Topology identification and line parameter estimation method and system for power distribution network

Abstract

The invention discloses a topology identification and line parameter estimation method and system of a power distribution network, and relates to the field of network topology identification of the power distribution network, and the method comprises the steps: carrying out the preliminary estimation of an admittance matrix of the power distribution network, and obtaining the initial values of a network topology structure and line parameters; based on the obtained initial values of the network topology structure and the line parameters, an improved Newton-Raphson iteration method is adopted, an admittance matrix and a node voltage phase angle are optimized at the same time, and the topology structure and the line parameters are updated; in the iterative optimization process, abnormal values in an admittance matrix are monitored in real time, and a topological structure is adjusted to eliminate mistakenly recognized lines; according to the method, phase angle information in the power distribution network is not needed, the topological structure can be identified and the line parameters can be estimated by directly utilizing the power and voltage amplitude measurement data, and the method is suitable for a conventional power distribution network and has relatively high calculation efficiency and robustness.

Description

Technical Field

[0001] The present invention relates to the field of distribution network topology identification, and in particular to a distribution network topology identification and line parameter estimation method and system. Background Art

[0002] With increasing fossil fuel consumption and greenhouse gas emissions, environmental degradation and resource scarcity will severely constrain rapid social development. Renewable energy has garnered significant attention in recent years, and associated energy management systems have become increasingly indispensable. The widespread penetration of new players in distribution networks, such as renewable energy, energy storage, and controllable loads, presents both opportunities and challenges for distribution network operation and optimization. For example, this operational optimization will improve distribution network safety and reliability. However, the widespread integration of renewable energy also increases the complexity of distribution network operation and analysis. For example, system topology and line parameters are crucial for assessing distribution network status. However, the time-varying nature of renewable energy makes these topological and line parameter information difficult to obtain. Distribution networks have fewer monitoring devices than transmission networks, and the vast majority of measurement equipment is automated metering (AMI).

[0003] Some existing methods rely on phase angle information for topology identification, requiring the installation of additional high-precision phase angle measurement equipment, increasing costs and operational complexity. Furthermore, phase angle measurements are susceptible to interference, affecting data accuracy and thus reducing the reliability of topology identification results. When using traditional measurement equipment (such as non-smart meters) for topology identification, the low data sampling frequency and limited accuracy make it difficult to capture the rapid topological changes in distribution networks caused by the time-varying nature of renewable energy, resulting in a lag in topology identification. Some existing methods rely on a large number of distributed measurement devices to obtain rich data to estimate line parameters, which is difficult to implement in distribution networks with limited testing equipment. Furthermore, the installation and maintenance costs of numerous measurement devices are high, making them less cost-effective. Some methods construct complex mathematical models to estimate line parameters, which are computationally intensive, time-consuming, and inefficient. Given the time-varying nature of renewable energy, it is difficult to respond quickly and accurately estimate line parameters in real time. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is: how to identify the topology and accurately estimate the line parameters only through the data measured by the AMI.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides a method for topology identification and line parameter estimation of a distribution network, comprising:

[0008] Make a preliminary estimate of the admittance matrix of the distribution network and obtain the initial values of the network topology and line parameters;

[0009] Based on the obtained initial values of network topology and line parameters, an improved Newton-Raphson iterative method is used to simultaneously optimize the admittance matrix and node voltage phase angles, and update the topology and line parameters.

[0010] During the iterative optimization process, outliers in the admittance matrix are monitored in real time, and the topology is adjusted to eliminate misidentified lines.

[0011] As a preferred solution for distribution network topology identification and line parameter estimation, the following methods are proposed:

[0012] The preliminary estimation of the admittance matrix of the distribution network and obtaining the initial values of the network topology and line parameters include:

[0013] Based on the relationship between power and voltage amplitude, a linear regression model is established to preliminarily estimate the admittance matrix of the distribution network;

[0014] The power flow equation is simplified based on multiple observations including power, voltage amplitude and reactive power, and the conductance matrix and susceptance matrix are estimated by linear regression method.

[0015] As a preferred solution for distribution network topology identification and line parameter estimation, the following methods are proposed:

[0016] The preliminary estimation of the admittance matrix of the distribution network and obtaining the initial values of the network topology and line parameters also includes:

[0017] Based on the estimated conductance matrix and susceptance matrix, branch contribution is introduced to determine the denoising threshold.

[0018] As a preferred solution for distribution network topology identification and line parameter estimation, the following methods are proposed:

[0019] The introducing branch contribution to determine the denoising threshold comprises:

[0020] Calculate the contribution ratio of each branch to the diagonal elements of the admittance matrix of the node to which it is connected. If the contribution of a branch is lower than the set lower threshold, it is considered that the branch may be misidentified and will be removed.

[0021] The denoising process is transformed into a constrained linear least squares problem and solved using an iterative denoising algorithm.

[0022] As a preferred solution for distribution network topology identification and line parameter estimation, the following methods are proposed:

[0023] The iterative denoising algorithm is used to solve the problem, which includes:

[0024] The branch parameters whose contribution values are less than the threshold are set to zero, and linear regression is performed on each line. The matrix data is updated and the matrix symmetry is ensured. The iterative solution process is repeated until the constraints are met, and the line parameters are obtained through the denoised matrix.

[0025] As a preferred solution for distribution network topology identification and line parameter estimation, the following methods are proposed:

[0026] The iterative optimization of the admittance matrix and the node voltage phase angle based on the obtained network topology and initial values of the line parameters, and the updating of the topology and line parameters include:

[0027] An improved Newton-Raphson model based on multiple sets of measurement data is established, and the admittance matrix is solved using the generalized inverse matrix.

[0028] As a preferred solution for distribution network topology identification and line parameter estimation, the following methods are proposed:

[0029] The iterative optimization of the admittance matrix and the node voltage phase angle based on the obtained network topology structure and initial values of the line parameters, and updating the topology structure and line parameters further includes:

[0030] Pseudo power flow calculation is performed at the beginning and end of each iteration. Buses other than the reference bus are regarded as PQ nodes. The admittance parameters estimated by preliminary identification and the measured power data are used to calculate the missing voltage phase angle, and the calculated voltage phase angle is used to replace the original value for subsequent iterations.

[0031] In a second aspect, an embodiment of the present invention provides a system for topology identification and line parameter estimation of a distribution network, including:

[0032] The initial value acquisition module is used to make a preliminary estimate of the admittance matrix of the distribution network and obtain the initial values of the network topology and line parameters;

[0033] An iterative optimization module is used to simultaneously optimize the admittance matrix and node voltage phase angle based on the obtained network topology and initial values of line parameters, and update the topology and line parameters using an improved Newton-Raphson iterative method;

[0034] The monitoring and adjustment module is used to monitor the abnormal values in the admittance matrix in real time during the iterative optimization process and adjust the topology structure to eliminate the incorrectly identified lines.

[0035] In a third aspect, an embodiment of the present invention provides a computing device, including:

[0036] memory and processor;

[0037] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the one or more programs are executed by the one or more processors, the one or more processors implement the topology identification and line parameter estimation method of the distribution network as described in any embodiment of the present invention.

[0038] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the method for topology identification and line parameter estimation of the distribution network.

[0039] The beneficial effects of the present invention are as follows: the present invention does not require phase angle information in the distribution network, and can directly use power and voltage amplitude measurement data to identify the topology structure and estimate line parameters. This method is applicable to conventional distribution networks and has high computational efficiency and robustness; the admittance matrix is preliminarily estimated through linear regression, and symmetry and iterative denoising are used to effectively remove measurement and calculation errors, identify possible line topologies, and obtain accurate initial values of line parameters; an improved Newton-Raphson model based on multiple sets of measurement data is combined with a generalized inverse matrix solution to reduce errors caused by erroneous measurements, further optimize the admittance matrix and node voltage phase angle, and make the identification of topology structure and line parameters more accurate; a pseudo-power flow calculation method is introduced to correct the voltage phase angle at the beginning of the iteration and in each iteration. Taking into account the influence of high-order terms of the first-order approximation of the Newton method when the deviation is large, and the nonlinear characteristics of the phase angle on power, the voltage phase angle is corrected through pseudo-current calculation, which improves the convergence speed and robustness of the iterative process and avoids falling into the local optimal solution; real-time monitoring is carried out during the iteration process, and the admittance matrix with abnormal values is corrected. When the iteration is close to termination and certain conditions are met, the topology structure is dynamically adjusted and inaccurate branches are removed to ensure that the final topology structure and line parameters are more in line with the actual situation; through error checking and iterative optimization, the admittance matrix, voltage phase angle and topology structure are continuously adjusted until the error is less than the allowable range, ensuring the reliability and stability of the entire model and identification results. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0041] Figure 1 This is an overall flow chart of the method for topology identification and line parameter estimation of a distribution network according to the present invention;

[0042] Figure 2It is a topology diagram of the IEEE33 test system in a simulation example of the method for topology identification and line parameter estimation of a distribution network according to the present invention;

[0043] Figure 3 is a 24-hour load variation curve of the 33rd node in the simulation example of the method for topology identification and line parameter estimation of the distribution network according to the present invention;

[0044] Figure 4 The noise-reduced topology identification and line parameter estimation method of the distribution network described in the present invention is obtained by basic identification in the simulation example.

[0045] Figure 5 The noise-reduced topology of the distribution network obtained by basic identification in the simulation example of the method for topological identification and line parameter estimation of the present invention is

[0046] Figure 6 The topology identification and line parameter estimation method of the distribution network described in the present invention is obtained by basic identification without noise reduction.

[0047] Figure 7 The topology identification and line parameter estimation method of the distribution network described in the present invention is obtained by basic identification without noise reduction.

[0048] Figure 8 is the number of identified erroneous branches in each iteration of the simulation example of the method for topology identification and line parameter estimation of the distribution network according to the present invention;

[0049] Figure 9 1 is a schematic diagram of changes in the recovery voltage angle at different times in a simulation example of the method for topology identification and line parameter estimation of a distribution network according to the present invention;

[0050] Figure 10 It is a schematic diagram of the changes in the errors of the admittance parameter g and the susceptance parameter b under different measurement error conditions in a simulation example of the distribution network topology identification and line parameter estimation method described in the present invention. DETAILED DESCRIPTION

[0051] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0052] Example 1, reference Figure 1 , which is the first embodiment of the present invention, provides a method for topology identification and line parameter estimation of a distribution network, comprising:

[0053] S1: Make a preliminary estimate of the admittance matrix of the distribution network and obtain the initial values of the network topology and line parameters;

[0054] S2: Based on the obtained initial values of the network topology and line parameters, the improved Newton-Raphson iterative method is used to simultaneously optimize the admittance matrix and node voltage phase angles to update the topology and line parameters;

[0055] S3: During the iterative optimization process, outliers in the admittance matrix are monitored in real time, and the topology is adjusted to eliminate misidentified lines.

[0056] It should be noted that, through S1-S3, this embodiment selects appropriate input parameters: the preliminary estimates of the admittance matrix g, b, and θ obtained from the basic identification step, as well as the node voltage phase angle θ obtained from the pseudo-power flow calculation, are substituted as input parameters into the fine-grained identification algorithm model. A modified Newton-Raphson iteration method is used to establish a set of constraint equations based on multiple sets of measurement data (active power P, reactive power Q, and voltage amplitude V), thereby improving the robustness of the model. To address the problem of the model being unable to be directly solved due to the greater number of measurement data sets than unknown variables, the generalized inverse matrix technique is used to calculate the changes in the admittance matrix ΔG, ΔB and the voltage phase angle Δθ. The calculated changes are then used to update the admittance matrix and voltage phase angle. Pseudo-power flow calculation is introduced: To improve iterative convergence, the pseudo-power flow is recalculated in each iteration based on the updated admittance matrix and voltage phase angle, and the voltage phase angle is updated using this result. Error checking and topology correction iterations are performed: First, the error between the updated admittance matrix and voltage phase angle and the measured data is calculated. If the error is larger than the preset allowable range, it indicates that there is a topology identification error, and topology correction is performed: random perturbations are added to the elements with outliers in the admittance matrix to escape the local optimum; when the iteration converges, the admittance matrix is traversed and the lines corresponding to the elements with too small values are eliminated; after correction, iterative optimization is performed again until the error is less than the allowable range.

[0057] Example 2, reference Figure 1 , which is an embodiment of the present invention, provides a method for topology identification and line parameter estimation of a distribution network based on the previous embodiment, including:

[0058] In this embodiment, the preliminary estimation of the admittance matrix of the distribution network and the acquisition of the initial values of the network topology and line parameters in the above step S1 include:

[0059] Based on the relationship between power and voltage amplitude, a linear regression model is established to make a preliminary estimate of the admittance matrix of the distribution network. The specific process is as follows:

[0060] According to the power equation in polar coordinates, it is as follows:

[0061]

[0062] Among them, p i and q i Represent the active and reactive power of node i respectively; ν i and ν j is the voltage amplitude of nodes i and j; G ij and B ij Expressed as conductance and susceptance between nodes i and j; cosθ ij and sinθ ij are the cosine and sine values of the voltage phase angle difference between nodes i and j;

[0063] Rewrite the above formula into matrix form as follows:

[0064]

[0065] Among them, ν i is the node voltage amplitude without phase angle information; [p / v] = [p1 / v1…p n / v n ] T , [q / v]=[q1 / v1…q n / v n ] T and [v] = [v1…v n ] T is a vector containing relevant information of each node;

[0066] It should be noted that in the actual distribution network, G ij and B ij The values of G are of the same order of magnitude (such as IEEE 33 and 123 feeder nodes). When node i and node j are directly connected, G ij and B ij are all non-zero. And since G ij # and B ij # It's G ij and B ij A linear combination of G ij # and B ij # When nodes i and j are directly connected, G ij # and Bij # are all non-zero items. That is, G ij # and B ij # With G ij and B ij Reflects the same topology information of the system.

[0067] Since in actual operation |θ ij | Usually less than 5° or 0.1rad. Therefore, in the embodiment, sinθ≈θ, cosθ≈1 are approximated accordingly. Thus, G ij # and B ij # It can be rewritten as follows:

[0068]

[0069] It should be noted that even if |θ ij | Very large, such as ±15°, assuming G ij =-B ij In the case of , the error between the theoretical value and the estimated value is also within 30%, and the result within this error range is still a good initial value for the subsequent steps.

[0070] Suppose there are M0 observations containing P, V, and Q, and use linear regression to estimate and Therefore, the power flow equation is expressed as:

[0071]

[0072] in,

[0073] By using the linear regression method, we estimate and

[0074]

[0075] At the same time, due to G ij and B ij In actual power systems, the matrix is symmetric and the symmetry of the matrix is maintained by the following method, where S represents symmetry:

[0076]

[0077] The symmetry of the admittance matrix is used to denoise the estimated results and identify possible line topologies. The specific steps are as follows:

[0078] It should be noted that due to the non-negligible errors in the measurement and linear regression process, the obtained and There may be errors. One method of denoising is to remove items that are smaller than a fixed value, namely the denoising threshold, and introduce the concept of branch contribution to determine the denoising threshold.

[0079] Calculate the contribution ratio γ(i,j) of each branch to the diagonal elements of the admittance matrix of the node to which it is connected. The calculation formula is as follows:

[0080]

[0081] According to the relevant knowledge of circuit principles, the diagonal elements of the admittance matrix are equal to the sum of the negative values of the admittances of the branches directly connected to the node. Therefore, if the admittance of a branch accounts for a small proportion of the total admittance of its corresponding node, the branch can be considered to be misidentified; however, due to the lack of accurate information about the branch admittance and its conductivity parameters g and b, it is difficult to accurately determine the upper limit values of g and b. In contrast, for a certain node, the actual number of branches connected to it is usually much smaller than the theoretical maximum possible number of connections (n-1), and the lower limit of the branch contribution γ is set to γ. Top Set to |G ii |, in order to filter out unconnected branches. |G ii The definition of | is as follows:

[0082]

[0083] If the contribution of a branch to the node it connects to is lower than the threshold γ Top , then remove the branch, that is, its corresponding G ij Set to zero.

[0084] The denoising process is converted into a constrained linear least squares problem as follows:

[0085]

[0086] in, represents the estimated node conductance matrix; γ Top represents the lower limit of branch contribution; γ (i,j) Indicates the contribution of each branch.

[0087] However, due to the nonlinear nature of this problem, it is not a simple constrained least squares problem. This embodiment proposes an iterative denoising algorithm to solve it:

[0088] Calculate the contribution of each branch γ (i,j) And the contribution value is less than γ Top Branch parameters and Set to zero. Perform linear regression on each line as follows:

[0089] [P / V (i) ]=[G #(i) ][V]

[0090] Among them, [P / V (i) ] and [G #(i) ] is the [P / V] of the line whose contribution value in the i-th row is greater than the threshold and data.

[0091] Use [G #(i) ] to update The relevant data in the formula is used to ensure that Symmetry of:

[0092]

[0093] Repeat the above iterative process until Satisfy the constraints; after denoising and To obtain the line parameters g,b.

[0094] In this embodiment, in step S2, based on the obtained initial values of the network topology and line parameters, iteratively optimizing the admittance matrix and the node voltage phase angle, and updating the topology and line parameters include:

[0095] An improved Newton-Raphson model based on multiple sets of measurement data is established, and the admittance matrix is solved using the generalized inverse matrix. The specific steps are as follows:

[0096] Applying the Newton method to the power flow equation yields the following:

[0097]

[0098] Where p=[p1…p n ] T , q=[q1…q n ] T , θ=[θ1…θ n ] T They are respectively the active, reactive and voltage phase angle vectors of n nodes; g=[g1…g m ] T , b=[b1…b m ] T are the conductance and susceptance vectors of the m branches respectively; and are the partial derivatives of the node active power with respect to g, b and θ respectively; and are the partial derivatives of node reactive power with respect to g, b and θ respectively;

[0099] Due to measurement errors and the number of variables (2m+n+1) being greater than the number of constraints (2n), only a set of [Δg, Δb, Δθ] T Therefore, this embodiment uses multiple sets of sampling [Δg, Δb, Δθ] T In order to reduce the error caused by erroneous measurement, the new solution equation is as follows:

[0100]

[0101] Where P = [p 1 ;…;p M ] T , Q=[q 1 ;…;q M ] T , Θ=[θ 1 ;…;θ M ].

[0102] However, the Jacobian matrix constructed by this equation is not a square matrix, so this embodiment uses the generalized inverse method to solve and obtain [Δg, Δb, Δθ] T The only optimal approximate solution is as follows:

[0103]

[0104] in, represents the generalized inverse of a matrix;

[0105] A pseudo-power flow calculation method is introduced to correct the voltage phase angle at the beginning of the iteration and in each iteration, thereby improving the iterative convergence. The specific steps are as follows:

[0106] The Newton method is based on the first-order approximation of Taylor expansion, as shown in the following formula:

[0107]

[0108] When there is a large deviation between the estimated results and the actual results, the impact of high-order terms cannot be ignored.

[0109] Assume that x represents g, b and y represents θ, and f is used to describe the mapping relationship between p, q and g, b. Then, the power flow equation based on the node power flow equation is rewritten into the power flow equation expressed in line parameters. The specific equation is as follows:

[0110]

[0111] Among them, w iJ Represents the connection relationship between branch J and node i. If w iJ =1 means that branch J is connected to node i, otherwise wiJ =0; Node k represents the node opposite to node i in branch J;

[0112] According to the mapping relationship above, we can see that p and q can be linearly represented by g and b, but cannot be linearly represented by θ. This shows that there is a linear relationship between the change in power in the system and the admittance parameters g and b, while the effect of phase angle θ on power is nonlinear, that is:

[0113]

[0114] Substituting the above formula into the Taylor expansion formula, we can get the following formula:

[0115]

[0116] It can be found that when Δx and Δy are not small enough, high-order terms (2≤k) will introduce significant errors in the Newton method, especially when there are a large number of variables. This is mainly due to the nonlinear characteristics of the power flow equation, which leads to the calculated [Δg, Δb, Δθ] T This method fails to accurately reflect the true iterative descent direction. Specifically, g and b are constrained by 2 × M n equations, while only 2 n equations are directly related to each θ because the voltage phase angle is independent across different data sets. Therefore, errors in the data or calculations significantly impact the accuracy of θ. Therefore, this embodiment proposes using pseudo power flow calculations to correct θ after each iteration.

[0117] Specifically, the pseudo power flow calculation involves treating all buses except the reference bus as PQ nodes and calculating the missing voltage phase angle using g and b estimated from the initial identification and p and q measured. The original θ is then replaced with the θ obtained from the pseudo power flow calculation in each subsequent iteration.

[0118] It should be noted that, through such processing, the obtained θ is more accurate than the unprocessed θ, and the convergence speed and robustness of the Newton method iterative process are also improved.

[0119] In the embodiment of the present application, in the iterative optimization process in step S3, real-time monitoring of outliers in the admittance matrix and adjusting the topology to eliminate misidentified lines include:

[0120] Eliminate outliers in the admittance matrix during the iteration process, including: During the iteration process of the improved Newton method, negative g or positive b values may appear. This is usually because the model falls into a local optimal solution, which is impossible in the actual distribution network. In order to avoid this situation, it is processed by introducing a stronger disturbance into the model. That is, the wrong g is modified to a small positive random value, and the wrong b is modified to a small negative random value. The absolute values of these small random disturbances are sampled from the uniform distribution (0, ξ). These randomly generated values play an appropriate incentive role, so that the iterative process can continue to converge to a better solution.

[0121] The topology is continuously adjusted during the iteration process, including: During the iteration process, branches with sufficiently small g values are considered to be incorrectly identified branches. Therefore, when the iteration is close to termination - that is, the deviation is less than a preset threshold ζ (called the topology modification threshold, whose value should be greater than the convergence threshold φ) - each branch is scanned to identify these branches. Specifically, if the g value of a branch is less than a threshold, the branch is considered inaccurate and is removed, that is, the corresponding g and b values are permanently set to zero. After the topology modification is completed, a new iteration will be started based on the updated topology. In this way, the topology can be dynamically corrected during the iteration process.

[0122] Example 3. The above is a schematic scheme of the distribution network topology identification and line parameter estimation method of this embodiment. It should be noted that the technical scheme of the distribution network topology identification and line parameter estimation system and the technical scheme of the distribution network topology identification and line parameter estimation method described above are based on the same concept. For details not described in detail in the technical scheme of the distribution network topology identification and line parameter estimation system in this embodiment, please refer to the description of the technical scheme of the distribution network topology identification and line parameter estimation method described above.

[0123] This embodiment further provides a system based on a distribution network topology identification and line parameter estimation method, including:

[0124] The initial value acquisition module is used to make a preliminary estimate of the admittance matrix of the distribution network and obtain the initial values of the network topology and line parameters;

[0125] An iterative optimization module is used to simultaneously optimize the admittance matrix and node voltage phase angle based on the obtained network topology and initial values of line parameters, and update the topology and line parameters using an improved Newton-Raphson iterative method;

[0126] The monitoring and adjustment module is used to monitor the abnormal values in the admittance matrix in real time during the iterative optimization process and adjust the topology structure to eliminate the incorrectly identified lines.

[0127] This embodiment further provides a computing device applicable to a method for topology identification and line parameter estimation of a distribution network, including:

[0128] Memory and processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the topology identification and line parameter estimation method of the distribution network proposed in the above embodiment.

[0129] This embodiment further provides a storage medium having a computer program stored thereon. When the program is executed by a processor, the method for topology identification and line parameter estimation of the distribution network proposed in the above embodiment is implemented.

[0130] The storage medium proposed in this embodiment and the distribution network topology identification and line parameter estimation method proposed in the above embodiment belong to the same inventive concept. For technical details not described in detail in this embodiment, please refer to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0131] Example 4, with reference to Figure 2-Figure 10 , which is an embodiment of the present invention, provides a method for topology identification and line parameter estimation of a distribution network. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0132] This embodiment will take the IEEE 33-node test system as an example and test the proposed model through two examples under different conditions. In the first case, the historical power p, q and voltage amplitude v data are used to identify the preliminary topology and parameters through the basic identification model. In the second case, based on the preliminary identification data obtained in the first case and the entire historical data set, the corresponding topology identification of the distribution network is performed in real time. Among them, the line topology of the example is as follows Figure 2 shown.

[0133] In the first example, the EEE 33-node test system is a 12.66kV distribution network system with 32 transmission lines and 5 tie lines. In this test example, this embodiment uses 24 hours of historical data with a data sampling interval of 12 minutes, that is, a total of 5×24 historical data, to preliminarily identify the topology and line parameters of the 33-node distribution network. In addition, a 2% external error is introduced to p and q, and a 0.01% error is introduced to the voltage amplitude v to simulate the influence of measurement errors and abnormal data. In the preliminary identification, this example sets γ Top =5%. Figure 3 Shown is the 24-hour load variation curve of the 33rd node.

[0134] Figure 4 and Figure 5 The image is displayed after noise reduction processing. Conductivity matrix and Susceptance matrix. Figure 6 and Figure 7 The image is displayed without noise reduction. Conductivity matrix and Susceptance matrix. Compared with the admittance and susceptance matrices before noise reduction, the network topology has been largely identified through preliminary identification. The resulting topology includes 39 possible branches, including 33 distribution trunk lines and tie lines in the IEEE 33-node distribution system. However, it also includes six incorrectly identified branches: 2-20, 3-6, 8-25, 9-12, 13-16, and 25-28.

[0135] After completing the basic identification, the average absolute errors of the admittance parameters g and b are 43.42% and 37.06% respectively. This result confirms that G ij # With G ij The assumption that the gap is smaller.

[0136] The second example is a fine identification based on the first example. The three important parameters ξ, ζ, and φ are set to 0.05, 0.01, and 1×10 -8 Since the Newton method has the characteristic of quadratic convergence, it only takes a few iterations to reach convergence. For example, in this example, only nine iterations are required to complete convergence, which further illustrates the fast convergence characteristics of the proposed method. In addition, the method shows strong robustness in dealing with erroneous topologies, such as Figure 8 As shown, six branches are corrected in only three iterations. Figure 9 The voltage angles recovered by the test system at 0:00, 6:00, 12:00, and 18:00 are shown. In this example, the average deviation of the voltage angle is 0.05°. Because the proposed method can provide a high-precision estimate of the admittance parameters, the voltage angle estimate is also highly accurate.

[0137] In addition, a sensitivity analysis of the impact of measurement error on estimation accuracy was conducted. Figure 10 As shown in the figure, the proposed method can maintain a low estimation error while maintaining high measurement accuracy. This result can provide an important reference for optimizing the operation of distribution networks.

[0138] In summary, this method can maintain high-precision analysis even with large amounts of measurement data. Distribution network operators can use historical data to make preliminary estimates of line parameters and accurately identify lines that are operating under different historical operating conditions.

[0139] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for topology identification and line parameter estimation of a distribution network, characterized in that: include: Make a preliminary estimate of the admittance matrix of the distribution network and obtain the initial values of the network topology and line parameters; Based on the obtained initial values of network topology and line parameters, an improved Newton-Raphson iterative method is used to simultaneously optimize the admittance matrix and node voltage phase angles, and update the topology and line parameters. During the iterative optimization process, outliers in the admittance matrix are monitored in real time, and the topology is adjusted to eliminate misidentified lines.

2. The method for topology identification and line parameter estimation of a distribution network according to claim 1, wherein: The preliminary estimation of the admittance matrix of the distribution network and obtaining the initial values of the network topology and line parameters include: Based on the relationship between power and voltage amplitude, a linear regression model is established to preliminarily estimate the admittance matrix of the distribution network; The power flow equation is simplified based on multiple observations including power, voltage amplitude and reactive power, and the conductance matrix and susceptance matrix are estimated by linear regression method.

3. The method for topology identification and line parameter estimation of a distribution network according to claim 2, wherein: The preliminary estimation of the admittance matrix of the distribution network and obtaining the initial values of the network topology and line parameters also includes: Based on the estimated conductance matrix and susceptance matrix, branch contribution is introduced to determine the denoising threshold.

4. The method for topology identification and line parameter estimation of a distribution network according to claim 3, wherein: The introducing branch contribution to determine the denoising threshold comprises: Calculate the contribution ratio of each branch to the diagonal elements of the admittance matrix of the node to which it is connected. If the contribution of a branch is lower than the set lower threshold, it is considered that the branch may be misidentified and will be removed. The denoising process is transformed into a constrained linear least squares problem and solved using an iterative denoising algorithm.

5. The method for topology identification and line parameter estimation of a distribution network according to claim 4, characterized in that: The iterative denoising algorithm is used to solve the problem, which includes: The branch parameters whose contribution values are less than the threshold are set to zero, and linear regression is performed on each line. The matrix data is updated and the matrix symmetry is ensured. The iterative solution process is repeated until the constraints are met, and the line parameters are obtained through the denoised matrix.

6. The method for topology identification and line parameter estimation of a distribution network according to claim 5, wherein: The iterative optimization of the admittance matrix and the node voltage phase angle based on the obtained network topology and initial values of the line parameters, and the updating of the topology and line parameters include: An improved Newton-Raphson model based on multiple sets of measurement data is established, and the admittance matrix is solved using the generalized inverse matrix.

7. The method for topology identification and line parameter estimation of a distribution network according to claim 6, wherein: The iterative optimization of the admittance matrix and the node voltage phase angle based on the obtained network topology structure and initial values of the line parameters, and updating the topology structure and line parameters further includes: Pseudo power flow calculation is performed at the beginning and end of each iteration. Buses other than the reference bus are regarded as PQ nodes. The admittance parameters estimated by preliminary identification and the measured power data are used to calculate the missing voltage phase angle, and the calculated voltage phase angle is used to replace the original value for subsequent iterations.

8. A distribution network topology identification and line parameter estimation system, applying the method according to any one of claims 1 to 7, characterized in that: include: The initial value acquisition module is used to make a preliminary estimate of the admittance matrix of the distribution network and obtain the initial values of the network topology and line parameters; An iterative optimization module is used to simultaneously optimize the admittance matrix and node voltage phase angle based on the obtained network topology and initial values of line parameters, and update the topology and line parameters using an improved Newton-Raphson iterative method; The monitoring and adjustment module is used to monitor the abnormal values in the admittance matrix in real time during the iterative optimization process and adjust the topology structure to eliminate the incorrectly identified lines.

9. A computing device comprising: memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions, when executed by a processor, implement the steps of the method according to any one of claims 1 to 7.

Images