A simplified calculation method and system for urban distribution network based on multi-level equivalent modeling

Abstract

The present invention discloses a simplified calculation method and system for an urban distribution network based on multi-level equivalent modeling, which relates to the technical field of digital simulation of distribution networks. The method comprises: defining the boundary of the distribution network based on distribution network data; generating an equivalent model based on the complex hierarchy, equivalent principle and boundary of the distribution network; adjusting the electrical characteristics and grid structure of the equivalent model to generate the distribution network model; the specific steps of generating the equivalent model comprise: obtaining sensitive areas and several partitions of the distribution network; constructing a single equivalent component; obtaining a high level based on a Thevenin equivalent circuit model, an equivalent principle and a boundary, obtaining a middle level based on a Ward equivalent model, an equivalent principle and a boundary, obtaining a low level based on a three-phase admittance matrix, an equivalent principle and a boundary, and generating an equivalent model based on sensitive areas, partitions, a single equivalent component, a high level, a middle level and a low level, thereby solving the problem of poor model accuracy due to a large number of nodes in the existing distribution network modeling method.

Description

Technical Field

[0001] The present invention relates to the technical field of digital simulation of distribution networks, and in particular to a simplified calculation method and system for urban distribution networks based on multi-level equivalent modeling. Background Art

[0002] The urban distribution network is an important part of the urban power system, responsible for transmitting electricity from the substation to the end user. Traditional distribution network modeling methods often require detailed modeling of all nodes and connections from the high-voltage substation to the user meter, and the number of nodes often exceeds 104 levels. Although such methods are accurate, they also bring significant performance bottlenecks, especially when performing power flow calculations. Due to the large number of nodes, the power flow calculation time of traditional modeling methods is often too long, and a single calculation may exceed 5 minutes, which seriously affects the real-time operation and control of the distribution network. At present, equivalent modeling methods are usually used to simplify the model, such as REI equivalence. Although it can simplify the model, when the substation contains a high proportion of photovoltaic power sources, the model accuracy is poor due to the characteristics of the photovoltaic power source itself, the characteristics of distributed energy access, and the limitations of the modeling method itself. The specific reasons are:

[0003] 1) The output power of photovoltaic power sources is affected by natural factors and is intermittent and fluctuating. The MPPT algorithms they use also have differences and uncertainties. Traditional equivalent modeling methods have difficulty accurately capturing these characteristics and are often based on simplified assumptions, resulting in accumulated deviations.

[0004] 2) Distributed energy sources such as photovoltaics are distributed and connected to the distribution network in large numbers at different locations. They interact with traditional loads, changing network power flow and voltage characteristics. Traditional equivalent modeling methods suffer from oversimplification and unreasonable assumptions when dealing with this complex connection and source-load interaction, and are unable to accurately reflect actual operating conditions.

[0005] 3) Traditional modeling methods have their own limitations, such as model assumptions that do not match reality, difficulty in determining equivalent parameters, and insufficient dynamic characteristics. They are difficult to truly reflect the actual operating characteristics of the distribution network, especially in terms of dynamic process simulation, which results in reduced accuracy and increased errors. Summary of the Invention

[0006] In order to solve the problem of poor model accuracy caused by the large number of nodes and low computational efficiency in existing distribution network modeling methods, the present invention provides a simplified calculation method for urban distribution networks based on multi-level equivalent modeling, the method comprising: obtaining distribution network data, defining the boundary of the distribution network based on the distribution network data; generating an equivalent model based on the complex hierarchy of the distribution network, the equivalent principle and the boundary; adjusting the electrical characteristics and grid structure of the equivalent model to generate a distribution network model; the complex hierarchy includes high level, middle level and low level, and the equivalent principle includes node equivalent principle and line equivalent principle;

[0007] The specific steps of generating the equivalent model include: obtaining sensitive areas and several partitions of the distribution network; constructing a single equivalent component; obtaining the high level based on the Thevenin equivalent circuit model, the equivalent principle and the boundary, obtaining the middle level based on the Ward equivalent model, the equivalent principle and the boundary, obtaining the low level based on the three-phase admittance matrix, the equivalent principle and the boundary, and generating the equivalent model based on the sensitive areas, the partitions, the single equivalent component, the high level, the middle level and the low level.

[0008] This method defines the boundaries of the distribution network, obtains the areas that need detailed modeling through sensitive areas, and divides the distribution network by zoning, which can improve the model accuracy and data processing efficiency, and better adapt to the diverse needs of urban distribution networks; by formulating differentiated equivalent principles and multi-level equivalent modeling to simplify the distribution network model, it is applied to trunk line simulation, greatly reducing the number of nodes and the amount of calculation, reducing the complexity of distribution network flow calculation, significantly improving the calculation speed, meeting the needs of real-time scheduling and control, and optimizing the utilization of power grid resources by simplifying unnecessary details in the distribution network; by simplifying electrical characteristics and geometric structures, simplifying the substations and trunk line voltages of the equivalent model, etc., the simulation calculation process is optimized, the simulation efficiency and real-time response capability of the distribution network are improved, adapting to the needs of large-scale distribution network modeling, and ensuring calculation accuracy and resource utilization.

[0009] This method defines boundary conditions, which may include power supply boundaries, load boundaries, and equipment capacity boundaries. In a complex distribution network, multiple substations serve as power supply boundaries, and the equivalent impedance of their outgoing lines can be calculated using the fixed current of the protection equipment at the beginning of the line. For the load boundaries of multiple substations, the TTU data collected by the substations are used to define them. This refined processing of boundary conditions enables the accurate division of boundaries between different areas when faced with complex power grid structures, providing accurate input data for subsequent equivalent modeling, thereby improving model accuracy and ensuring that the equivalent model can correctly reflect the operating status of the actual power grid.

[0010] This method uses the equivalent principle to simplify the model. This principle can be applied in various ways, including node equivalents, line equivalents, and equipment equivalents. In complex distribution networks, the node equivalent method can be used to simplify the processing of key 10kV feeders, while the line equivalent method can be used to simplify the calculation of substation lines. This flexible selection of equivalent principles enables targeted equivalent modeling based on the characteristics of different parts of the power grid, effectively handling the complex connections between multiple substations and substations, thereby improving model accuracy.

[0011] This method uses a multi-level equivalent modeling approach, dividing the distribution network into high-level, mid-level, and low-level layers. For complex distribution networks with multiple substations and substations, this hierarchical approach can model the structure and characteristics of the grid at different levels separately, effectively improving model accuracy. The high-level model is based on the Thevenin equivalent model, which can equate the entire substation and the complex network in its vicinity to an equivalent power source and impedance, simplifying the computational complexity of the substation. The mid-level model is based on the Ward equivalent model, which can equate the power grid near the substation. In this way, when dealing with multiple substations, each substation and its surrounding complex network can be simplified into a corresponding equivalent model, thereby reducing the overall computational complexity while retaining key electrical characteristics. The low-level model is based on the three-phase admittance matrix, which equates a complex radial network (such as a feeder or substation) to a multi-port admittance matrix, significantly reducing the number of nodes in the overall model. The three-phase admittance matrix (in the form of a 3×3 submatrix) naturally supports phase coupling (such as phase mutual inductance and unbalanced load) and can accurately equate the three-phase unbalanced characteristics of the original network. In addition, single-phase loads and asymmetric lines are common in distribution networks, and traditional single-phase equivalent models have large errors. The three-phase admittance matrix is ​​a more practical choice.

[0012] The three-phase admittance matrix (3-Phase Admittance Matrix) is a mathematical tool that describes the electrical coupling relationships between all nodes in a three-phase power network. Essentially, it is an extension of the single-phase admittance matrix for three-phase systems. It accurately represents the linear relationship between three-phase voltages and currents in matrix form and is a core foundation for analyzing distribution network power flow, short circuits, and stability.

[0013] In summary, this method can perform efficient simulation under limited resources through equivalent modeling and simplified calculation. It can enhance the computational adaptability of large-scale systems and handle large-scale distribution network simulation. It is especially suitable for complex distribution networks containing multiple substations and multiple substations.

[0014] Furthermore, the specific steps of obtaining sensitive areas and several partitions of the distribution network include: obtaining the sensitivity of voltage to power based on the distribution network data, and obtaining the sensitive areas based on the sensitivity; clustering the distribution network data based on preset parameters to obtain the partitions.

[0015] Combining sensitivity analysis with clustering algorithms for multi-level equivalent modeling partitioning can calculate the sensitivity of voltage to power through sensitivity analysis, determine the areas that require detailed modeling (high sensitivity areas), and further optimize the partitioning based on electrical distance or load density through clustering algorithms.

[0016] Furthermore, the specific steps of constructing a single equivalent component include: generating the single equivalent component based on state enumeration and Monte Carlo simulation;

[0017] The calculation method for generating the single equivalent component is:

[0018] ;

[0019] in, represents the equivalent parameter, Indicates status, Represents a collection of states, Indicates status The probability of Indicates status Parameter value.

[0020] A single equivalent component is used to simplify complex power system models and improve analysis efficiency through state probability and parameter calculation.

[0021] Furthermore, the method further comprises: obtaining the number of simulations of the single equivalent component, and obtaining the reliability of the equivalent model based on the number of simulations;

[0022] The calculation formula for obtaining the reliability is:

[0023] ;

[0024] in, Represents reliability, Indicates the number of simulations, Indicates the The fault indication function of the simulation, Represents an integer greater than or equal to 1.

[0025] Conduct reliability analysis on the equivalent model to measure its reliability and ensure the performance of the model.

[0026] Furthermore, the method further comprises: acquiring energy data of distributed energy resources, constructing a dynamic characteristic probability model based on the energy data, and obtaining distributed energy output characteristics based on the dynamic characteristic probability model;

[0027] The calculation formula of the dynamic characteristic probability model is:

[0028] ;

[0029] in, represents the output power of distributed energy resources, represents the mean value of distributed energy resources, represents the variance of distributed energy resources, Represents the probability distribution of distributed energy output power.

[0030] DERs (such as photovoltaics and energy storage) are typically modeled as static negative loads, without considering their dynamic and stochastic nature. This approach models DERs as dynamic loads at the low-level modeling level and introduces a probabilistic model to describe the stochastic nature of their power output (e.g., the impact of weather on photovoltaics). Combined with multi-level modeling, this approach gradually transfers the dynamic characteristics of DERs to the middle and higher levels, addressing the issues of DER fluctuations and complex topologies.

[0031] Furthermore, the method further includes: obtaining a detailed model of the distribution network, comparing the detailed model with the equivalent model to obtain an error value, determining whether the error value exceeds a preset threshold, and if so, adjusting a first model parameter of the equivalent model, obtaining a first model based on the first model parameter, and updating the equivalent model to the first model;

[0032] The calculation formula of the error value is:

[0033] ;

[0034] in, Indicates the error value, and Represent the voltage vectors of the equivalent model and the detailed model respectively.

[0035] Compare the power flow and voltage distribution of the equivalent model and the detailed model. When the error exceeds the preset threshold, adjust the equivalent model parameters and re-verify. Used to verify the difference between the equivalent model and the detailed model, and to verify the accuracy of the model.

[0036] Furthermore, the specific steps of adjusting the electrical characteristics and grid structure of the equivalent model include:

[0037] Simplifying the distributed parameters of the feeders in the equivalent model based on a lumped parameter model, processing the load dynamics of the equivalent model based on an average value model, optimizing the grid structure of the equivalent model based on a minimum spanning tree algorithm, calculating the line impedance of the equivalent model based on a Casson equation, and optimizing the second model parameter of the equivalent model based on a genetic algorithm;

[0038] The specific steps to generate a distribution network model include:

[0039] A three-phase power flow model is obtained based on the Newton-Raphson algorithm, and the distribution network model is obtained based on the adjusted equivalent model and the three-phase power flow model.

[0040] By simplifying the electrical characteristics, a simplified simulation model with fast calculation is constructed; the geometric structure of the power grid is simplified, and a genetic algorithm optimization algorithm is used to accelerate the simulation process and achieve simulation calculation optimization; the three-phase power flow model and the average value model are combined to improve the simulation accuracy and efficiency.

[0041] The three-phase power flow model has the following key uses:

[0042] Handling unbalanced loads: Urban distribution networks often experience three-phase imbalances due to single-phase loads (such as residential and commercial electricity) or distributed energy resources (such as photovoltaics). The three-phase power flow model accurately analyzes the voltage, current, and power flow of each phase, resolving imbalances and ensuring stable grid operation.

[0043] Improved simulation accuracy: Compared to single-phase models, three-phase models take into account interphase coupling and mutual inductance, providing more realistic simulation of grid behavior. This is crucial for power flow calculation, voltage distribution analysis, and fault detection in complex urban power grids.

[0044] Support for real-time dispatch: By quickly calculating three-phase power flows, the model supports real-time dispatch decisions such as load distribution, voltage regulation, and DERs (distributed energy resources) management to optimize grid efficiency.

[0045] Enhanced fault analysis capabilities: The three-phase model can simulate common urban power grid problems such as single-phase grounding faults and unbalanced faults, providing accurate data for fault location and restoration.

[0046] Adapting to complex topologies: Urban distribution networks have complex structures such as multi-branch and ring networks. The three-phase power flow model combined with a simplified equivalent model can efficiently handle these topologies and reduce computational complexity.

[0047] Furthermore, the method further comprises:

[0048] Based on a preset cloud computing simulation framework and objective function, updating the three-phase power flow model and the equivalent model;

[0049] The objective function is:

[0050] ;

[0051] in, represents the objective function, and Both represent weights, Indicates accuracy, Indicates speed.

[0052] Using the cloud computing framework, the model can be updated in real time on the cloud. The objective function comprehensively considers accuracy and speed. Combining the cloud computing framework with the genetic algorithm enhances the real-time and scalability of the model. Therefore, this method is more suitable for complex distribution networks containing multiple substations and multiple substations.

[0053] For complex distribution networks involving multiple substations and distribution areas, the computational workload can still be substantial. Cloud computing-based simulation frameworks can leverage the powerful computing capabilities of cloud computing to update models in real time. This allows complex computational tasks to be distributed across multiple computing nodes in the cloud, rapidly completing simulations of large-scale distribution networks and improving computational efficiency to meet the simulation needs of complex distribution networks.

[0054] Furthermore, the method further comprises:

[0055] Acquire real-time power grid data, construct a state space model, and update the boundary and the equivalent model based on the state space model and the real-time power grid data;

[0056] The calculation formula for updating the equivalent model is:

[0057] ;

[0058] in, Indicates the power adjustment amount, represents the proportional gain, represents the integral gain, Indicates the actual measured power, Indicates analog power;

[0059] The calculation formula of the state space model is:

[0060] ;

[0061] in, express The system status at any moment, express The system status at any moment, express Real-time grid data input at all times, express Output grid data at all times, represents the state matrix, represents the input matrix, represents the output matrix, Represents a straight-through matrix.

[0062] During dynamic simulation, the equivalent model is adjusted based on real-time feedback, and the equivalent method is continuously optimized through a real-time feedback mechanism. A feedback mechanism is established to monitor simulation results in real time, adjusting boundary conditions and the equivalent model based on feedback to ensure that the entire power grid model remains consistent with actual operating conditions. A closed-loop control system is used to adjust the equivalent model based on real-time measurement data. A state-space model is also established to represent the feedback process, using an existing improved LSTM to predict load changes. This improves the forward-looking and adaptable nature of the feedback mechanism and adapts to the complexity and dynamic nature of urban power grids.

[0063] The operating state of distribution networks is dynamic, especially in complex networks involving multiple substations and distribution areas. A dynamic feedback mechanism adjusts the equivalent model based on real-time feedback. For example, when the load at a distribution area changes significantly or the outgoing power of a substation fluctuates, the closed-loop control system and state-space model adjust the equivalent model parameters in real time to ensure that the entire grid model remains consistent with the actual operating state. This dynamic adjustment capability enables the model to adapt to the dynamic changes in various components of a complex distribution network, ensuring the accuracy and reliability of the calculation results.

[0064] The present invention also provides a simplified calculation system for an urban distribution network based on multi-level equivalent modeling, the system further comprising:

[0065] Boundary unit: used for acquiring distribution network data and defining the boundary of the distribution network based on the distribution network data;

[0066] An equivalent model unit: used for generating an equivalent model based on the complex hierarchy of the distribution network, the equivalent principle and the boundary;

[0067] The complexity level includes a high level, a middle level and a low level, and the equivalence principle includes a node equivalence principle and a line equivalence principle;

[0068] The equivalent model unit specifically includes:

[0069] Partition unit: used to obtain sensitive areas and several partitions of the distribution network;

[0070] Component unit: used to construct a single equivalent component;

[0071] Hierarchical unit: used for obtaining the high level based on the Thevenin equivalent circuit model, the equivalence principle and the boundary, obtaining the middle level based on the Ward equivalent model, the equivalence principle and the boundary, and obtaining the low level based on the three-phase admittance matrix, the equivalence principle and the boundary;

[0072] A generating unit, configured to generate the equivalent model based on the sensitive area, the partition, the single equivalent component, the high level, the middle level, and the low level;

[0073] Simplified model unit: used to adjust the electrical characteristics and grid structure of the equivalent model to generate a distribution network model.

[0074] The principles and effects of this system are similar to those of this method, so no further description will be given of this system.

[0075] One or more technical solutions provided by the present invention have at least the following technical effects or advantages:

[0076] 1. This method defines the boundaries of the distribution network, obtains the areas that require detailed modeling through sensitive areas, and divides the distribution network by zoning, which can improve the model accuracy and data processing efficiency, and better adapt to the diverse needs of urban distribution networks; simplifies the distribution network model through the equivalent principle and multi-level equivalent modeling, and applies it to trunk line simulation, greatly reducing the number of nodes and the amount of calculation, reducing the complexity of distribution network flow calculation, and significantly improving the calculation speed to meet the needs of real-time scheduling and control. By simplifying unnecessary details in the distribution network, the utilization rate of the power grid resource can be optimized; by simplifying the electrical characteristics and geometric structure, the simulation calculation process is optimized, the simulation efficiency and real-time response capability of the distribution network are improved, adapting to the needs of large-scale distribution network modeling, and ensuring the calculation accuracy and resource utilization.

[0077] 2. This method adopts a multi-level equivalent modeling method to divide the distribution network into high-level, middle-level and low-level. For complex distribution networks containing multiple substations and multiple substations, this hierarchical approach can model the grid structure and characteristics of different levels separately, effectively improving the accuracy of the model. The high-level is based on the Thevenin equivalent model, which can equate the entire substation and the complex network near it to an equivalent power source and impedance, simplifying the calculation complexity of the substation; the middle-level is based on the Ward equivalent model, which can equate the power grid near the substation. In this way, when dealing with multiple substations, each substation and the surrounding complex network can be simplified into a corresponding equivalent model, thereby reducing the overall calculation complexity while retaining key electrical characteristics.

[0078] 3. Combining sensitivity analysis with clustering algorithms for multi-level equivalent modeling partitioning. Sensitivity analysis can be used to calculate the sensitivity of voltage to power, determine areas requiring detailed modeling (high-sensitivity areas), and further optimize the partitioning based on electrical distance or load density using clustering algorithms.

[0079] 4. In low-level modeling, DERs are modeled as dynamic negative loads, and a probabilistic model is introduced to describe the randomness of their power output (such as the influence of weather on photovoltaics). Combined with multi-level modeling, the dynamic characteristics of DERs are gradually transferred to the middle and high levels to solve the problems of DER fluctuations and complex topologies.

[0080] 5. By simplifying the electrical characteristics, a simplified simulation model with fast calculation is constructed; the geometric structure of the power grid is simplified, and a genetic algorithm optimization algorithm is used to accelerate the simulation process and achieve simulation calculation optimization; the three-phase power flow model and the average value model are combined to improve the simulation accuracy and efficiency.

[0081] 6. Using the cloud computing framework, the model can be updated in real time on the cloud. The objective function comprehensively considers accuracy and speed. Combining the cloud computing framework with the genetic algorithm enhances the real-time and scalability of the model. Therefore, this method is more suitable for complex distribution networks containing multiple substations and multiple substations.

[0082] 7. During the dynamic simulation process, the equivalent model is adjusted based on real-time feedback, and the equivalent method is continuously optimized through a real-time feedback mechanism. A feedback mechanism is established to monitor simulation results in real time, adjusting boundary conditions and the equivalent model based on feedback to ensure that the entire power grid model remains consistent with the actual operating state. A closed-loop control system is used to adjust the equivalent model based on real-time measurement data. A state-space model is also established to represent the feedback process, and an improved LSTM is used to predict load changes. This improves the forward-looking and adaptable nature of the feedback mechanism to accommodate the complexity and dynamic nature of urban power grids. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of the present invention, and do not constitute a limitation of the embodiments of the present invention;

[0084] Figure 1 This is a flow chart of a simplified calculation method for an urban distribution network based on multi-level equivalent modeling in the present invention;

[0085] Figure 2 It is a schematic diagram of the overall process of a simplified calculation method for urban distribution network based on multi-level equivalent modeling in the present invention. DETAILED DESCRIPTION

[0086] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features therein can be combined with each other without conflict.

[0087] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0088] Example 1

[0089] refer to Figure 1 and Figure 2 This embodiment provides a simplified calculation method for an urban distribution network based on multi-level equivalent modeling, the method comprising:

[0090] The distribution network data is acquired, and the boundaries of the distribution network are defined based on the distribution network data. In this embodiment, the boundary conditions may include power supply boundaries, load boundaries, and equipment boundaries.

[0091] For example, based on graph theory, the distribution network can be represented as G=(V,E), where V represents the node set, which can include substations, load points, etc., and E represents the edge set, which can include cables, lines, etc. Nodes can also be grouped based on geographic location, load density, and electrical distance based on the K-means clustering algorithm. The calculation method of the K-means clustering algorithm can be:

[0092] ; (1)

[0093] in, represents the number of clusters, represents an integer greater than or equal to 1, Indicates the clusters, represents a data point, Indicates the The cluster center of a cluster is the average eigenvector of the nodes in the cluster.

[0094] In this embodiment, the method may further include data preprocessing, such as Z-score processing:

[0095] ; (2)

[0096] in, represents the standard score, represents a data point, represents the mean, Represents standard deviation.

[0097] In this embodiment, the method may further include boundary verification and boundary adjustment:

[0098] The boundary verification process can be:

[0099] ; (3)

[0100] ; (4)

[0101] in, and Represents nodes respectively The active power and reactive power of Representation node The voltage amplitude, Representation node The voltage amplitude, Representation node node The voltage phase angle difference, represents the real part of the admittance matrix, represents the imaginary part of the admittance matrix, and are integers greater than or equal to 1, Indicates the number of nodes.

[0102] The boundary adjustment process can be:

[0103] ; (5)

[0104] in, Indicates the adjusted active power, Indicates the active power before adjustment.

[0105] The voltage constraint is:

[0106] ; (6)

[0107] in, Indicates the minimum voltage amplitude, Indicates the maximum voltage amplitude.

[0108] The power constraint is:

[0109] ; (7)

[0110] in, Indicates the minimum active power, Indicates the maximum active power.

[0111] generating an equivalent model based on the complexity hierarchy of the distribution network, the equivalent principle and the boundaries;

[0112] The electrical characteristics and grid structure of the equivalent model are adjusted to generate a distribution network model; in this embodiment, the electrical characteristics include impedance, voltage drop, etc.

[0113] The complexity level includes a high level, a middle level and a low level, and the equivalence principle includes a node equivalence principle and a line equivalence principle;

[0114] The specific steps of generating the equivalent model include:

[0115] Obtain sensitive areas and several subareas of the distribution network;

[0116] Constructing a single equal-value component;

[0117] The high-level is obtained based on the Thevenin equivalent circuit model, the equivalence principle and the boundary; for example, the Thevenin equivalent circuit model can be:

[0118] ; (8)

[0119] in, Indicates equivalent voltage, represents equivalent impedance, Represents the self-admittance parameter of the node.

[0120] The middle level is obtained based on the Ward equivalence model, the equivalence principle and the boundary; for example, the Ward equivalence model can be:

[0121] ; (9)

[0122] in, represents the equivalent admittance matrix of the external system, represents the admittance matrix of the internal system, represents the mutual admittance matrix between the nodes of the external system and the nodes of the internal system, The mutual admittance matrix between the nodes of the internal system and the nodes of the external system is The transposed matrix of represents the admittance matrix of the external system.

[0123] The low level is obtained based on the three-phase admittance matrix, the equivalence principle and the boundary; for example, the calculation formula of the three-phase admittance matrix can be:

[0124] ; (10)

[0125] in, Representation node No. The injected current of the phase, and All represent any one of the three phases. Indicates different three phases, represents the elements of the three-phase admittance matrix, Representation node No. Phase voltage, Indicates the number of nodes.

[0126] The equivalent model is generated based on the sensitive area, the partition, the single equivalent component, the high level, the middle level, and the low level.

[0127] For example, sensitive areas and low levels are used to construct areas that require the highest precision simulation, and the boundary nodes of the area are identified. These nodes are nodes directly connected to the external network (i.e., the mid-level Ward equivalent area); the mid-level area is between the core detailed area (low level) and the external main network (high level), and an intermediate buffer area is constructed to identify the internal nodes and boundary nodes of the mid-level area. The boundary nodes are divided into two categories: inner boundary nodes, which are nodes connected to the boundary nodes of the low-level area, and outer boundary nodes, which are nodes connected to the Thevenin equivalent points of the high-level area; the entire upstream system of the high-level area outside the mid-level area can be converted from Thevenin equivalent to Norton equivalent to facilitate integration with the admittance matrix model. The three-layer model is combined through boundary nodes. The low-level boundary nodes and the inner boundary nodes of the Ward equivalent are the same physical nodes and can be regarded as connection nodes. The outer boundary nodes of the Ward equivalent are the action points of the higher-level Norton equivalent. The low-level detailed admittance matrix, the middle-level Ward equivalent admittance matrix and the Norton admittance converted from the higher level are superimposed on the corresponding rows and columns according to the node relationship they describe. The self-admittance of the shared boundary node is the sum of contributions from multiple parts. The injection current source, the injection current of the low-level internal node comes from the load / power model inside the area, and the injection current of the connection node comes from the local load / power source (if any) that the node may be connected to, plus the contribution of the node current in the low-level and middle-level (usually included in the matrix equation, the explicit current source term may be zero). The injection current of the Ward outer boundary node is equal to the current source of the higher-level Norton equivalent, thereby integrating the three levels to obtain a three-level integrated model.

[0128] By combining single equivalent components, key variables that affect the equivalent model (such as switch state combinations, typical DG output levels, typical load levels, and key component fault states) are identified to form a discrete set of states. Each state represents a specific configuration or operating point of the system. Each state is assigned a probability of occurrence. For each state, a deterministic equivalent parameter is calculated for that state. The equivalent parameters for all states are weighted averaged according to their probability of occurrence to obtain a single, static, expected equivalent parameter. This compresses the dynamic, three-level complex equivalent model that relies on a large number of time-varying / random states into a static, single equivalent parameter in the expected sense. This parameter itself can be used as a single, fixed-parameter equivalent component (such as an equivalent impedance, an equivalent admittance, or an equivalent current source + admittance Norton circuit), replacing the three-level integrated model that previously required maintenance and calculation, thereby obtaining a distribution network model. The specific steps for obtaining sensitive areas and several partitions of the distribution network include:

[0129] Based on the distribution network data, sensitivity analysis is performed to obtain sensitivity of voltage to power, and the sensitive area is obtained based on the sensitivity;

[0130] The distribution network data is clustered based on preset parameters to obtain the partitions, such as grouping nodes based on geographic location, load density, and electrical distance based on a K-means clustering algorithm.

[0131] The specific steps for building a single equivalent component include:

[0132] generating the single equivalent component based on state enumeration and Monte Carlo simulation;

[0133] The calculation method for generating the single equivalent component is:

[0134] ; (11)

[0135] in, represents the equivalent parameter, Indicates status, Represents a collection of states, Indicates status The probability of Indicates status Parameter value.

[0136] The specific steps of adjusting the electrical characteristics and grid structure of the equivalent model include:

[0137] The distributed parameters of the feeder in the equivalent model are simplified based on a lumped parameter model; the lumped parameter model can be:

[0138] ; (12)

[0139] in, represents the equivalent series impedance, represents the equivalent parallel capacitance, represents the impedance per unit length, represents the capacitance per unit length, represent the hyperbolic sine function and the hyperbolic tangent function respectively, represents the propagation constant, Indicates the line length.

[0140] In this embodiment, the feeder mainly refers to a long feeder, and the distributed parameters may include series resistance, series inductance, parallel capacitance, parallel conductance, and the like.

[0141] The load dynamics of the equivalent model are processed based on the Average-Value Model (AVM). Load dynamics refers to the characteristics of load power (active / reactive) or current changing over time. The calculation formula can be:

[0142] ; (13)

[0143] in, represents the average current, represents the integration period, Indicates instantaneous current.

[0144] The grid structure of the equivalent model is optimized based on the minimum spanning tree algorithm; the calculation formula can be:

[0145] ; (14)

[0146] in, represents the minimum spanning tree, represents an edge in the edge set, express function, represents the edge set, Indicates the weight of the edge, which can be impedance or distance.

[0147] Calculating the line impedance of the equivalent model based on the Casson equation accurately reflects the electrical characteristics of the line. This method modifies the traditional line impedance formula to incorporate factors such as earth conductivity and frequency, making the impedance calculation more realistic and quantifying the electromagnetic coupling effect between conductors. It is applicable to three-phase unbalanced lines, multi-circuit lines on the same tower, and distribution networks with densely packed cables. It supports precise simplification of equivalent models by providing parameters for π-type and T-type equivalent circuits, allowing flexible handling of overhead lines (three-phase, multi-circuit lines on the same tower, split conductors), underground cables (multi-core cables, metal sheath effects), and hybrid lines (connected overhead lines and cable segments). It also improves the accuracy of distribution network analysis. Accurate impedance values ​​ensure the accuracy of voltage amplitudes and phase angles in power flow calculations, especially in long feeder or high-load situations. This provides a more realistic reflection of fault current distribution and improves protection setting reliability. The mutual impedance calculated using the Casson equation can be used in sequence impedance matrices (positive, negative, and zero sequence) to analyze asymmetric faults (such as single-phase ground faults) or load imbalance.

[0148] Therefore, applying the Casson equation to the equivalent model can correct the earth return effect and improve the accuracy of impedance calculation; quantify multi-conductor coupling and support unbalanced system analysis; provide key parameters for the equivalent circuit (π type / T type), ensuring that the model can both simplify calculations and retain actual physical characteristics, making it more suitable for distribution network planning, operation and fault analysis.

[0149] The second model parameters of the equivalent model are optimized based on a genetic algorithm. In this embodiment, the first model parameters and the second model parameters may include series parameters (longitudinal impedance): resistance, inductance, and series impedance, etc.; parallel parameters (transverse admittance): capacitance and capacity, conductance, and parallel admittance, etc.

[0150] The specific steps to generate a distribution network model include:

[0151] A three-phase power flow model is obtained based on the Newton-Raphson algorithm. Based on the adjusted equivalent model and the three-phase power flow model, the equivalent model and the three-phase power flow model are fused using an existing fusion method (such as a current injection method and a voltage constraint method) to obtain the distribution network model.

[0152] The three-phase power flow model can be:

[0153] ; (15)

[0154] ; (16)

[0155] ; (17)

[0156] ; (18)

[0157] in, and Represents nodes respectively The active power and reactive power of and Represents nodes respectively node The voltage, Representation node node The voltage phase angle difference, represents the real part of the admittance matrix, represents the imaginary part of the admittance matrix, and are integers greater than or equal to 1, and All represent three phases. Indicates different three phases, Indicates the number of nodes.

[0158] Example 2

[0159] On the basis of the first embodiment, in this embodiment, the method further includes:

[0160] Obtaining the number of simulations of the single equivalent component, and obtaining the reliability of the equivalent model based on the number of simulations;

[0161] The calculation formula for obtaining the reliability is:

[0162] ; (19)

[0163] in, Represents reliability, Indicates the number of simulations, Indicates the The fault indication function of the simulation, Represents an integer greater than or equal to 1.

[0164] Example 3

[0165] Based on the above embodiment, in this embodiment, the method further includes:

[0166] Acquiring energy data of distributed energy resources, constructing a dynamic characteristic probability model based on the energy data, and obtaining distributed energy output characteristics based on the dynamic characteristic probability model;

[0167] The calculation formula of the dynamic characteristic probability model is:

[0168] ; (20)

[0169] in, represents the output power of distributed energy resources, represents the mean value of distributed energy resources, represents the variance of the energy resources of the distributed partitioned units, Represents the probability distribution of distributed energy output power.

[0170] Distributed energy output characteristics refer to the power output patterns exhibited by distributed energy resources (DERs, such as photovoltaics, wind power, energy storage, and micro gas turbines) during operation. Their core characteristics are time-varying, random, and controllable. Specifically, these characteristics include timescale characteristics, short-term fluctuations (seconds to minutes), intraday variations (hourly), seasonal patterns (monthly to annual), probabilistic statistical characteristics, probability distributions, spatiotemporal correlations (output complementarity between multiple DERs or regional agglomeration effects), controllability parameters, and adjustable features such as energy storage charge and discharge rates and gas turbine ramping capabilities. Existing technologies can be used to embed distributed energy output characteristics into equivalent models.

[0171] Example 4

[0172] Based on the above embodiment, in this embodiment, the method further includes:

[0173] Obtaining a detailed model of the distribution network, comparing the detailed model with the equivalent model to obtain an error value, determining whether the error value exceeds a preset threshold, and if so, adjusting a first model parameter of the equivalent model, obtaining a first model based on the first model parameter, and updating the equivalent model to the first model;

[0174] In this embodiment, the detailed model is a complete power grid model including all nodes and parameters.

[0175] The calculation formula of the error value is:

[0176] ;(twenty one)

[0177] in, Indicates the error value, and Represent the voltage vectors of the equivalent model and the detailed model respectively.

[0178] Example 5

[0179] Based on the above embodiment, in this embodiment, the method further includes: updating the three-phase power flow model and the equivalent model based on a preset cloud computing simulation framework and objective function;

[0180] The objective function is:

[0181] ;(twenty two)

[0182] in, represents the objective function, and Both represent weights, Indicates accuracy, Indicates speed.

[0183] In this embodiment, the simulation framework may be an existing cloud computing simulation framework that can implement real-time updating of models in the cloud.

[0184] Example 6

[0185] Based on the above embodiment, in this embodiment, the method further includes:

[0186] Acquire real-time power grid data, construct a state space model, and update the boundary and the equivalent value model based on the state space model and the real-time power grid data to implement equivalent value adjustment in dynamic simulation;

[0187] The calculation formula for updating the equivalent model is:

[0188] ;(twenty three)

[0189] in, Indicates the power adjustment amount, represents the proportional gain, represents the integral gain, Indicates the actual measured power, Indicates analog power;

[0190] The calculation formula of the state space model is:

[0191] ;(twenty four)

[0192] in, express The system status at any moment, express The system status at any moment, express Real-time grid data input at all times, express Output grid data at all times, represents the n×n state matrix, where n is the node, represents an n×m input matrix, where m is the dimension of the input variable, represents a p×n output matrix, where p is the dimension of the output variable, Represents a straight-through matrix.

[0193] In this embodiment, the method further includes:

[0194] Based on the real-time power grid data, the Bayesian method and the particle swarm optimization algorithm are used to adjust the model parameters to achieve model optimization. The particle swarm optimization algorithm can be:

[0195] ; (25)

[0196] in, express The particle speed at time express The particle speed at time represents the inertia weight factor, and Both represent learning factors, and Both represent random numbers between [0,1]. represents the local optimal solution, represents the global optimal solution, express The particle position at the moment.

[0197] The model stability can also be analyzed based on the Lyapunov stability method, and the calculation formula can be:

[0198] ; (26)

[0199] in, represents the energy function, represents the time derivative. If the above conditions are met, the model is stable.

[0200] The solution that adopts dynamic feedback mechanism and automatic correction model can make instant adjustments according to the actual operating status of the power grid and optimize the real-time operation control of the distribution network.

[0201] The proposed equivalent method uses precise boundary condition definition, dynamic adjustment, and correction to reduce computational complexity while preserving the core characteristics of grid operation. This allows the model's calculation results to more closely approximate actual grid operation, thereby improving the accuracy and reliability of simulation results.

[0202] Core features include:

[0203] Preservation of key electrical characteristics: During the equivalent modeling process, key electrical characteristics of the power grid, such as impedance and voltage drop, are preserved. While the distributed parameters of long feeders are simplified using a centralized parameter model during electrical characteristic simplification, this simplification is based on reasonable engineering approximations. Computational optimization algorithms (such as parallel computing and multi-threaded processing) are used to ensure that the simplified model accurately reflects the actual electrical characteristics of the power grid. Furthermore, during the equivalent model verification process, by comparing the power flow and voltage distribution of the equivalent model with the detailed model, if the error exceeds a preset threshold, the equivalent model parameters are adjusted and re-verified, ensuring that the core electrical characteristics of the power grid operation are preserved.

[0204] Accuracy of equivalent modeling: In the equivalent modeling process, both the high-level Thevenin equivalent model and the mid-level Ward equivalent model are based on accurate electrical parameters such as the admittance matrix and voltage phase angle difference of the power grid. These models can accurately describe the core characteristics of the power grid, such as the voltage-power relationship and power flow distribution. For example, the equivalent voltage V th and equivalent impedance Z th It is accurately extracted from the node admittance matrix and can truly reflect the electrical characteristics of the power grid at the node.

[0205] Dynamic characteristic modeling: Distribution networks have dynamic characteristics during operation, such as changes in load and fluctuations in the output of distributed energy resources. The present invention uses feedback from dynamic simulation and equivalent methods to adjust the equivalent model based on real-time feedback during the dynamic simulation process. This dynamic simulation feedback mechanism ensures that the equivalent model can reflect the dynamic operating characteristics of the power grid in real time. For example, by establishing a state-space model to represent the feedback process, using real-time measurement data as input, and adjusting the parameters of the equivalent model, the model can accurately capture the dynamic changes of the power grid and retain the dynamic core characteristics of the power grid operation.

[0206] Focus on Core Equipment and Lines: While some aspects of the grid structure are simplified during equivalent modeling, this paper focuses on core equipment and lines within the grid, such as the voltage distribution of the trunk lines. A more precise modeling approach is employed for these critical components, ensuring that core characteristics of grid operation, such as trunk line voltage quality and power transmission characteristics, are preserved while simplifying the model.

[0207] The model of the present invention is more adaptable in areas with a high proportion of photovoltaic power stations, can effectively reduce calculation errors, and provide more accurate simulation results. It is particularly suitable for simulation of distribution networks with high proportions of photovoltaic power stations and other distributed energy sources. The specific reasons are as follows:

[0208] Modeling the dynamic characteristics of distributed energy resources:

[0209] The equivalent method in this embodiment specifically considers the probabilistic modeling of distributed energy resources, such as photovoltaics, due to their dynamic characteristics. For example, the output power of photovoltaic power sources is intermittent and fluctuating due to natural factors. This invention uses a dynamic probabilistic model to describe the output power of distributed energy resources, including parameters such as its mean and variance. This probabilistic model can more accurately reflect the actual operating characteristics of distributed energy resources. By taking these characteristics into account during the equivalent modeling process, the equivalent model is more adaptable to distribution networks with a high proportion of distributed energy resources, such as photovoltaics.

[0210] Compatibility of the equivalent model with distributed energy resources:

[0211] The equivalent modeling method of this embodiment allows for the integration of new energy sources and energy storage devices within the substation and on the main distribution lines. In the equivalent model, through reasonable boundary condition definitions and equivalent principles, the distributed energy access point and the surrounding power grid structure can be reasonably treated as equivalent. For example, for the substation lines simplified using the line equivalent method, the impact of distributed energy access on the line electrical characteristics is also considered, making the equivalent model compatible with the existence of distributed energy and accurately reflecting its impact on grid operation.

[0212] Adapting to source-load interaction:

[0213] In distribution networks with a high proportion of distributed energy resources, such as photovoltaics, complex source-load interactions exist between distributed energy resources and traditional loads. This embodiment calculates the sensitivity of voltage to power and identifies areas requiring detailed modeling, enabling better handling of this source-load interaction. Furthermore, by utilizing state enumeration and Monte Carlo simulation to generate a single equivalent component, the impact of different operating states of distributed energy resources on the grid is considered. This consideration of source-load interactions enables the equivalent model to adapt to the changing operational characteristics of the grid after a high proportion of distributed energy resources is integrated.

[0214] Error control and accuracy assurance:

[0215] To address the significant errors experienced by traditional equivalence methods when integrating distributed energy resources such as photovoltaics at high penetration rates, this embodiment controls these errors through an equivalence model verification and correction process. During the equivalence model verification process, the power flow and voltage distribution of the equivalence model are compared with those of the detailed model. When the error exceeds a preset threshold, the equivalence model parameters are adjusted and revalidated. Furthermore, model corrections using methods such as Bayesian methods and particle swarm optimization can effectively reduce these errors and improve the accuracy of the equivalence model in simulations of high-proportion distributed energy distribution networks.

[0216] Example 7

[0217] Based on the above embodiment, this embodiment further provides a simplified calculation system for an urban distribution network based on multi-level equivalent modeling, the system further comprising:

[0218] Boundary unit: used for acquiring distribution network data and defining the boundary of the distribution network based on the distribution network data;

[0219] An equivalent model unit: used for generating an equivalent model based on the complex hierarchy of the distribution network, the equivalent principle and the boundary;

[0220] The complexity level includes a high level, a middle level and a low level, and the equivalence principle includes a node equivalence principle and a line equivalence principle;

[0221] The equivalent model unit specifically includes:

[0222] Partition unit: used to obtain sensitive areas and several partitions of the distribution network;

[0223] Component unit: used to construct a single equivalent component;

[0224] Hierarchical unit: used for obtaining the high level based on the Thevenin equivalent circuit model, the equivalence principle and the boundary, obtaining the middle level based on the Ward equivalent model, the equivalence principle and the boundary, and obtaining the low level based on the three-phase admittance matrix, the equivalence principle and the boundary;

[0225] A generating unit, configured to generate the equivalent model based on the sensitive area, the partition, the single equivalent component, the high level, the middle level, and the low level;

[0226] Simplified model unit: used to adjust the electrical characteristics and grid structure of the equivalent model to generate a distribution network model.

[0227] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0228] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A simplified calculation method for urban distribution network based on multi-level equivalent modeling, characterized in that: The method comprises: acquiring distribution network data, and defining a boundary of the distribution network based on the distribution network data; generating an equivalent model based on the complexity hierarchy of the distribution network, the equivalent principle and the boundaries; Adjusting the electrical characteristics and grid structure of the equivalent model to generate a distribution network model; The complexity level includes a high level, a middle level and a low level, and the equivalence principle includes a node equivalence principle and a line equivalence principle; The specific steps of generating the equivalent model include: Obtain sensitive areas and several subareas of the distribution network; Constructing a single equal-value component; The high-level is obtained based on the Thevenin equivalent circuit model, the equivalent principle and the boundary, the middle-level is obtained based on the Ward equivalent model, the equivalent principle and the boundary, the low-level is obtained based on the three-phase admittance matrix, the equivalent principle and the boundary, and the equivalent model is generated based on the sensitive area, the partition, the single equivalent component, the high-level, the middle-level and the low-level; The specific steps of adjusting the electrical characteristics and grid structure of the equivalent model include: Simplifying the distributed parameters of the feeders in the equivalent model based on a lumped parameter model, processing the load dynamics of the equivalent model based on an average value model, optimizing the grid structure of the equivalent model based on a minimum spanning tree algorithm, calculating the line impedance of the equivalent model based on a Casson equation, and optimizing the second model parameter of the equivalent model based on a genetic algorithm; The specific steps to generate a distribution network model include: A three-phase power flow model is obtained based on the Newton-Raphson algorithm, and the distribution network model is obtained based on the adjusted equivalent model and the three-phase power flow model.

2. A simplified calculation method for urban distribution network based on multi-level equivalent modeling according to claim 1, characterized in that: The specific steps to obtain sensitive areas and several zones of the distribution network include: Obtaining sensitivity of voltage to power based on the distribution network data, and obtaining the sensitive area based on the sensitivity; The distribution network data is clustered based on preset parameters to obtain the partitions.

3. The simplified calculation method for urban distribution network based on multi-level equivalent modeling according to claim 2 is characterized in that: The specific steps to build a single equivalent component include: generating the single equivalent component based on state enumeration and Monte Carlo simulation; The calculation method for generating the single equivalent component is: ； in, represents the equivalent parameter, Indicates status, Represents a collection of states, Indicates status The probability of Indicates status Parameter value.

4. The simplified calculation method for urban distribution network based on multi-level equivalent modeling according to claim 3 is characterized in that: The method further comprises: Obtaining the number of simulations of the single equivalent component, and obtaining the reliability of the equivalent model based on the number of simulations; The calculation formula for obtaining the reliability is: ； in, Represents reliability, Indicates the number of simulations, Indicates the The fault indication function of the simulation, Represents an integer greater than or equal to 1.

5. The simplified calculation method for urban distribution network based on multi-level equivalent modeling according to claim 1 is characterized in that: The method further comprises: Acquiring energy data of distributed energy resources, constructing a dynamic characteristic probability model based on the energy data, and obtaining distributed energy output characteristics based on the dynamic characteristic probability model; The calculation formula of the dynamic characteristic probability model is: ； in, represents the output power of distributed energy resources, represents the mean value of distributed energy resources, represents the variance of distributed energy resources, Represents the probability distribution of distributed energy output power.

6. The simplified calculation method for urban distribution network based on multi-level equivalent modeling according to claim 1 is characterized in that: The method further comprises: Obtaining a detailed model of the distribution network, comparing the detailed model with the equivalent model to obtain an error value, determining whether the error value exceeds a preset threshold, and if so, adjusting a first model parameter of the equivalent model, obtaining a first model based on the first model parameter, and updating the equivalent model to the first model; The calculation formula of the error value is: ； in, Indicates the error value, and Represent the voltage vectors of the equivalent model and the detailed model respectively.

7. The simplified calculation method for urban distribution network based on multi-level equivalent modeling according to claim 1 is characterized in that: The method further comprises: Based on a preset cloud computing simulation framework and objective function, updating the three-phase power flow model and the equivalent model; The objective function is: ； in, represents the objective function, and Both represent weights, Indicates accuracy, Indicates speed.

8. The simplified calculation method for urban distribution network based on multi-level equivalent modeling according to claim 1 is characterized in that: The method further comprises: Acquire real-time power grid data, construct a state space model, and update the boundary and the equivalent model based on the state space model and the real-time power grid data; The calculation formula for updating the equivalent model is: ； in, Indicates the power adjustment amount, represents the proportional gain, represents the integral gain, Indicates the actual measured power, Indicates analog power; The calculation formula of the state space model is: ； in, express The system status at any moment, express The system status at any moment, express Real-time grid data input at all times, express Output grid data at all times, represents the state matrix, represents the input matrix, represents the output matrix, Represents a straight-through matrix.

9. A simplified calculation system for urban distribution network based on multi-level equivalent modeling, characterized in that: The system further comprises: Boundary unit: used for acquiring distribution network data and defining the boundary of the distribution network based on the distribution network data; An equivalent model unit: used for generating an equivalent model based on the complex hierarchy of the distribution network, the equivalent principle and the boundary; The complexity level includes a high level, a middle level and a low level, and the equivalence principle includes a node equivalence principle and a line equivalence principle; The equivalent model unit specifically includes: Partition unit: used to obtain sensitive areas and several partitions of the distribution network; Component unit: used to construct a single equivalent component; Hierarchical unit: used for obtaining the high level based on the Thevenin equivalent circuit model, the equivalence principle and the boundary, obtaining the middle level based on the Ward equivalent model, the equivalence principle and the boundary, and obtaining the low level based on the three-phase admittance matrix, the equivalence principle and the boundary; A generating unit, configured to generate the equivalent model based on the sensitive area, the partition, the single equivalent component, the high level, the middle level, and the low level; Simplified model unit: used for adjusting the electrical characteristics and grid structure of the equivalent model to generate a distribution network model; The specific steps of adjusting the electrical characteristics and grid structure of the equivalent model include: Simplifying the distributed parameters of the feeders in the equivalent model based on a lumped parameter model, processing the load dynamics of the equivalent model based on an average value model, optimizing the grid structure of the equivalent model based on a minimum spanning tree algorithm, calculating the line impedance of the equivalent model based on a Casson equation, and optimizing the second model parameter of the equivalent model based on a genetic algorithm; The specific steps to generate a distribution network model include: A three-phase power flow model is obtained based on the Newton-Raphson algorithm, and the distribution network model is obtained based on the adjusted equivalent model and the three-phase power flow model.

Images