Three-phase voltage source converter harmonic dynamic characteristic identification method based on sparse identification

Through the harmonic dynamic characteristic recognition method of three-phase voltage source converter based on sparse recognition, the harmonic resonance problem in the new energy grid-connected system is solved, and the harmonic dynamic characteristic recognition with high accuracy is realized, effectively guiding the harmonic resonance stability Analysis and risk monitoring to ensure the safe and stable operation of the system.

CN119047394APending Publication Date: 2024-11-29STATE GRID ANHUI ELECTRIC POWER CO LTD +3
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410967675.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-18
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

The existing methods cannot effectively solve the harmonic resonance problem in new energy grid-connected systems, resulting in the impact of the power quality problem and the stability of the grid-connected inverter, especially when the system impedance network does not match, series/parallel resonance may be triggered. Endangering the safe and stable operation of the system.

Method used

The harmonic dynamic characteristic recognition method of three-phase voltage source converter based on sparse recognition is adopted. By selecting state variables and constraint candidates based on physical structure and small signal response, system characteristic data is obtained, and the sparse recognition method is used to identify the converter. Harmonic dynamic characteristics constrain the equation model and measure the accuracy of the model by determining coefficients.

Benefits of technology

This method combines physical models and mathematical driving to accurately obtain the harmonic dynamic characteristics of power electronic power converters. It has small errors and good applicable effects, and can effectively guide the harmonic resonance stability analysis and risk monitoring of new energy grid-connected systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119047394A_ABST
    Figure CN119047394A_ABST
Patent Text Reader

Abstract

The invention discloses a three-phase voltage source converter harmonic dynamic characteristic identification method based on sparse identification, belongs to the technical field of harmonic dynamic characteristic identification, and solves the harmonic resonance problem of a renewable energy three-phase voltage source converter grid-connected system. According to the physical structure and small signal response of the three-phase voltage source converter, the current response constraint of the three-phase voltage source converter under voltage disturbance is obtained, and a state variable and a constraint candidate item are selected based on the current response constraint; acquiring state data with system characteristics; identifying a harmonic dynamic characteristic constraint equation model of the converter by using a sparse identification method; measuring the accuracy of the model by adopting a decision coefficient; according to the method, a physical model and mathematical driving are combined to obtain the harmonic dynamic characteristics of the power electronic power converter, the model precision is very high, the harmonic dynamic characteristics calculated through sparse recognition have very small errors compared with a simulation result, and the method has a very good application effect on harmonic dynamic recognition.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of harmonic dynamic characteristic identification, and relates to a method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification. Background Art

[0002] With the increasing capacity of renewable energy such as photovoltaic and wind power connected to the grid, a large number of power electronic equipment are connected to the grid. Compared with traditional electromagnetic energy conversion equipment such as generators and transformers, power electronic power conversion equipment has huge differences in physical structure, control method, dynamic response, etc. When there is a resonance peak in the system and the background harmonics match the resonance frequency, it will cause resonant overvoltage and overcurrent in large photovoltaic grid-connected systems, causing power quality problems such as excessive harmonic content. At the same time, the resonance phenomenon also has a great impact on the stable operation of the grid-connected inverter. The series / parallel resonance caused by the mismatch of the system impedance network may even cause the grid-connected inverter to trip without fault, endangering the safe and stable operation of the system.

[0003] Impedance analysis is generally considered to be an effective method for harmonic analysis, and how to accurately obtain the harmonic impedance characteristics of the converter is a key issue. At present, the converter harmonic impedance calculation method based on physical drive is very complex, and there is a certain error in the linear processing of the signal. It is very difficult to apply this method to systems with complex control loops. In addition, the converter harmonic impedance acquisition method based on data drive lacks model interpretability and is often used in real-time online harmonic risk monitoring.

[0004] In summary, the existing methods cannot provide good guidance for harmonic resonance stability analysis and harmonic resonance risk analysis of renewable energy grid-connected systems, as well as solving power quality problems of renewable energy grid-connected systems. Summary of the invention

[0005] The technical solution of the present invention is used to solve the harmonic resonance problem of a renewable energy three-phase voltage source converter grid-connected system.

[0006] The present invention solves the above technical problems through the following technical solutions:

[0007] A method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification includes the following steps:

[0008] Step 1: According to the physical structure and small signal response of the three-phase voltage source converter, the current response constraint under the voltage disturbance is obtained, and the state variables and constraint candidates are selected based on the constraint constraint;

[0009] Step 2: Obtain status data with system characteristics;

[0010] Step 3: Using a sparse identification method to identify the harmonic dynamic characteristic constraint equation model of the converter;

[0011] Step 4: Use the coefficient of determination to measure the accuracy of the model.

[0012] Furthermore, according to the physical structure and small signal response of the three-phase voltage source converter in step 1, the current response constraint under the voltage disturbance is obtained, and the method for selecting state variables and constraint candidates based on this is specifically as follows:

[0013] According to the physical structure and small signal response of the three-phase voltage source converter, its current response constraint under voltage disturbance is obtained as follows:

[0014]

[0015] Where m a 、m b 、m c are the three-phase pulse width modulation reference signals, K m is the modulator gain;

[0016] The state variables for sparse identification are selected as follows:

[0017]

[0018] Constraint candidate selection: Select existing constraint candidates based on state variables X and U, and select all combinations of state variables, including single state variables and combinations of state variables, totaling 63, as follows:

[0019]

[0020] Furthermore, the specific method for obtaining state data with system characteristics described in step 2 is: first, let the converter transient simulation model run to a steady state, then inject a small signal disturbance of three-phase voltage or current from the AC measurement, and run it to a steady state again, and finally measure the corresponding current or voltage. Finally, according to the superposition principle, the response signal with and without the injected disturbance is subtracted to obtain the response caused by the small signal.

[0021] Furthermore, the method of using the sparse identification method in step 3 to identify the harmonic dynamic characteristic constraint equation model of the converter is specifically as follows:

[0022] Using the acquired harmonic characteristic data, the sparse coefficient matrix is ​​identified through the sparse identification method, and the sparse coefficient matrix is ​​substituted into the formula The harmonic dynamic characteristic constraint equation model of the converter is as follows:

[0023] s 2 I p =k1V dc s 3 V p3 +k2V dc s 3 V p 2 I p +k3V dc s 3 V p 2 +k4V dc s 3 V p I p 2

[0024] Among them, s is the Laplace operator, I p is the disturbance current, k1, k2, k3, k4 are sparse identification coefficients, V p is the disturbance voltage, V dc is the converter DC voltage.

[0025] Furthermore, the calculation formula of the determination coefficient described in step 4 is as follows:

[0026]

[0027] Among them, R 2 is the coefficient of determination, S res is the residual sum of squares of the data, S tot is the total sum of squares of the data, h is the regression value, y i is the measured value, is the average value.

[0028] An electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above-mentioned method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification, and the processor is configured to execute the program stored in the memory.

[0029] A storage medium stores a computer program, which, when executed by a processor, executes the steps of the above-mentioned method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification.

[0030] The advantages of the present invention are:

[0031] The present invention obtains the current response constraint of the three-phase voltage source converter under voltage disturbance according to its physical structure and small signal response, and selects state variables and constraint candidates based on this; obtains state data with system characteristics; applies sparse recognition method to identify the harmonic dynamic characteristic constraint equation model of the converter; and uses determination coefficient to measure the accuracy of the model; the present invention combines the physical model with mathematical drive to obtain the harmonic dynamic characteristics of the power electronic power converter, the model has high accuracy, and the harmonic dynamic characteristics calculated by sparse recognition have a small error compared with the simulation result. The method of the present invention has a good applicable effect in harmonic dynamic recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 A flowchart of a method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification;

[0033] Figure 2 is a topology diagram of a three-phase voltage source converter system;

[0034] Figure 3 This is the effect diagram of the dynamic identification of harmonic response current. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and specific embodiments:

[0037] Embodiment 1

[0038] 1. Three-phase voltage source converter

[0039] like Figure 1 As shown in FIG. 1 , the three-phase voltage source converter of the present invention adopts current closed-loop control, wherein S a1 , S a2 , S b1 , S b2 , S c1 , S c2 are the switching signals of the three-phase upper bridge arm and lower bridge arm respectively, V dc is the converter DC voltage, v a 、v b 、v c 、i a 、i b 、ic are three-phase voltage and three-phase current, I dr ,I qr are d-axis reference current and q-axis reference current respectively, K f is the voltage feedback gain, K f =1÷(K m ×V dc ), K m is the modulation gain, θ is the AC phase of the phase-locked loop (PLL) output, and L is the reactor.

[0040] In addition to the system fundamental voltage, the AC part of the converter also contains harmonic disturbance voltage that needs to be injected. Its time domain signal is:

[0041]

[0042] Its frequency domain signal is:

[0043]

[0044] Among them, V1 is the amplitude of the fundamental signal, f1 is the frequency of the fundamental signal, V p is the amplitude of the harmonic disturbance signal, f p is the harmonic disturbance signal frequency, φ vp is the phase of the harmonic disturbance signal, and V p Much smaller than V1.

[0045] 2. Principle of sparse recognition method

[0046] For a nonlinear dynamic system with control, its dynamic characteristics can be expressed as:

[0047]

[0048] where x(t)=[x1(t)x2(t)…x n (t)] T is the state of the system at time t; u(t)=[u1(t)u2(t)…u n (t)] T is the input of the dynamic system at time t; f(x(t),u(t)) is the constraint of the dynamic system with control.

[0049] The specific form of the function f(x(t),u(t)) is determined by data-driven and physical knowledge, and it is necessary to collect t1, t2, t3…t with dynamic system characteristics. n The state data x at each moment and the state matrix are constructed as follows:

[0050]

[0051] in, is the differential of X, where X is the data obtained by simulation or online and has the characteristics of the dynamic system.

[0052] Then, based on the physical knowledge of the dynamic system, possible function candidates of the system are constructed. The function candidates can be composed of polynomials, trigonometric functions, etc. composed of the system state and input state:

[0053]

[0054] Among them, X1 and X2 are the state variables of the dynamic system, U1 is the control input of the dynamic system, and each column of Θ(X,U) represents a vector of candidate items. In fact, the dynamic system constraints containing control are only described by the combination of some state variables, which makes the control equations in the high-dimensional nonlinear function space sparse. This sparse identification method can find the sparse constraint equations of the system from a complex and unknown dynamic system. This method balances the complexity and accuracy of the model and obtains the simplest dynamic equation as follows:

[0055]

[0056] Where Ξ=[ξ1ξ2…ξ n ] is the identified sparse matrix, ξ k Represents the coefficient of the k-th state function in Θ(X,U).

[0057] The objective function of the sparse recognition method is defined by L1 norm regression, which is expressed as follows:

[0058]

[0059] Among them, λ is the penalty factor, which is used to balance the complexity and accuracy of the model.

[0060] 3. Figure 2 As shown, a method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification includes:

[0061] Follow these steps:

[0062] Step 1: Select state variables and constraint candidates as follows:

[0063] According to the physical structure and small signal response of the three-phase voltage source converter, its current response constraint under voltage disturbance is obtained as follows:

[0064]

[0065] Where m a 、m b 、m c are the three-phase pulse width modulation reference signals, K m is the modulator gain.

[0066] The state variables for selecting sparse recognition are:

[0067]

[0068] Constraint candidate selection, select possible constraint candidates based on state variables X and U. Here, all combinations of state variables are selected, including single state variables and combinations of state variables, totaling 63, as follows:

[0069]

[0070] It is necessary to select key state variables that can describe the dynamic system with control according to the dynamic characteristics of the converter. The selected state variables should highly reflect the state of the corresponding system. Since the object to be identified is the harmonic dynamic characteristics of the converter, the selected state variables should at least include i p (harmonic current), v p (Harmonic voltage). And according to the physical structure and control method of the converter, select the possible constraint candidates for its harmonic dynamic characteristics. The selection of constraint candidates profoundly affects the accuracy of the recognition results. The key lies in two points. First, the combination of candidates must be based on the physical characteristics and control method of the converter, otherwise the constraint equations of the converter cannot be identified, or the recognition result error is large; second, the complexity of the candidate must match the complexity of the data. If the complexity of the data used for recognition is high and the complexity of the candidate is low, then the correct model cannot be identified. If the complexity of the data is low and the complexity of the candidate is high, overfitting problems will easily occur.

[0071] Step 2: Obtain harmonic characteristic data with system characteristics.

[0072] In order to obtain system dynamic data for identification, build Figure 1 The transient simulation model shown in the figure has a reactor L = 3.0 × 10 -3 H; Phase voltage peak DC voltage V dc =550V; d-axis reference current I dr =10A, q-axis reference current I qr =0A; modulation gain K m =1 / 2; the current controller is a proportional integral controller (PI controller), and the transfer function is:

[0073]

[0074] The proportional gain K P =0.04578, integral gain K i =43.1.

[0075] The method of obtaining data based on the transient simulation model is as follows. First, the converter simulation model needs to run to a steady state; then a small signal disturbance of the three-phase voltage (or current) is injected from the AC measurement, and it runs to a steady state again; finally, the corresponding current (or voltage) is measured; finally, according to the superposition principle, the response signal with and without the injected disturbance is subtracted to obtain the response caused by the small signal. It should be noted that the amplitude of the disturbance signal needs to be appropriately selected. If the amplitude of the disturbance signal is too large, the converter will deviate too far from the steady-state operation point; if the disturbance signal is too small, it is easy to be affected by the interference signal. The amplitude is recommended to be selected between 0.1 and 0.5% of the voltage (or current) reference value.

[0076] When acquiring data, the acquired data must be able to contain the dynamic characteristics of the system. If the acquired data fails to contain all the dynamic characteristics of the system, then the dynamic constraint equations identified using the sparse identification method will not be able to reflect all the dynamic characteristics of the system accordingly.

[0077] Step 3: Use the sparse identification method to identify the harmonic dynamic characteristic constraint equation model of the converter.

[0078] Using the acquired harmonic characteristic data, the sparse coefficient matrix is ​​identified by the sparse identification method. The sparse coefficient matrix Ξ is shown in Table 1. Substituting the sparse coefficient matrix into the formula The harmonic dynamic characteristic constraint equation model of the converter is as follows:

[0079] s 2 I p =k1V dc s 3 V p 3 +k2V dc s 3 V p 2 I p +k3V dc s 3 V p 2 +k4V dc s 3 V p I p 2 (15)

[0080] Among them, s is the Laplace operator, I p is the disturbance current, k1, k2, k3, k4 are sparse identification coefficients, V p is the disturbance voltage, V dc is the converter DC voltage.

[0081] like Figure 3As shown in the figure, the recognition result is compared with the simulation data collection result. It can be seen from the figure that the harmonic dynamic characteristics calculated by sparse recognition have a small error compared with the simulation results. This method has a good applicability effect in harmonic dynamic recognition.

[0082] Table 1 Sparse coefficient matrix

[0083]

[0084] Step 4: Verify the accuracy of the obtained model. If the result is accurate, the harmonic impedance characteristics of the converter can be inferred; if the result is inaccurate, it is necessary to return to step 1 and re-examine the selected state variables, constraint candidates and the scientificity of data collection, and modify them until an accurate result is obtained. The present invention adopts the determination coefficient R 2 To measure the accuracy of the model, as shown in equations (17) to (19), when the model accuracy is high, R 2 Tends to 1.

[0085]

[0086] Among them, S res is the residual sum of squares of the data, S tot is the total sum of squares of the data, h is the regression value, y i is the measured value, and y is the average value. After calculation, the model determination coefficient R 2 =0.999999231214, indicating that the model is very accurate.

[0087] Embodiment 2

[0088] An electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification in Example 1, and the processor is configured to execute the program stored in the memory.

[0089] Embodiment 3

[0090] A storage medium stores a computer program, which, when executed by a processor, executes the steps of the method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification in embodiment 1.

[0091] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification, characterized in that: The following steps are involved: Step 1: According to the physical structure and small signal response of the three-phase voltage source converter, the current response constraint under the voltage disturbance is obtained, and the state variables and constraint candidates are selected based on the constraint constraint; Step 2: Obtain status data with system characteristics; Step 3: Using a sparse identification method to identify the harmonic dynamic characteristic constraint equation model of the converter; Step 4: Use the coefficient of determination to measure the accuracy of the model.

2. The method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification according to claim 1 is characterized in that: According to the physical structure and small signal response of the three-phase voltage source converter described in step 1, the current response constraint under the voltage disturbance is obtained. Based on this, the method of selecting state variables and constraint candidates is as follows: According to the physical structure and small signal response of the three-phase voltage source converter, its current response constraint under voltage disturbance is obtained as follows: Where m a 、m b 、m c are the three-phase pulse width modulation reference signals, K m is the modulator gain; The state variables for sparse identification are selected as follows: Constraint candidate selection: Select existing constraint candidates based on state variables X and U, and select all combinations of state variables, including single state variables and combinations of state variables, totaling 63, as follows:

3. The method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification according to claim 2 is characterized in that: The specific method for obtaining state data with system characteristics described in step 2 is: first, let the converter transient simulation model run to a steady state, then inject a small signal disturbance of three-phase voltage or current from the AC measurement, and run it to a steady state again, and finally measure the corresponding current or voltage. Finally, according to the superposition principle, the response signal with and without the disturbance is subtracted to obtain the response caused by the small signal.

4. The method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification according to claim 3 is characterized in that: The method of using the sparse identification method to identify the harmonic dynamic characteristic constraint equation model of the converter described in step 3 is specifically as follows: Using the acquired harmonic characteristic data, the sparse coefficient matrix is ​​identified through the sparse identification method, and the sparse coefficient matrix is ​​substituted into the formula The harmonic dynamic characteristic constraint equation model of the converter is as follows: s 2 I p =k1V dc s 3 V p 3 +k2V dc s 3 V p 2 I p +k3V dc s 3 V p 2 +k4V dc s 3 V p I p 2 Among them, s is the Laplace operator, I p is the disturbance current, k1, k2, k3, k4 are sparse identification coefficients, V p is the disturbance voltage, V dc is the converter DC voltage.

5. The method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification according to claim 4 is characterized in that: The coefficient of determination described in step 4 is calculated as follows: Among them, R 2 is the coefficient of determination, S res is the residual sum of squares of the data, S tot is the total sum of squares of the data, h is the regression value, y i is the measured value, is the average value.

6. An electronic device, comprising a memory and a processor, characterized in that: The memory is used to store a program that supports the processor to execute the method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification as described in any one of claims 1 to 5, and the processor is configured to execute the program stored in the memory.

7. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for identifying harmonic dynamic characteristics of a three-phase voltage source converter based on sparse identification according to any one of claims 1 to 5 are executed.