a powerful power flow calculation tool for the optimized operation, planning, and control of distribution systems.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of power flow calculation models for distribution systems, and relates to a method for calculating the semi-AC power flow of a distribution network that takes into account reactive power distribution and network losses. Background Technology

[0002] Power flow calculation is a powerful tool for power distribution system analysis. Power flow calculation can generally be divided into DC power flow and AC power flow. [1] While AC power flow algorithms are accurate, their nonlinear nature leads to increasing computational complexity and efficiency as the scale of the power distribution system grows, potentially even resulting in non-convergence. Simplifying the AC power flow model linearly can reduce computational load and improve efficiency. DC power flow models are currently the most widely used linear power flow models, but they suffer from low accuracy and cannot calculate voltage, reactive power distribution, or network losses.

[0003] [References]

[0004] [1] Zhao Jinquan, Ye Junling, Deng Yong. Comparative analysis of DC power flow and AC power flow[J]. Power System Technology, 2012, 36(10):147-152.DOI:10.13335 / j.1000-3673.pst.2012.10.029.

[0005] [2] Gao Shanshan, Qin Wenping, Han Xiaoqing, et al. Two-level real-time reconfiguration of distribution network considering load time-varying characteristics and distributed generation output fluctuations [J]. Power System Technology, 2014, 38(11):3260-3265. DOI:10.13335 / j.1000-3673.pst.2014.11.049. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention proposes a semi-AC power flow calculation method for distribution networks that considers reactive power distribution and network losses. The semi-AC power flow model involved in this method is a linear model, which, compared to the DC power flow model, can simultaneously calculate reactive power distribution and network losses, and improves the accuracy of the calculation. Furthermore, the computational load and time are much smaller than those of the nonlinear AC power flow model, thereby significantly reducing the power flow calculation time, accelerating the calculation speed, and improving the convergence of the calculation. This makes it a powerful power flow calculation tool for the optimized operation, planning, and control of distribution systems.

[0007] To address the aforementioned technical problems, this invention proposes a method for calculating the semi-AC power flow of a distribution network that considers reactive power distribution and network losses, comprising the following steps:

[0008] Step 1) Input power distribution system operating parameters: Determine the power distribution system topology, the unit resistance r0 of the conductors, the unit reactance x0 of the conductors, and the maximum current carrying capacity I of the conductors.max The resistance r of each branch in the power distribution system ij and reactance x ij Voltage level U of the power distribution system N The active load P connected to each node i and reactive load Q i and the types of each node;

[0009] Step 2) Construct an N-row, 1-column active power matrix P and an N-row, 1-column reactive power matrix Q: N is the number of nodes, and N is a positive integer greater than or equal to 2; if the i-th node is not a slack node, 1≤i≤N, the element P in the i-th row and 1-th column of the active power matrix P... i1 = -P i The element Q in the i-th row and 1-th column of the reactive power matrix Q i1 =-Q i If the i-th node is a saturated node, 1≤i≤N, then the element P in the i-th row and 1-th column of the active power matrix P is... i1 =0, the element Q in the i-th row and 1-th column of the reactive power matrix Q i1 =0;

[0010] Step 3) Construct an N-row, N-column node admittance matrix Y: In the node admittance matrix Y, for the element Y in the i-th row and j-th column... ij , 1≤i≤N, 1≤j≤N, i≠j; if there is no connection between node i and node j, then Y ij =0; if there is a connection between node i and node j, then Y = 0. ij = -1 / x ij For the element Y in the i-th row and i-th column ii 1≤i≤N, if node i is a balanced node, then Y ii =0; If node i is not a balanced node, and node i is connected to the other n nodes, 1≤n≤N-1, then Y = 0. ii =n / x ij ;

[0011] Step 4) Calculate the voltage phase angle vector θ and the voltage virtual phase angle vector θ Q :

[0012] P=Yθ (1)

[0013] Q = Yθ Q (2)

[0014] According to equation (1), θ = Y -1 P, according to equation (2), θ Q =Y -1 Q;

[0015] Step 5) Calculate the branch active power distribution P without considering network losses. ijThe reactive power distribution of the branch circuit Q ij ',include:

[0016] First, establish the distribution equations (3) and (4) for the power flow between node i and node j:

[0017]

[0018]

[0019] In equations (3) and (4): P ij-AC Q ij-AC These represent the AC active power distribution and AC reactive power distribution at the beginning of the branch circuit, respectively; G ij and B ij These represent the mutual conductance and mutual susceptance between nodes i and j, respectively; U i and U j Let θ represent the voltage magnitudes at nodes i and j, respectively; ij The voltage phase angle difference between nodes i and j;

[0020] Secondly, since urban power distribution networks mostly have short lines and generally use cables, the network loss ratio is low and the voltage drop is small. In addition, the power grid itself and distributed generation (DG) have a certain voltage regulation capability. Therefore, the voltage U at each node can be approximated. i ≈1.00 pu, inter-node voltage phase angle difference θ ij Very small, has sinθ ij ≈θ ij ≈θ i -θ j cosθ ij ≈1, neglecting branch resistance r ij And all the branches to the ground, so the simplified equation (3) becomes equation (5), which is used to calculate the active power distribution P of the branches without considering network losses. ij ′:

[0021]

[0022] In equation (5): θ i and θ j Let be the voltage phase angles of nodes i and j, respectively;

[0023] Unlike DC power flow models, the method proposed in this invention does not ignore reactive power distribution. By analogy with equation (5), equation (6) is obtained, which is used to calculate the reactive power distribution Q of branches without considering network losses. ij ′:

[0024]

[0025] In formula (6): and These are the virtual phase angles of the voltages at nodes i and j, respectively.

[0026] Step 6) Calculate the active power loss ΔP of each branch. ij and reactive power loss ΔQ ij Based on the active power distribution P of each branch without considering network losses ij Divide by node voltage U i The current I flowing through each branch is obtained. ij :

[0027]

[0028] Calculate the active power loss ΔP of each branch. ij and reactive power loss ΔQ ij :

[0029]

[0030] Step 7) Calculate the active power distribution P of each branch considering network losses. ij and reactive power distribution Q ij :

[0031]

[0032] Compared with the prior art, the beneficial effects of the present invention are:

[0033] (1) This invention proposes a semi-AC power flow model for power distribution system power flow calculation. The model is simple, is a linear model, and is easy to converge.

[0034] (2) The present invention uses a semi-AC power flow model to calculate reactive power distribution and network loss at the same time, which improves the accuracy of calculation compared with the DC power flow model.

[0035] (3) Since the power flow model used in the calculation method of the present invention is a semi-AC power flow model, the calculation method of the present invention is fast. Compared with the AC power flow model, it can greatly improve the power flow calculation speed and realize fast calculation. It is a powerful power flow calculation tool for the optimization, planning and control of power distribution system. Attached Figure Description

[0036] Figure 1 This is a flowchart of the semi-AC power flow algorithm of the present invention;

[0037] Figure 2 This is a schematic diagram of the IEEE 33-node example provided in the research materials of this invention. Detailed Implementation

[0038] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but the following embodiments are by no means intended to limit the present invention.

[0039] like Figure 1 As shown, the present invention proposes a method for calculating the semi-AC power flow of a distribution network that takes into account reactive power distribution and network losses, comprising the following steps:

[0040] Step 1) Input power distribution system operating parameters: Determine the power distribution system topology, the unit resistance r0 of the conductors, the unit reactance x0 of the conductors, and the maximum current carrying capacity I of the conductors. max The resistance r of each branch in the power distribution system ij and reactance x ij Voltage level U of the power distribution system N The active load P connected to each node i and reactive load Q i and the types of each node;

[0041] Step 2) Construct an N-row, 1-column active power matrix P and an N-row, 1-column reactive power matrix Q: N is the number of nodes, and N is a positive integer greater than or equal to 2; if the i-th node is not a slack node, 1≤i≤N, the element P in the i-th row and 1-th column of the active power matrix P... i1 = -P i The element Q in the i-th row and 1-th column of the reactive power matrix Q i1 =-Q i If the i-th node is a saturated node, 1≤i≤N, then the element P in the i-th row and 1-th column of the active power matrix P is... i1 =0, the element Q in the i-th row and 1-th column of the reactive power matrix Q i1 =0.

[0042] Step 3) Construct an N-row, N-column node admittance matrix Y: In this node admittance matrix Y, for the element Y in the i-th row and j-th column... ij , 1≤i≤N, 1≤j≤N, i≠j; if there is no connection between node i and node j, then Y ij =0; if there is a connection between node i and node j, then Y = 0. ij = -1 / x ij For the element Y in the i-th row and i-th column ii 1≤i≤N, if node i is a balanced node, then Y ii =0; If node i is not a balanced node, and node i is connected to the other n nodes, 1≤n≤N-1, then Y = 0. ii =n / x ij .

[0043] Step 4) Calculate the voltage phase angle vector θ and the voltage virtual phase angle vector θ Q :

[0044] P=Yθ (1)

[0045] Q = Yθ Q(2)

[0046] In equation (1), P is the N-row, 1-column active power matrix, Y is the N-row, N-column nodal admittance matrix, and θ is the N-row, 1-column voltage phase angle vector. According to equation (1), the voltage phase angle vector θ can be calculated as θ = Y. -1 P.

[0047] In equation (2), Q is an N-row, 1-column reactive power matrix, Y is an N-row, N-column nodal admittance matrix, and θ Q Let θ be the voltage virtual phase angle vector with N rows and 1 column. According to equation (2), the voltage virtual phase angle vector θ can be calculated. Q θ Q =Y -1 Q.

[0048] As can be seen from equation (2), the method of the present invention does not ignore reactive power like the DC power flow model, but calculates the voltage virtual phase angle vector θ based on the nodal admittance matrix Y and the reactive power matrix Q. Q This lays the foundation for subsequent calculations of reactive power distribution in the power distribution system.

[0049] Step 5) Calculate the branch active power distribution P without considering network losses. ij The reactive power distribution of the branch circuit Q ij ',include:

[0050] First, establish the distribution equations (3) and (4) for the power flow between node i and node j:

[0051]

[0052]

[0053] In equations (3) and (4): P ij-AC Q ij-AC These represent the AC active power distribution and AC reactive power distribution at the beginning of the branch circuit, respectively; G ij and B ij These represent the mutual conductance and mutual susceptance between nodes i and j, respectively; U i and U j Let θ represent the voltage magnitudes at nodes i and j, respectively; ij The voltage phase angle difference between nodes i and j;

[0054] Secondly, since urban power distribution networks mostly have short lines and generally use cables, the network loss ratio is low and the voltage drop is small. In addition, the power grid itself and distributed generation (DG) have a certain voltage regulation capability. Therefore, the voltage U at each node can be approximated. i ≈1.00 pu, inter-node voltage phase angle difference θ ij Very small, has sinθ ij ≈θij ≈θ i -θ j cosθ ij ≈1, neglecting branch resistance r ij And all the branches to the ground, so the simplified equation (3) becomes equation (5), which is used to calculate the active power distribution P of the branches without considering network losses. ij ′:

[0055]

[0056] In equation (5): θ i and θ j Let be the voltage phase angles of nodes i and j, respectively;

[0057] Unlike DC power flow models, the method proposed in this invention does not ignore reactive power distribution. By analogy with equation (5), equation (6) is obtained, which is used to calculate the reactive power distribution Q of branches without considering network losses. ij ′, where Q is the reactive power distribution of each branch considering network loss in subsequent calculations. ij Lay the foundation:

[0058]

[0059] In formula (6): and These are the virtual phase angles of the voltages at nodes i and j, respectively.

[0060] Step 6) Calculate the active power loss ΔP of each branch. ij and reactive power loss ΔQ ij :

[0061] First, based on the active power distribution P of each branch without considering network losses. ij Divide by node voltage U i The current I flowing through each branch is obtained. ij :

[0062]

[0063] Calculate the active power loss ΔP of each branch. ij and reactive power loss ΔQ ij :

[0064]

[0065] In this invention, unlike DC power flow models that ignore network losses, the active power distribution P of each branch, which does not consider network losses, is utilized. ij ′ and branch head node voltage U i The current I flowing through each branch was calculated. ij Then, the active power loss ΔP of each branch was calculated.ij and reactive power loss ΔQ ij The active power distribution P of each branch is considered for subsequent calculations to account for network losses. ij and reactive power distribution Q ij It laid the foundation.

[0066] Step 7) Calculate the active power distribution P of each branch considering network losses. ij and reactive power distribution Q ij :

[0067]

[0068] Research Materials

[0069] (1) Basic information about the example used

[0070] The power distribution system topology used in this invention is an IEEE 33-bus system, such as... Figure 2 As shown. The conductor type is LGJ-185, the conductor's unit resistance r0 is 0.18Ω / km, the conductor's unit reactance x0 is 0.37Ω / km, and the conductor's maximum current carrying capacity I... max The voltage level is 515A. Distribution system voltage level U N The voltage is 10kV, and the capacity is 8.920MVA; the per-unit base capacity is taken as 8.920MVA. Assuming each branch length is 0.10km, then the resistance r of each branch is... ij The resistance is 0.018Ω, and the reactance of each branch is x. ij The resistance is 0.037Ω. The active and reactive loads connected to each node are shown in Table 1. All nodes are PQ type nodes. Node 1 is not connected to any load and is a balanced node.

[0071] Table 1 Load Parameters

[0072] Node number Load value / kW+jkvar Node number Load value / kW+jkvar 2 100+j60 18 90+j40 3 90+j40 19 90+j40 4 120+j80 20 90+j40 5 60+j30 21 90+j40 6 60+j20 22 90+j40 7 200+j100 23 90+j50 8 200+j100 24 420+j200 9 60+j20 25 420+j200 10 60+j20 26 60+j25 11 45+j30 27 60+j25 12 60+j35 28 60+j20 13 60+j35 29 120+j70 14 120+j80 30 200+j100 15 60+j10 31 150+j70 16 60+j20 32 210+j100 17 60+j20 33 60+j40

[0073] (2) For the above example, the voltage amplitude set in the DC power flow model and the semi-AC power flow model is compared with the voltage amplitude of each node calculated by the AC power flow model. The comparison results are shown in Table 2.

[0074] Table 2 Comparison of Voltage Amplitude

[0075]

[0076]

[0077] As shown in Table 2, when comparing the node voltage amplitudes, although there are errors in the setting results of the DC power flow model and the calculation results of the AC power flow model, the maximum error is only 0.6948% (as shown in node numbers 17 or 18 in Table 2), which is relatively small. Therefore, in terms of voltage amplitude processing, the present invention adopts the same processing method as the DC power flow model, setting the voltage amplitude of each node to 1.00 pu. The calculation result error is small, and the power flow calculation speed is improved.

[0078] (3) For the above example, the power distribution results of each branch calculated using the DC power flow model, the AC power flow model and the semi-AC power flow model of the present invention are compared, as shown in Table 3.

[0079] Table 3 Power Distribution

[0080]

[0081]

[0082] As shown in Table 3, the DC power flow model can only calculate the active power distribution, and more than half of the errors exceed or approach 10%, with the maximum error being 16.9385%, indicating a large error. Furthermore, it cannot calculate the reactive power distribution. The semi-AC power flow model of this invention can calculate both active and reactive power distributions simultaneously, with very small errors; the maximum error is only 0.3851%, and most errors are below 0.3%.

[0083] (4) For the above example, the active power loss and power loss rate calculated by the DC power flow model, the AC power flow model and the semi-AC power flow model of the present invention are compared, as shown in Table 4.

[0084] Table 4 Comparison of Network Losses

[0085] Comparison items Active power loss / kW <![CDATA[Network loss rate [2] > Network loss error Network loss rate error DC power flow algorithm 0 0 100% 100% Semi-AC power flow algorithm 8.5165 0.2287% 19.9344% 19.8888% AC power flow algorithm 10.6369 0.2855% 0 0

[0086] The relative errors of active power loss and network loss rate are calculated based on the AC power flow algorithm. As shown in Table 4, since the active power loss and network loss rate of the DC power flow algorithm are both set to 0, the error is 100%. In this invention, the network loss error calculated by the semi-AC power flow algorithm is 19.9344%, the network loss rate error is 19.8888%, and the absolute error of the network loss rate is only 0.0568%, which greatly improves the accuracy compared to the DC power flow algorithm.

[0087] (5) Comparison of computing speed

[0088] The simulation calculations of this invention are based on the MATLAB R2018B environment, with an Intel(R) Core(TM) i7-7700HQ processor with a main frequency of 2.80GHz and 16GB of RAM. For the IEEE 33-node system, three power flow algorithms were used to calculate 1000 times each, and the average calculation time for one power flow result is shown in Table 5 below.

[0089] Table 5 Comparison of Calculation Time

[0090] Comparison items Average computation time / ms DC power flow algorithm 0.030787 Semi-AC power flow algorithm 0.052745 AC power flow algorithm 1.541621

[0091] As shown in Table 5, the computation time of the semi-AC power flow algorithm is similar to that of the DC power flow algorithm, at only 0.053ms, which is much smaller than the computation time of the AC power flow algorithm (1.542ms), reducing the computation time by 96.58%.

[0092] To compare the computational complexity in a more general sense, Table 6 shows the number of operations and time complexity of the three power flow algorithms, using the IEEE 33-node example.

[0093] Table 6. Number of operations and time complexity

[0094] Comparison items Number of operations Number of IEEE 33-node computations Time complexity DC power flow algorithm <![CDATA[3N 2 +16N-4]]> 3791 <![CDATA[O(N 2 )]]> This invention is a semi-AC tidal flow algorithm. <![CDATA[4N 2 +16N-3]]> 4881 <![CDATA[O(N 2 )]]> AC power flow algorithm <![CDATA[24N 3 +21N 2 -43N+117]]> 884055 <![CDATA[O(N 3 )]]>

[0095] In Table 6, N represents the number of nodes in the power distribution system; the number of operations performed by the AC power flow algorithm is calculated based on an average of 3 iterations.

[0096] Table 6 shows the number of operations performed using the IEEE 33-node example. It is evident that the computational complexity of the semi-AC power flow algorithm is significantly lower than that of the AC power flow algorithm, similar to that of the DC power flow algorithm, and its time complexity is quadratic O(N). 2 ), compared to the cubic order O(N) of the AC power flow algorithm. 3 The time complexity is reduced by one order, thus the semi-AC power flow algorithm proposed in this invention has higher computational efficiency.

[0097] Although the present invention has been described above in conjunction with the accompanying drawings, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many modifications under the guidance of the present invention without departing from the spirit of the present invention, and these modifications are all within the protection scope of the present invention.

Claims

1. A method for calculating the semi-AC power flow of a distribution network considering reactive power distribution and network losses, characterized in that, Includes the following steps: Step 1) Input the power distribution system operating parameters: Determine the topology of the power distribution system, the unit resistance r0 of the conductors, the unit reactance x0 of the conductors, and the maximum current carrying capacity I of the conductors. max The resistance r of each branch in the power distribution system ij and reactance x ij Voltage level U of the power distribution system N The active load P connected to each node i and reactive load Q i and the types of each node; Step 2) Establish an N-row, 1-column active power matrix P and an N-row, 1-column reactive power matrix Q: N is the number of nodes, and N is a positive integer greater than or equal to 2; If the i-th node is not a slack node, 1≤i≤N, then the element P in the i-th row and 1-th column of the active power matrix P is... i1 =-P i The element Q in the i-th row and 1-th column of the reactive power matrix Q i1 =-Q i ; If the i-th node is a saturated node, 1≤i≤N, then the element P in the i-th row and 1-th column of the active power matrix P is... i1 =0, the element Q in the i-th row and 1-th column of the reactive power matrix Q. i1 =0; Step 3) Construct an N x N nodal admittance matrix Y: In the nodal admittance matrix Y, For element Y in row i and column j ij , 1≤i≤N, 1≤j≤N, i≠j; If there is no connection between node i and node j, then Y ij =0; if there is a connection between node i and node j, then Y ij =-1 / x ij ; For element Y in row i and column i ii 1≤i≤N, if node i is a balanced node, then Y ii =0; If node i is not a balanced node, and node i is connected to the other n nodes, 1≤n≤N-1, then Y ii =n / x ij ; Step 4) Calculate the voltage phase angle vector θ and the voltage virtual phase angle vector θ Q : (1) (2) According to equation (1) According to equation (2). ; Step 5) Calculate the active power distribution of the branch without considering network losses. Reactive power distribution of branches ,include: First, establish the distribution equations (3) and (4) for the power flow between node i and node j: (3) (4) In equations (3) and (4): P ij-AC Q ij-AC These represent the AC active power distribution and AC reactive power distribution at the beginning of the branch circuit, respectively; G ij and B ij These represent the mutual conductance and mutual susceptance between nodes i and j, respectively; U i and U j Let θ represent the voltage magnitudes at nodes i and j, respectively; ij The voltage phase angle difference between nodes i and j; Secondly, it is approximated that the voltage U at each node i ≈1.00 pu, inter-node voltage phase angle difference θ ij Very small, has sin θ ij ≈θ ij ≈θ i -θ j cos θ ij ≈1, neglecting branch resistance r ij For all branches connected to the ground, simplified equation (3) becomes equation (5), which is used to calculate the active power distribution of branches without considering network losses. : (5) In equation (5): θ i and θ j Let be the voltage phase angles of nodes i and j, respectively; Equation (6) is obtained by simplifying equation (4), which is used to calculate the reactive power distribution of branches without considering network losses. : (6) In formula (6): and These are the virtual phase angles of the voltages at nodes i and j, respectively. Step 6) Calculate the active power loss ΔP of each branch. ij and reactive power loss ΔQ ij : Based on the active power distribution of each branch without considering network losses Divide by node voltage U i The current I flowing through each branch is obtained. ij : (7) Calculate the active power loss ΔP of each branch. ij and reactive power loss ΔQ ij : (8) Step 7) Calculate the active power distribution P of each branch considering network losses. ij and reactive power distribution Q ij : (9)。

Citation Information

Patent Citations

  • Direct current load flow computing method with network loss considered

    CN103904643A

  • Method and system for computing distribution network line loss

    CN110212521A