阅读说明：本技术 电力电子换流器实时仿真的电磁暂态仿真方法 (Electromagnetic transient simulation method for real-time simulation of power electronic converter ) 是由 汪可友 李子润 徐晋 吴盼 李国杰 于 2021-02-25 设计创作，主要内容包括：一种电力电子换流器实时仿真的电磁暂态仿真方法,其最大特点在于将可以采用极小的纳秒级步长进行建模,同时考虑硬件计算能力采用微秒级仿真步长进行仿真。为此首先以纳秒级步长构建电力电子换流器离散状态空间模型,以历史电流为状态变量,节点电压与支路电流为输出变量,描述电力电子换流器的工作状态,通过对其系数矩阵进行重构,并以微秒级步长对电力电子换流器离散状态空间模型进行仿真求解。本发明采用微秒级仿真步长进行仿真,并达到纳秒级步长的仿真精度的同时,可以使L/C开关模型在开关动作后的暂态误差快速收敛,有效解决了传统实时仿真开关模型存在的虚拟功率损耗问题,极大提高了仿真精度。(The electromagnetic transient simulation method for real-time simulation of the power electronic converter is mainly characterized in that modeling can be carried out by adopting a very small nanosecond step length, and simulation is carried out by adopting a microsecond simulation step length in consideration of hardware computing capacity. The method comprises the steps of constructing a discrete state space model of the power electronic converter by nanosecond step length, describing the working state of the power electronic converter by taking historical current as a state variable and taking node voltage and branch current as output variables, reconstructing a coefficient matrix of the power electronic converter, and carrying out simulation solving on the discrete state space model of the power electronic converter by microsecond step length. According to the invention, microsecond-level simulation step length is adopted for simulation, and the simulation precision of nanosecond-level step length is achieved, meanwhile, the transient error of the L/C switch model after switching action can be rapidly converged, the problem of virtual power loss of the traditional real-time simulation switch model is effectively solved, and the simulation precision is greatly improved.)

1. An electromagnetic transient simulation method for real-time simulation of a power electronic converter is characterized by comprising the following steps:

numbering power electronic converters and branches and nodes of circuits where the power electronic converters are located respectively, wherein the number of a grounding node is 0; forming a correlation matrix K of the power electronic converter system, wherein the row number of the correlation matrix K is the node number N_nThe number of columns is the number of branches N_bThe formation rule is as follows:

1) if node i is connected to branch j, then K (m) when current flows from node i to branch j_ij)＝1；

2) If node i is connected to branch j, then K (m) when current flows from branch j to node i_ij)＝-1；

3) K (m) if node i is not connected to branch j_ij)＝0；

Discretizing a resistance branch, an inductance branch, a capacitance branch and a switch branch of the power electronic converter system to obtain a discrete equivalent model formed by connecting a historical current source and an equivalent admittance in parallel, replacing the independent voltage source branch with the discrete equivalent model formed by connecting the historical current source and the equivalent admittance in parallel, wherein discretizing is carried out by adopting a step length h:

the equivalent admittance of the discrete equivalent model of the resistance branch is expressed as:

the resistance branch discrete equivalent model does not contain a historical current source: i is_hR＝0

Wherein, R is the resistance value of the resistance branch circuit;

the equivalent admittance of the discrete equivalent model of the inductance branch is expressed as:

the historical current source of the inductance branch discrete equivalent model is represented as: i is_hL(t)＝i_L(t-h)

Where L is the inductance of the inductor branch, i_L(t-h) is the branch current of the inductive branch at the last simulation moment;

the equivalent admittance of the discrete equivalent model of the capacitive branch is expressed as:

the historical current source of the capacitor branch discrete equivalent model is represented as: i is_hC(t)＝-Y_bCu_C(t-h)

Where C is the capacitance value of the capacitive branch, u_C(t-h) is the branch voltage of the capacitor branch at the last simulation moment;

let the equivalent admittance of the switch be: y is_bSW＝1；

When the switch is switched on, the historical current source of the discrete equivalent model of the switch branch is represented as: i is_hSW(t)＝i_SW(t-h)；

When the circuit is switched off, the historical current source of the discrete equivalent model of the switch branch circuit is represented as: i is_hSW(t)＝-Y_bSWu_SW(t-h)；

Wherein i_SW(t-h) is the branch current of the switching branch at the last simulation instant, u_SW(t-h) is the branch voltage of the last simulation time switch branch;

if the current simulation time t is the simulation initial time, the size of the historical current sources of all the branches is zero; if the current simulation time is not the initial simulation time, calculating the size of the historical current source of each branch at the current simulation time according to the branch voltage and the branch current of each branch at the previous simulation time;

step (3) according to the discretized power electronic converter system, a node admittance matrix Y is constructed_nAnd branch admittance matrix Y_b；

Step (4) using history current source I_hFor state variables, injecting electricityStream source I_sFor input variable, node voltage V_nAnd branch current I_bFor the output variables, a discrete state space model of the power electronic converter system is constructed as follows:

coefficient matrices A, B, C and D of the discrete state space model of the power electronic converter are formed according to the following rules:

wherein M is_LNumber of branches N_bX number of branches N_bWhen the branch circuit is in an inductance or switch conducting state, the corresponding element is 1, otherwise, the corresponding element is 0; m_CNumber of branches N_bX number of branches N_bWhen the branch circuit is in a capacitance or switch off state, the corresponding element is 1, otherwise, the corresponding element is 0;

and (5) selecting a simulation step length delta t-kh, wherein k is a positive integer, and reconstructing the state equation coefficient matrixes A and B of the discrete state space model of the power electronic converter according to the following rules to obtain the coefficient matrixes as follows:

wherein A is_i' reconstruction of A by i times to obtain a matrix, A₀' is A_iInitial value of `, B_i' is a matrix obtained by reconstructing B i times, B₀' is B_i' initial value;

therefore, the state equation of the small step synthesis model of the discrete state space of the power electronic converter system after small step synthesis is obtained, wherein the modeling step length is h, the simulation step length is delta t, and the state equation is as follows:

wherein, A 'and B' are coefficient matrixes obtained after reconstructing A and B for k times respectively;

step (6) updating the historical current of each branch at the next moment according to the sizes of the historical current and the injected current at the current simulation moment and in combination with the coefficient matrixes A 'and B';

step (7) updating the node voltage of each node at the current moment according to the historical current and the injected current at the current simulation moment and in combination with the coefficient matrixes C and D;

step (8) updating the branch current of each branch at the current moment according to the historical current and the injected current at the current simulation moment and by combining the coefficient matrixes C and D;

if the simulation is not finished, the next simulation period is entered, and the step (6) is returned; and if the simulation ending time is reached, ending the simulation.

Technical Field

The invention relates to a power system, in particular to an electromagnetic transient simulation method for real-time simulation of a power electronic converter.

Background

In the late 60 s of the 20 th century, doctor h.w.dommel, canada developed digital computer algorithms, commonly known as emtp (electro magnetic transitions program), for electromagnetic transient simulation studies of electric power systems. Electromagnetic transient simulation can be mainly divided into off-line simulation and real-time simulation. Generally, the computation time required for off-line simulation is far longer than the duration of the transient phenomenon under study, and particularly for a power electronic converter system with high-frequency switches, the simulation efficiency is severely limited. The real-time simulation can ensure the accurate synchronization of the internal clock of the simulator and the real world clock, and the mutual cooperation of the simulation model of the computer power electronic converter and the actual protection and control device provides a highly simulated on-site actual test environment, thereby greatly improving the simulation efficiency.

For real-time simulation with high precision requirement, a power electronic converter generally uses a power electronic switch as a basic unit for modeling, and common power electronic switch models are mainly divided into the following two types:

1. binary resistance (R)_on/R_off) Model: when the switch is turned on, the switch is equivalent to a resistor with a small resistance value, and when the switch is turned off, the switch is equivalent to a resistor with a large resistance value. When the switch acts, the binary resistance model can cause recalculation of an admittance matrix of the power electronic converter system, and the requirement of real-time performance is difficult to meet, so that the binary resistance model is generally used in off-line simulation software such as PSCAD (power system computer aided design).

2. Inductance/capacitance (L/C) based constant admittance model: the switch is equivalent to a small inductor when the switch is turned on, and is equivalent to a small capacitor when the switch is turned off. The equivalent admittance of the inductor when the circuit is switched on is equal to that of the capacitor when the circuit is switched off, so that the frequent change of an admittance matrix caused by the action of a high-frequency switch is avoided, the simulation efficiency is greatly improved, and the circuit is widely applied to real-time simulation platforms such as RTDS (real time digital system) and ADPSS (advanced digital simulator).

In the EMTP, the L/C switch model generates non-negligible transient errors after switching action, which are inconsistent with the actual switching action, so that the problem of virtual power loss is caused, and the simulation precision is seriously influenced. The electromagnetic transient simulation method suitable for real-time simulation of the power electronic converter can compress the transient process of the L/C switch model after switching action into a simulation step length, neglect the transient process, ensure the constant admittance characteristic and simultaneously enable the model to be similar to an ideal switch, thereby effectively solving the problem of virtual power loss. Meanwhile, the method can be used for modeling in nanosecond step length, and simultaneously adopts microsecond simulation step length for simulation, thereby taking hardware computing capacity and simulation precision into consideration.

Disclosure of Invention

The invention aims to provide an electromagnetic transient simulation method for real-time simulation of a power electronic converter aiming at the defects of the conventional power electronic switch constant admittance model, which is used for modeling the power electronic converter model by adopting nanosecond step length and simulating by adopting microsecond step length, so that the mutual restriction of hardware computing capacity and simulation precision is effectively solved.

The technical solution of the invention is as follows:

an electromagnetic transient simulation method for real-time simulation of a power electronic converter is characterized by comprising the following steps:

numbering power electronic converters and branches and nodes in circuits of the power electronic converters respectively, wherein the number of the grounding node is 0; forming a correlation matrix K of the power electronic converter system, wherein the row number of the correlation matrix K is the node number N_nThe number of columns is the number of branches N_bThe formation rule is as follows:

1) if node i is connected to branch j, then K (m) when current flows from node i to branch j_ij)＝1；

2) If node i is connected to branch j, then K (m) when current flows from branch j to node i_ij)＝-1；

3) K (m) if node i is not connected to branch j_ij)＝0；

Discretizing a resistance branch, an inductance branch, a capacitance branch and a switch branch of the power electronic converter system to obtain a discrete equivalent model formed by connecting a historical current source and an equivalent admittance in parallel, replacing the independent voltage source branch with the discrete equivalent model formed by connecting the historical current source and the equivalent admittance in parallel, wherein discretizing is carried out by adopting a step length h:

the equivalent admittance of the discrete equivalent model of the resistance branch can be expressed as:

the resistance branch discrete equivalent model does not contain a historical current source: i is_hR＝0

Wherein, R is the resistance value of the resistance branch.

The equivalent admittance of the discrete equivalent model of the inductive branch can be expressed as:

the historical current source of the inductance branch discrete equivalent model can be expressed as: i is_hL(t)＝i_L(t-h)

Where L is the inductance of the inductor branch, i_LAnd (t-h) is the branch current of the inductive branch at the last simulation moment.

The equivalent admittance of the discrete equivalent model of the capacitive branch can be expressed as:

the historical current source of the capacitor branch discrete equivalent model can be represented as: i is_hC(t)＝-Y_bCu_C(t-h)

Where C is the capacitance value of the capacitive branch, u_CAnd (t-h) is the branch voltage of the capacitor branch at the last simulation moment.

The switch branch circuit is equivalent to a small inductor when the switch is switched on and is equivalent to a small capacitor when the switch is switched off, so that the discrete equivalent model of the switch branch circuit is ensured to be switched on and switched off in order to avoid the change of a system admittance matrix caused by the switching actionTime equivalent admittance Y_bSWEquality, in general, let the equivalent admittance Y of the switch_bSW＝1。

When the switch is turned on, the historical current source of the discrete equivalent model of the switch branch can be represented as: i is_hSW(t)＝i_SW(t-h)；

When the switch is turned on, the historical current source of the discrete equivalent model of the switch branch can be represented as: i is_hSW(t)＝-Y_bSWu_SW(t-h)

Wherein i_SW(t-h) is the branch current of the switching branch at the last simulation instant, u_SWAnd (t-h) is the branch voltage of the switching branch at the last simulation moment.

If the current simulation time t is the simulation initial time, the size of the historical current sources of all the branches is zero; and if the current value is not the simulation initial moment, calculating the size of the historical current source of each branch at the current simulation moment according to the branch voltage and the branch current of each branch at the previous simulation moment.

Step (3) according to the discretized power electronic converter system, a node admittance matrix Y is constructed_nAnd branch admittance matrix Y_b。

Step (4) using history current source I_hFor state variables, injecting a current source I_sFor input variable, node voltage V_nAnd branch current I_bConstructing a discrete state space model of the power electronic converter system for the output variable:

coefficient matrices A, B, C and D of the discrete state space model of the power electronic converter are formed according to the following rules:

wherein M is_LNumber of branches N_bX number of branches N_bWhen the branch circuit is in an inductance or switch conducting state, the corresponding element is 1, otherwise, the corresponding element is 0; m_CNumber of branches N_bX number of branches N_bWhen the branch is in a capacitance or switch off state, the corresponding element of the diagonal matrix is 1, otherwise, the diagonal matrix is 0.

And (5) selecting the simulation step size to be delta t-kh, wherein k is a positive integer. Reconstructing the state equation coefficient matrixes A and B of the discrete state space model of the power electronic converter according to the following rules to obtain coefficient matrixes:

therefore, the state equation of the discrete state space small-step synthesis model of the power electronic converter system after small-step synthesis is obtained, wherein the modeling step length is h, the simulation step length is delta t, and the state equation is as follows:

step (6) updating the historical current of each branch at the next moment according to the sizes of the historical current and the injected current at the current simulation moment and in combination with the coefficient matrixes A 'and B';

step (7) updating the node voltage of each node at the current moment according to the historical current and the injected current at the current simulation moment and in combination with the coefficient matrixes C and D;

step (8) updating the branch current of each branch at the current moment according to the historical current and the injected current at the current simulation moment and by combining the coefficient matrixes C and D;

if the simulation is not finished, the next simulation period is entered, and the step (6) is returned; and if the simulation ending time is reached, ending the simulation.

The invention has the technical effects that:

1) according to the method, the discrete state space model of the power electronic converter is constructed, the coefficient matrix of the power electronic converter is redesigned, the power electronic converter is modeled in nanosecond step length and simulated in microsecond step length, and the problem that the hardware computing capacity and the real-time simulation precision are mutually restricted is effectively solved.

2) The waveform of the power electronic switch when the method is adopted to carry out real-time simulation on the power electronic converter is closer to an ideal switch than that of the power electronic converter adopting the traditional EMTP method, and the real-time simulation precision of the power electronic converter is greatly improved. For the system-level analysis of the power electronic converter, the switches are generally considered to be ideal switches without virtual power loss, while the L/C switch model adopted in the conventional EMTP algorithm generates non-negligible virtual power loss, and the virtual power loss varies with the variation of the equivalent admittance parameter, as shown in table 1,

TABLE 1 virtual Power loss contrast

Table 1 shows the virtual power loss for two-level converters, three-level converters and five-level converters when the two-level converters, the three-level converters and the five-level converters are simulated at different switching frequencies.

Experiments show that the method can effectively solve the problem of virtual power loss in the L/C switch model, the virtual power loss is basically not influenced by switch equivalent admittance parameters and switching frequency, and the simulation precision is closer to that of a current converter consisting of ideal switches.

Drawings

FIG. 1 is a schematic diagram of a switching leg;

FIG. 2 is a schematic diagram of a two-level converter;

fig. 3 is a schematic diagram of a three-level inverter;

fig. 4 is a schematic diagram of a five-level converter;

FIG. 5 is a graph of virtual power loss for two-level, three-level, and five-level converters under two approaches when simulated at different switch equivalent admittance parameters;

FIG. 6 is a comparison of simulation results for two methods;

FIG. 7 is a schematic diagram of a simulation cycle of the method of the present invention;

fig. 8 is a flowchart of an electromagnetic transient simulation method for real-time simulation of a power electronic converter according to the present invention.

Detailed Description

For the sake of understanding, the present invention will be described below by taking a two-level inverter as shown in fig. 2 as an example, but the scope of the present invention should not be limited thereby.

When the mixed electromagnetic transient simulation method suitable for the real-time simulation of the micro-grid is used for carrying out the real-time simulation on the micro-grid, the hardware part mainly comprises the following steps: (ii) a PXIe controller (model: PXIe-8135) of National Instruments (NI) Inc. of USA: the simulation of a micro-grid control system is mainly responsible, meanwhile, the simulation can be carried out with an upper computer through the Ethernet, and real-time simulation waveforms are displayed on the upper computer. ② FPGA module (model: PXIe-7975R) of American National Instruments (NI): the simulation system is mainly responsible for the simulation of a circuit part of the micro-grid, and can be connected with an external controller and an oscilloscope through an I/O port to perform hardware-in-loop simulation. The simulation system and the simulation method are communicated through a PXIe bus to complete real-time simulation.

The software part is mainly Labview development environment of National Instruments (NI) company in the United states. Programs in the upper computer, the PXIe controller and the FPGA module are programmed through Labview. The program in the upper computer completes the functions of communication with the PXIe controller, simulation waveform display and the like; the PXIe controller has the functions of completing communication with an upper computer, reading and writing data from and into the FPGA module, simulating a control system of the converter and the like; and the program in the FPGA module completes the functions of circuit simulation cycle calculation and the like. The above-described process is not within the scope of the present invention, and the National Instruments (NI) company provides a related example of the process on the official website, and therefore will not be described in detail. The FPGA module is a specific implementation carrier of the present invention, and is shown in fig. 8 in detail, and fig. 8 is a flowchart of an electromagnetic transient simulation method for real-time simulation of the power electronic converter according to the present invention.

Referring to fig. 8, fig. 8 is a flowchart of an electromagnetic transient simulation method for real-time simulation of a power electronic converter according to an embodiment of the present invention. It can be seen from the figure that the electromagnetic transient simulation method for the real-time simulation of the power electronic converter comprises the following steps:

step (1) numbering two-level converters and branches and nodes in a circuit where the two-level converters are located respectively, wherein the number of a grounding node is 0; forming an incidence matrix K of the two-level converter system:

the row number of the incidence matrix K is the node number N_nThe number of columns is the number of branches N_bThe formation rule is as follows:

1) if node i is connected to branch j, then K (m) when current flows from node i to branch j_ij)＝1；

2) If node i is connected to branch j, then K (m) when current flows from branch j to node i_ij)＝-1；

3) K (m) if node i is not connected to branch j_ij)＝0；

Wherein, the number of nodes of the two-level converter is 11, and the number of branches is 19.

Step (2), discretizing a resistance branch, an inductance branch, a capacitance branch and a switch branch of the two-level converter system by adopting a step length h of 10ns, as shown in fig. 1;

step (3) constructing a node admittance matrix Y of the discretized two-level converter system_nAnd branch admittance matrix Y_b：

And (4) constructing a discrete state space model of the two-level converter system, and forming coefficient matrixes A, B, C and D of the discrete state space model of the two-level converter:

and (5) selecting a simulation step length delta t as 100h, wherein k is 100, and reconstructing the state equation coefficient matrixes A and B of the two-level converter discrete state space model to obtain a coefficient matrix:

therefore, a simulation calculation schematic diagram of the two-level converter system after the modeling step length h is 10ns, the simulation step length Δ t is 1 μ s, and the small-step synthesis is obtained as shown in fig. 7, and a state equation of a discrete state space small-step synthesis model is as follows:

and (6) updating the historical current of each branch at the next moment according to the sizes of the historical current and the injection current at the current simulation moment and by combining the coefficient matrixes A 'and B':

and (7) updating the node voltage of each node at the current moment according to the magnitudes of the historical current and the injected current at the current simulation moment and in combination with the coefficient matrixes C and D:

and (8) updating the branch current of each branch at the current moment according to the historical current and the injected current at the current simulation moment and by combining the coefficient matrixes C and D:

and (9) returning to the step (7), and entering the next simulation cycle:

and (5) circulating the calculation steps (6) to (9) until the simulation is finished, as shown in FIG. 8.