Electromagnetic transient simulation method based on system dynamic characteristic difference decoupling

By dividing the power system into slow-change subsystem and fast-change subsystem and performing independent solutions and information interaction, the problem of insufficient computational complexity and flexibility of traditional electromagnetic transient simulation is solved, and efficient electromagnetic transient simulation is achieved, which is suitable for Renovation and promotion of existing simulation platforms.

CN119004840APending Publication Date: 2024-11-22NORTH CHINA ELECTRIC POWER UNIV +1
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411189136.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-28
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

Traditional electromagnetic transient simulation methods have large calculation volume and long simulation time in large-scale power systems, which is difficult to meet real-time requirements. The existing decoupling methods have problems with insufficient computational complexity and flexibility.

Method used

By dividing the system into slow-change subsystem and fast-change subsystem, using node analysis method for discretization and matrix transfer mapping, independent solution and information interaction of each subsystem are realized, the equivalent process is simplified, and the additional components and contact line currents are avoided. Ask for solution.

Benefits of technology

It significantly reduces the computational complexity and time, improves simulation efficiency, is suitable for the transformation of existing simulation platforms, is universal and efficient, and is suitable for electromagnetic transient simulation of power systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure BDA0005014543520000041
    Figure BDA0005014543520000041
  • Figure BDA0005014543520000042
    Figure BDA0005014543520000042
  • Figure BDA0005014543520000043
    Figure BDA0005014543520000043
Patent Text Reader

Abstract

The invention relates to an electromagnetic transient simulation method based on system dynamic characteristic difference decoupling. An electromagnetic transient decoupling simulation method is constructed from the dynamic characteristics of a power system. The system is divided into two parts according to the time constant difference of each device of the power system, the part with the continuously changing topological structure is regarded as a fast-changing subsystem, and the other part is a slow-changing subsystem. Transfer mapping of a node admittance matrix and an injection current source column vector is carried out on each side of boundary points of the two subsystems, two matrix descriptions of the same system are formed, and therefore decoupling simulation is achieved. The method is a coarse-grained simulation segmentation method with universality, and is characterized in that accurate matrixes for describing a slow-varying subsystem and a quick-varying subsystem can be respectively formed on a simulation level without introducing additional virtual elements and a complex equivalence method, electromagnetic transient decoupling simulation of a power system is realized, the calculation complexity is reduced, and the simulation efficiency is improved. And the simulation efficiency is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of electromagnetic transient simulation of power systems, and in particular relates to an electromagnetic transient simulation method based on differential decoupling of system dynamic characteristics. Background Art

[0002] With the proposal of the "dual carbon" goal, the grid connection of a high proportion of renewable energy has profoundly changed the characteristics of the power system, and the traditional power system is transforming into a new power system with "dual high" characteristics. Electromagnetic transient simulation is an important tool for designing and analyzing power systems. However, with the expansion of the system scale and the introduction of a large number of power electronic equipment, the simulation step size is further reduced, the dimension is expanded, and the amount of calculation is increased, resulting in unacceptable offline simulation time. The calculation pressure of real-time simulation is high and it is difficult to meet real-time requirements. This has brought huge challenges to electromagnetic transient simulation and put forward new requirements for the existing simulation system.

[0003] To meet such challenges, the industry is currently improving the efficiency of electromagnetic transient simulation from two aspects: simulation model optimization and simulation technology innovation.

[0004] In terms of modeling optimization, the Ward equivalent method is used for specific topologies to eliminate internal nodes, achieving external equivalence and improving efficiency while retaining all information of the eliminated nodes, making it possible to solve the internal state quantity. There is also a method based on step average, which transmits accurate trigger pulses based on step average to the power electronic switch model, which can increase the simulation step size and improve efficiency while ensuring the same accuracy.

[0005] In terms of simulation technology, system decoupling is an important way to improve the efficiency of electromagnetic transient simulation. The typical decoupling method in electromagnetic transient simulation is to use the transmission lines in the power system, and use the relationship between the wave propagation time in the long line and the simulation step size to naturally decouple the systems at both ends of the line, but there are certain requirements for the coordination of the decoupling position, line length and simulation step size. There is also a multi-region Thevenin equivalent method for system decoupling, which decouples the system by first solving the interconnection line current between subnets, and then uses this current to calculate the node voltage of each subnet in parallel. However, when this method is used for network division, it needs to be performed on concentrated parameter elements, and the network division flexibility is insufficient. The node splitting method can be divided at any point in the system, which greatly improves the flexibility of network division, but when the number of divided subsystems increases, the difficulty of solving the Thevenin equivalent of each subsystem and the interconnection line current increases, and the amount of calculation increases, which offsets the efficiency improvement advantage brought by decoupling. Delay insertion decoupling is to represent the ground branch as a parallel connection of capacitors, conductances and current sources, and the branch between two nodes as a series connection of inductors, resistors and voltage sources. The characteristics of inductors and capacitors are used to create a half-step delay to achieve decoupling, making the computational complexity linear with the simulation scale. However, few systems naturally meet the conditions required by the delay insertion method, and it is necessary to add non-real capacitors and inductors, which changes the original structure of the system.

[0006] In view of the shortcomings of the above decoupling method, an electromagnetic transient simulation method based on system dynamic characteristic difference decoupling has made up for the problems of the above decoupling method to a certain extent from the perspective of simulation technology. Summary of the invention

[0007] The present invention provides an electromagnetic transient simulation method based on decoupling of system dynamic characteristics differences, with the aim of overcoming the defects of the above-mentioned traditional decoupling methods. At the same time, it only requires lightweight transformation and upgrading of the traditional simulation platform to support the application of the present invention, which has practical promotion and application value.

[0008] The present invention can be implemented by the following technical solution: an electromagnetic transient simulation method based on system dynamic characteristic difference decoupling, comprising:

[0009] Step 1: Based on the differences in system dynamic characteristics, the entire system is coarsely divided into slow-changing subsystems and fast-changing subsystems;

[0010] Step 2: Based on the node analysis method, the components in each subsystem are discretized to form the node admittance matrix of each subsystem and the column vector of the injected current source;

[0011] Step 3: Transfer and map the matrices and column vectors of each subsystem to form two matrix descriptions of the same system;

[0012] Step 4: Solve each subsystem independently to obtain the system state quantity of this time step;

[0013] Step 5: Steps 2 to 4 are executed repeatedly until the simulation time meets the simulation end condition and the electromagnetic transient simulation is completed.

[0014] Preferably, the system division method in step 1 is specifically as follows:

[0015] The power electronic equipment that needs to frequently switch states to change the system topology, the AC filter device that filters out high-frequency components, the high-frequency transformer and other networks that process high-frequency electric energy are regarded as fast-changing subsystems. The network operating near the power frequency range is regarded as a slow-changing subsystem. Through the above division, the entire system is classified into two parts: the fast-changing subsystem and the slow-changing subsystem.

[0016] Preferably, the process of element discretization and matrix formation in step 2 is specifically as follows:

[0017] The node analysis method is the most widely used electromagnetic transient simulation method. This method discretizes all the electrical components of the system to obtain their respective accompanying circuit models, and forms the node admittance matrix and the injection current source column vector through the node-branch association relationship matrix, the branch admittance matrix and the branch current source column vector.

[0018] Preferably, the matrix transfer mapping in step 3 is specifically:

[0019] The detailed information inside the subsystem is composed of the node admittance matrix of the subsystem and the column vector of the injected current source in the simulation. The power system constitutes a loop, and the electromagnetic transient simulation is a solution to the system as a whole. However, the node admittance matrix and the column vector of the injected current of each subsystem mentioned above are relatively independent, with different dimensions. The subsystems are only connected by boundary nodes, and cannot exchange information to obtain information from other parts outside the subsystem. Therefore, the node analysis method cannot be used to solve it. The matrix transfer mapping is essentially to uniformly compress the subsystem information to the boundary nodes between subsystems. Through the mutual transmission and mapping of information between subsystems, each subsystem can observe each other, forming two matrix descriptions of the same system.

[0020] Preferably, the independent solution of the subsystem in step 4 is specifically as follows:

[0021] The two matrix descriptions of the same system refer to the whole system description formed by the fast-changing subsystem as the main body and the whole system description formed by the slow-changing subsystem as the main body. This realizes the decoupling of the fast-changing subsystem and the slow-changing subsystem. By applying the node analysis method to solve the two matrix equations separately, the state variables of each subsystem can be solved.

[0022] Preferably, the step 5 is to execute steps 2 to 4 in a loop, driving the simulation process with time, and by looping steps 2 to 4, the operating state of the power system at each time step is solved to complete the electromagnetic transient simulation.

[0023] Compared with the existing decoupling simulation technology, the present invention has the following advantages:

[0024] The present invention does not need to introduce any additional components, does not need to solve the interconnection line current, and has a more simplified equivalent process. It is a universal coarse-grained simulation segmentation method. After adopting this method, the traditional decoupling method can still be applied to each subsystem to further improve the simulation efficiency. At present, internationally renowned commercial power system electromagnetic transient offline and real-time simulation software such as PSCAD, RSCAD, etc. and some domestic simulation platforms all use the node analysis method for simulation. Since the overall simulation decoupling process of the present invention is based on the traditional node analysis method, the relevant practitioners have a high degree of awareness of this method, which makes the simulation method based on the decoupling of system dynamic characteristics differences involved in the present invention easy to promote and apply, suitable for the transformation of existing simulation platforms to improve efficiency, and has promotion and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 creative work.

[0026] Figure 1 Capacitor discretization adjoint circuit model.

[0027] Figure 2 Schematic diagram of the discretization and transfer mapping of each subsystem.

[0028] Figure 3 Schematic diagram of the dual active bridge test system.

[0029] Figure 4 Dual active bridge low voltage DC side voltage waveform.

[0030] Figure 5 Voltage waveform on the high-voltage DC side of the dual active bridge. DETAILED DESCRIPTION

[0031] The present invention provides an electromagnetic transient simulation method based on decoupling of system dynamic characteristics differences. The steps of the electromagnetic transient decoupling simulation method constructed by the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0032] Step 1: Based on the differences in system dynamic characteristics, the entire system is coarsely divided into slow-changing subsystems and fast-changing subsystems.

[0033] With the development of new energy, the power system is full of power electronic equipment, which greatly changes the characteristics of traditional power systems. In the electromagnetic transient simulation based on the node analysis method, the high-frequency changing power electronic equipment constantly changes the topology of the system, which means that the node admittance matrix needs to be frequently inverted, which brings great computing pressure to the computing equipment. The present invention classifies the part that processes high-frequency electric energy as a fast-changing subsystem, and the part that traditionally operates in the baseband AC range as a slow-changing subsystem. The complexity of directly inverting an N-dimensional matrix is ​​O(N 3 ), the complexity of inversion using improved methods such as Gaussian elimination is O(N 3 / 3), but it is still the cube of N. If the fast-changing subsystem and the slow-changing subsystem are solved independently, it means that the matrix with dimension N is decomposed into two matrices with dimension sum N and inverted separately. This can significantly reduce the computational complexity and time, and improve the simulation efficiency. This is the theoretical basis for improving the simulation efficiency of the present invention. Steps 2, 3, and 4 will focus on this goal, decouple the fast-changing subsystem and the slow-changing subsystem, and achieve independent solution.

[0034] Step 2: Based on the node analysis method, the components in each subsystem are discretized to form the node admittance matrix of each subsystem and the column vector of the injected current source.

[0035] The discretization of electrical components is to discretize them using numerical methods based on the actual dynamic characteristics of the components, and to express them with the circuit model. Taking the discretization of capacitors as an example, the change rate of its port current i(t) is proportional to the port voltage u(t):

[0036]

[0037] Integrate the equation (1) at both ends with Δt as the interval:

[0038]

[0039] The formula (2) is discretized using trapezoidal integral:

[0040]

[0041] Formula (3) can be used Figure 1 The accompanying circuit model expression is shown.

[0042] The accompanying circuits of the discretized electrical components form a branch admittance matrix and a branch current source column vector, and cooperate with the connection relationship of each electrical component, that is, the node-branch association matrix, to form a node admittance matrix and an injected current source column vector.

[0043] Step 3: Transfer and map the matrices and column vectors of each subsystem to form two matrix descriptions of the same system.

[0044] Transfer mapping is to compress the dimension of the subsystem node admittance matrix and the injected current column vector to the same dimension as the boundary node. The information of each subsystem is transferred and mapped through the boundary nodes so that adjacent subsystems can observe each other, forming a matrix description of the whole system as seen from each subsystem, that is, two matrix descriptions of the same system.

[0045] The node admittance matrix contains not only the system topology, but also the characteristics of each electrical component. By combining the injected current source of each node, the subsystem characteristics can be fully reflected. The information of the transfer mapping is the subsystem node admittance matrix and the injected current source column vector. Information interaction requires an interface as a medium for transmission, and the interface is shared by the required interaction subjects. Therefore, the information of the transfer mapping must be exchanged through the public part between subnets, and the boundary node is the interface for information transmission. By compressing the subsystem node admittance matrix and the injected current source column vector to the boundary node, the boundary node can be used as a medium to transmit subsystem information, thus realizing mutual observation between subsystems. Figure 2 Schematic diagram of the discretization and transfer mapping of each subsystem.

[0046] The following will take the mapping of the system information transfer of the slow-changing subsystem A to the fast-changing subsystem B as an example to explain how to make the fast-changing subsystem observe the external system information. For subsystem A, in the process of generating the node admittance matrix, the nodes are classified, the nodes connected to other subsystems are boundary nodes, and the remaining nodes are internal nodes. The node voltage equation of the slow-changing subsystem A can be written as follows:

[0047]

[0048] The essence of equation (4) is Kirchhoff's first law. The left side of the equation is the outflow current of each node of subsystem A, and the right side of the equation is the injection current of each node of the subnet and the current flowing into the boundary node of the external system. α,α , G a,a They represent the node admittance matrix between boundary nodes and the node admittance matrix between internal nodes of subsystem A, G α,a , G a,α represents the node admittance matrix between the boundary nodes and internal nodes of subsystem A, U α is the boundary node voltage of subsystem A, U a is the internal node voltage of subsystem A, J α , J a are the injection current sources of the boundary nodes and internal nodes of subsystem A, I α Represents the current flowing from B to A between subsystem A and subsystem B.

[0049] Decoupling is reflected in the matrix as only containing elements on the main diagonal, and the rest of the positions are 0. From equation (4), it can be seen that the off-diagonal elements of the node admittance matrix make the boundary node voltages and the internal node voltages coupled together. By transforming the matrix into G α,a By setting the elements to 0, the influence of the internal node voltage on the boundary node voltage can be eliminated:

[0050]

[0051] In formula (5): represents the compression matrix of the admittance information of node A of the subsystem, It represents the compressed column vector of the injected current information of each node of subsystem A. The two are the information required for transfer mapping of subsystem A. Since the dimension of the column is the same as the dimension of the boundary node, it is convenient to transmit information with the fast-changing subsystem B, so that the fast-changing subsystem can observe the slow-changing subsystem, and the slow-changing subsystem observes the fast-changing subsystem in the same way.

[0052] Step 4: Solve each subsystem independently to obtain the system state quantity of this time step.

[0053] Step 3 forms a matrix description of the whole system from the perspective of each subsystem. For the fast-changing subsystem, after compressing the information of the slow-changing subsystem, the information is transmitted to the fast-changing subsystem through the boundary nodes between the slow-changing subsystem and the fast-changing subsystem. At this time, the fast-changing subsystem observes the slow-changing subsystem, and the information of the whole system is known. The network equation can be solved to obtain the voltage of all nodes of the fast-changing subsystem. The information of the fast-changing subsystem to the slow-changing subsystem is similar, and the solution can obtain the voltage of all nodes of the slow-changing subsystem. In this way, the state variables of the whole system are obtained, and the system solution of this time step is completed.

[0054] Step 5: Loop through steps 2 to 4. The electromagnetic transient simulation based on the node analysis method is a time-driven simulation. By looping through steps 2 to 4, the operating status of the power system at each time step can be solved. When the simulation end conditions are met, the electromagnetic transient simulation is completed.

[0055] Figure 3 The dual active bridge system is simulated by applying the electromagnetic transient method based on system dynamic characteristic difference decoupling constructed by the present invention, and the simulation results are compared and analyzed with PSCAD / EMTDC. Figure 4 , Figure 5 The comparison of the voltage waveforms of the low-voltage DC side and the high-voltage DC side of the dual active bridge is shown in Figure 1. The relative errors of each electrical quantity are within 1%, which can accurately reflect the dynamic characteristics of the system. From the perspective of computational efficiency, at a simulation step size of 2.5μs, the method of the present invention has a 4.39-fold speed-up effect compared with PSCAD / EMTDC, which greatly improves the simulation efficiency.

[0056] Finally, it should be noted that the above description is only a preferred embodiment of the present invention that is easily understood by those skilled in the art, and is not intended to limit the present invention. Any modification, equivalent substitution, and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An electromagnetic transient simulation method based on decoupling of system dynamic characteristics differences, the core significance of which is that after classifying the power system according to the dynamic characteristics difference, the node analysis method is applied to form two descriptions of the same topology through the mutual transfer mapping of the node admittance matrix and the injection current source column vector between subsystems, and the independent solution between subsystems is realized. The computational complexity of the algorithm is reduced by matrix dimensionality reduction, thereby improving the efficiency of electromagnetic transient simulation. Implementation of the present invention The following steps are involved: Step 1: Based on the differences in system dynamic characteristics, the entire system is coarsely divided into slow-changing subsystems and fast-changing subsystems; Step 2: Based on the node analysis method, the components in each subsystem are discretized to form the node admittance matrix of each subsystem and the column vector of the injected current source; Step 3: Transfer and map the matrices and column vectors of each subsystem to form two matrix descriptions of the same system; Step 4: Solve each subsystem independently to obtain the system state quantity of this time step; Step 5: Steps 2 to 4 are executed repeatedly until the simulation time meets the simulation end condition and the electromagnetic transient simulation is completed.

2. According to claim 1, an electromagnetic transient simulation method based on system dynamic characteristic difference decoupling, It is characterized in that The system division method in step 1 is specifically as follows: All power electronic devices that quickly switch states and change system topology to convert electric energy through high-frequency changes are considered fast-changing subsystems; the part that only operates in the base frequency AC range is considered a slow-changing subsystem. The entire system is classified into two parts: fast-changing subsystems and slow-changing subsystems according to the difference in dynamic characteristics or the size of time constants.

3. According to claim 1, an electromagnetic transient simulation method based on system dynamic characteristics difference decoupling, It is characterized in that The process of element discretization and matrix formation in step 2 is specifically as follows: The electrical components are discretized according to their own dynamic characteristics to obtain the accompanying circuit model, which can be expressed in the form of a conductance parallel historical current source, and then the node admittance matrix and the injection current source column vector are generated according to the network topology connection relationship.

4. The electromagnetic transient simulation method based on system dynamic characteristic difference decoupling according to claim 1, It is characterized in that The matrix transfer mapping of step 3 is specifically as follows: The fast- and slow-changing subsystem information is compressed to the boundary nodes, and the compressed node admittance matrix of each subsystem and the injected current source vector are interacted with each other to transmit the information required for simulation, thereby realizing the decoupling of the fast-changing subsystem and the slow-changing subsystem.

5. The electromagnetic transient simulation method based on system dynamic characteristic difference decoupling according to claim 1, It is characterized in that The specific solution of the subsystem in step 4 is as follows: According to step 3, two matrix descriptions of the same system can be obtained. The state variables of each subsystem at the time step can be solved by applying the node voltage method.

6. The electromagnetic transient simulation method based on system dynamic characteristic difference decoupling according to claim 1, It is characterized in that The step 5 is to execute steps 2 to 4 in a loop until the simulation time meets the simulation end requirement.

Citation Information

Cited By

  • Electromagnetic transient simulation decoupling method and system based on small-step synthesis and medium

    CN120579344A

  • Data splitting processing method for electromagnetic simulation model of power system

    CN120764224A

  • A decoupled parallel simulation method, system, terminal and medium for a reactive power generator

    CN122242423B