A Stability Analysis Method for MMC-Based Grid Converter Hybrid Systems
By using a layer-by-layer stability analysis method, the problem of judging the overall small-signal stability of the MMC-based grid converter hybrid system was solved. This method achieves multi-frequency disturbance mapping and port relationship preservation without changing the physical relationship, and can locate sensitive subsystems or interconnection links.
Patent Information
- Application Number
- CN202611576934.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-13
- Publication Date
- 2026-11-10
AI Technical Summary
Existing technologies make it difficult to directly determine the overall small-signal stability of a hybrid MMC-grid converter system without changing the physical relationship between grid-type voltage source and grid-type current source. Furthermore, single average gain or equivalent control delay methods are insufficient to explicitly describe multi-frequency disturbance mapping.
A layer-by-layer stability analysis method is adopted to establish models of the controlled voltage source and output impedance of the grid-type MMC, the combined model of the controlled current source and input admittance of the grid-connected MMC, the load input admittance model, and the grid input model. The stability is judged layer by layer by the generalized Nyquist criterion, and the multi-frequency coupling relationship of modulation reference disturbance, switching time disturbance and bridge arm voltage disturbance is preserved.
Without altering the physical relationships, explicitly describing the multi-frequency disturbance mapping can distinguish the source-type port relationships and port boundaries of parallel branches, locate sensitive subsystems or interconnection links, and achieve overall small-signal stability judgment.
Smart Images

Figure CN123073733A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of electronic information technology, and specifically to a stability analysis method for a hybrid MMC-based grid converter system. Background Technology
[0002] The internal dynamics of a Modular Multilevel Converter (MMC) include bridge arm circulating current, submodule capacitor voltage variations, and the relationships between common-mode and differential-mode variables. Harmonic state-space modeling methods can describe the coupling between these variables through frequency expansion and can be connected to control loops such as phase-locked loops, AC current control, DC voltage control, and circulating current suppression control.
[0003] During PWM modulation, the switching timing changes with the disturbance of the modulation reference. The PWM modulation multi-frequency model describes the relationship between the modulation reference disturbance and the modulation response at different frequencies based on the actual MMC modulation method. For modulation methods using carrier comparison, the model is established through the linearization of the intersection of the modulation reference and the carrier, and the frequency expansion of the switching function disturbance. Using a single average gain or equivalent control delay method is insufficient to explicitly describe the above multi-frequency disturbance mapping. Furthermore, the specific circuit equivalents and return ratios of existing systems cannot be directly applied to grid-type MMCs, grid-connected MMCs, and grids and loads connected in parallel at the same common coupling point, making it impossible to complete the overall small-signal stability assessment without changing the physical relationship between grid-type voltage source and grid-connected current source types. Summary of the Invention
[0004] In view of this, the present invention provides a stability analysis method for MMC-grid converter hybrid systems to solve the problems that the single average gain or equivalent control delay method is difficult to explicitly describe the above-mentioned multi-frequency disturbance mapping, and that the existing technology cannot complete the overall small-signal stability judgment without changing the physical relationship between the grid-type voltage source and the grid-type current source.
[0005] A stability analysis method for a hybrid MMC-grid converter system is presented. This hybrid system comprises four types of branches connected in parallel to the AC common coupling point (PCC): the power grid, the grid-type MMC, the grid-connected MMC, and the load. The grid-type MMC employs a controlled voltage source and output impedance model, the grid-connected MMC employs a controlled current source and input admittance combination model, and the load employs an input admittance model. The PWM multi-frequency mapping obtained based on the actual modulation methods of the two types of MMCs is embedded into the corresponding converter closed-loop model, and is not treated as an independent subsystem or independent analysis layer.
[0006] The method comprises the following five steps in sequence: S1, establishing basic subsystem models of the power grid, grid-type MMC, grid-connected MMC, and load in the hybrid system; S2, determining the stability of the basic subsystems of the power grid, grid-type MMC, grid-connected MMC, and load in the hybrid system respectively; S3, determining the stability of the coupled subsystem composed of the grid-connected MMC and load; S4, determining the stability of the hybrid subsystem composed of the grid-type MMC, grid-connected MMC, and load; S5, determining the stability of the overall hybrid system composed of the power grid, grid-type MMC, grid-connected MMC, and load. Step S1 is not the stability analysis layer. The first layer is the basic subsystem layer, using the load as the starting point for subsequent combinations, and simultaneously checking the stability of the power grid, grid-type MMC, grid-connected MMC, and load; the second layer combines the grid-connected MMC3 with the load to form the first coupled system; the third layer adds the grid-type MMC to the first coupled system to form the second coupled system; the fourth layer interconnects the power grid with the second coupled system to form the overall interconnected system.
[0007] In step S1, the internal circuits and actual control relationships of the two types of MMCs are linearized using small-signal methods, and the coupling between the retained frequencies is organized using a harmonic state-space model. A PWM multi-frequency mapping is formed based on the actual MMC modulation method, from the modulation reference disturbance to the actual modulation response disturbance and then to the bridge arm voltage disturbance. This mapping is then embedded into a closed-loop model composed of the control model and the MMC power circuit model. The MMC power circuit harmonic state-space model is applicable to both networked and gridded MMCs. The difference between the two lies in the control model and the source-type port relationship. The networked MMC is represented as a controlled voltage source and a closed-loop output impedance model, while the gridded MMC is represented as a controlled current source and a closed-loop input admittance combination model.
[0008] Steps S2 to S5 follow the organization of layer-by-layer stability analysis, performing basic subsystem checks, coupled system combinations, and overall interconnection in the following order: the first coupled system consisting of load, grid-connected MMC, and load; the second coupled system consisting of grid-connected MMC and load; and the overall hybrid system consisting of power grid, grid-connected MMC, and load. The first layer performs stability assessments on the power grid, grid-connected MMC, grid-connected MMC, and load respectively. The second and third layers establish loop gain matrices formed by multiplying the corresponding input admittance matrix with the load's input impedance matrix. The fourth layer combines the power grid and grid-connected MMC into a controlled voltage source and equivalent output impedance, establishing the corresponding loop gain matrix. Each layer uses the generalized Nyquist criterion to determine stability; only when the previous layer meets the stability condition is the new interconnection effect of the next layer further explained, and the output of the corresponding sensitive subsystem or interconnection link with reduced stability margin is based on the judgment results of the adjacent two layers, without pre-limiting it to a specific interface.
[0009] This invention extends the layer-by-layer stability analysis approach to parallel systems that simultaneously incorporate MMC internal harmonics, mesh-type control, follow-mesh control, and PWM frequency coupling. Without altering the physical relationships between the GFM voltage source and GFL current source, the basic subsystem, the first coupled system, the second coupled system, and the overall interconnected system are examined sequentially, with GNC applied at each layer. The key technical focus of this invention is the integration of the layer-by-layer method with the analyzed object and the multi-frequency closed-loop model; each layer's matrix is used only to express the corresponding port boundaries and feedback relationships. Without actual calculation or measurement data, no definitive stability margin, instability frequency, or performance improvement is provided.
[0010] The stability analysis method for MMC-based grid converter hybrid systems provided by the present invention has the following beneficial effects: First, addressing the problem that existing technologies using a single average gain or equivalent control delay cannot explicitly describe the PWM multi-frequency disturbance mapping, this invention establishes a PWM modulation multi-frequency model based on the actual MMC modulation method and embeds the PWM frequency coupling into the closed-loop harmonic state-space model of two types of MMCs. Therefore, it can explicitly preserve the multi-frequency coupling relationship between modulation reference disturbance, switching timing disturbance, switching function disturbance, and bridge arm voltage disturbance. Second, addressing the problem that the specific circuit equivalents and return ratios of existing systems cannot be directly used for grid-type MMCs, follow-up grid-type MMCs, and grid and load connected in parallel at the same common coupling point, this invention establishes controlled voltage source and output impedance models for grid-type MMCs, controlled current source and input admittance combination models for follow-up grid-type MMCs, load input admittance models, and grid input models, respectively. Therefore, it preserves the source-type port relationships and port boundaries when the four types of branches are connected in parallel at the common coupling point. Furthermore, addressing the issue of locating sensitive components in layer-by-layer analysis, steps S2 to S5 sequentially combine components layer by layer in the following order: the first coupled system consisting of the load, the grid-connected MMC and the load; the second coupled system consisting of the grid-connected MMC and the load; and the overall hybrid system consisting of the power grid, the grid-connected MMC, and the load. Each layer establishes a corresponding loop gain matrix based on the stability conditions of the previous layer and employs the generalized Nyquist criterion. Therefore, it can distinguish the newly added interconnection effects of the basic subsystem, the first coupled system, the second coupled system, and the overall interconnected system, and output the corresponding sensitive subsystem or interconnection link with reduced stability margin. Finally, this invention combines the power grid and the grid-connected MMC into a controlled voltage source and equivalent output impedance, and uses the number of poles in the right half-plane achieved by compatible open-loop control to determine overall stability. Therefore, it can complete the overall small-signal stability judgment without changing the physical relationship between the grid-connected voltage source and the grid-connected current source. Attached Figure Description
[0011] Figure 1 This is a structural block diagram of a hybrid system; Figure 2A flowchart illustrating the stability analysis method for a hybrid MMC-grid converter system; Figure 3 This is a comparison chart of the theoretical curve and frequency sweep results of the MMC port impedance. Detailed Implementation
[0012] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain embodiments of the present invention, and should not be construed as limiting the present invention.
[0013] Embodiments of the present invention provide a stability analysis method for a hybrid MMC-based grid converter system. Please refer to [link to relevant documentation]. Figure 1 The hybrid system includes a power grid 1, a grid-type MMC 2, a grid-connected MMC 3, and a load 4 connected in parallel at the AC common coupling point, forming a four-layer structure. The first layer is the basic subsystem layer, with 4 as the starting point for combination. The second layer is the first coupling system composed of the grid-connected MMC 3 and the load 4. The third layer is the second coupling system formed after the addition of the grid-type MMC 2. The fourth layer is the overall interconnected system formed after the addition of the power grid 1.
[0014] Please see Figure 2 The method includes steps S1 to S5: S1. Establish basic subsystem models for the power grid, grid-type MMC, grid-connected MMC, and loads in the hybrid system.
[0015] Among them, a closed-loop small-signal model containing PWM frequency coupling is formed based on the internal circuit, control relationship and actual PWM modulation method of each multilevel converter. Under the common port reference and frequency coordinate, the controlled voltage source and output impedance model of the grid-type MMC and the controlled current source and input admittance combination model of the grid-connected MMC are obtained, and the input models of the power grid and load are established.
[0016] In step S1, common-mode and differential-mode variables are used to describe the arm current, submodule capacitor voltage, and modulation amount of the MMC. Small-signal linearization is performed near the operating trajectory, and the coupling between the retained frequency components is described using the harmonic state-space form. The harmonic state-space model of the same MMC power circuit is applicable to both grid-type and follow-grid-type MMCs. The difference between the two is formed by the control model and the source port relationship. The PWM frequency coupling is embedded in the closed-loop harmonic state-space model of the strain gauge current transformer and is not treated as an independent analysis subsystem.
[0017] Specifically, the network-type MMC 2 uses a controlled voltage source and closed-loop output impedance, the network-type MMC 3 uses a combination of a controlled current source and closed-loop input admittance, load 4 uses input admittance, and grid 1 uses an equivalent voltage source and output impedance, expressed as follows:
[0018]
[0019]
[0020]
[0021] in, For controlled voltage source disturbances in network-type MMC. and These are the port voltage disturbance and port current disturbance of the network-type MMC, respectively. This is a closed-loop output impedance matrix that includes the internal dynamics of the network-type MMC, droop control, phase angle dynamics, circulating current suppression control, actual virtual impedance control and / or current inner loop, and PWM frequency coupling. For complex frequency variables, and These are the port current disturbance and port voltage disturbance of the mesh-type MMC, respectively. To accommodate the controlled current source disturbance of the grid-type MMC. It is a closed-loop input admittance matrix that includes a phase-locked loop, an outer DC voltage loop, an inner AC current loop, circulating current suppression control, internal dynamics of the MMC, and PWM frequency coupling. To absorb current disturbances at the ports of the load branch. The load input admittance matrix, For the port voltage disturbance of the load branch, For the port voltage disturbance of the power grid branch, This refers to the equivalent voltage source disturbance of the power grid. The output impedance matrix of the power grid. For the port current disturbance of the power grid branch.
[0022] For the two types of MMC power circuits, common-mode, differential-mode, and bridge arm internal state variables are linearized to form state, modulation, and port input matrices. The conventional item-by-item relationships of bridge arm current, submodule capacitor voltage, and common-mode-differential-mode coupling are not elaborated, but only their effects on the port model and multi-frequency coupling are retained. Modulation disturbances, port voltage disturbances, and internal state disturbances are all miniaturized according to the actual operating trajectory.
[0023] When the running trajectory and the actual modulation timing have a common period season and index the set according to its frequency. Expanding the perturbation, we get:
[0024]
[0025]
[0026] in, The imaginary unit, For time, A set of indices with the same frequency. For output frequency index, This is the small-signal time-domain perturbation vector of the internal state variables of the MMC power circuit. For the corresponding dimension, , and These are frequency-stacked vectors representing the state, actual modulation response, and port input perturbation, respectively. Frequency Index The corresponding state perturbation phasor The state matrix is assembled from the Fourier coefficients of the periodic matrix. and The modulation input matrix and port input matrix are assembled from the Fourier coefficients of the periodic matrix. A diagonal matrix stacked for frequencies. Represents a diagonal matrix. Common angular frequency, For a common cycle.
[0027] When there is no common period, a compatible frequency set is established based on the actual timing sequence, instead of directly applying the above formula. The harmonic state-space model of the MMC power circuit constructed by the above formula is applicable to both networked MMC 2 and rooted MMC 3. The difference between the two is formed by the control model and the source port relationship.
[0028] The grid-type MMC 2 employs active-frequency and reactive-voltage droop control, transforming active and reactive power disturbances into controlled voltage sources through power measurement, droop control, phase angle dynamics, and port voltage mapping, while satisfying the following:
[0029]
[0030] in, The controlled voltage source dynamic mapping matrix consists of power measurement, droop control, phase angle dynamics, and port coordinate transformation. and These are active power disturbances and reactive power disturbances, respectively. This indicates transpose.
[0031] The grid-type MMC 3 employs a phase-locked loop, an outer DC voltage loop, an inner AC current loop, and circulating current suppression control to form controlled current sources from DC voltage disturbances, active power disturbances, and reactive power disturbances, while satisfying the following:
[0032]
[0033] in, The dynamic mapping matrix of the controlled current source. This is a DC voltage disturbance.
[0034] In this embodiment, both types of MMC establish PWM modulation multi-frequency models based on the actual MMC modulation method. For the modulation process that uses continuous modulation reference and carrier comparison and has non-degenerate steady-state intersection, the switching time disturbance is determined based on the slope difference between the modulation reference and carrier at the steady-state intersection, the switching function disturbance is determined based on the switching jump variable, and the multi-frequency mapping matrix is formed by the switching function disturbance. This allows the PWM modulation dynamics to be embedded into the converter closed-loop harmonic state space model in the order of modulation reference disturbance, actual modulation response disturbance, and bridge arm voltage disturbance.
[0035] Specifically, for the continuous modulation reference and carrier comparison method, with Indicates the first Taiwan MMC's A comparison channel modulation reference, with Indicates the corresponding carrier wave, with Indicates the first The time corresponding to each steady-state intersection point. The non-degenerate intersection points and their linearized relationship are as follows:
[0036]
[0037]
[0038] in, For MMC index, To compare channel indexes, For switch event index, As a reference for trajectory modulation, For the first Taiwan MMC's The carrier signal corresponding to the first comparison channel is in the first... At each steady-state intersection point The value at that location, For the disturbance at the switching time, For the first Taiwan MMC's The modulation reference signal of the first comparison channel is in the... At each steady-state intersection point Small signal perturbation at the location. The difference in slope on both sides of the intersection point. and These represent the time derivatives of the carrier and modulation reference, respectively; for modulation schemes that do not employ carrier comparison, a similar mapping is established based on the actual switching event conditions.
[0039] Let the first The jump variable of the switch function for each event is The frequency mapping coefficients, multi-frequency mapping matrix, and dynamic mapping from modulation reference to actual insertion amount are formed by the perturbation at the switching time:
[0040]
[0041]
[0042]
[0043]
[0044] in, and These are the event-side band index and the carrier comparison-side band index, respectively. For frequency mapping coefficients, It is a multi-frequency mapping matrix. Indexed by event-side frequency Frequency index compared with carrier side The frequency mapping coefficients determined by the difference and These are the frequency stacked vectors of the modulation reference and the switching function perturbation, respectively. and These are the modulation reference allocation matrix and the synthesis matrix from the switching function to the actual insertion amount, respectively. For PWM closed-loop mapping, Indicates the first Taiwan MMC's PWM multi-frequency mapping matrix, This represents the frequency stacking vector representing the actual modulation response perturbation of the MMC. This is the actual modulation response perturbation frequency stacking vector without embedded PWM multi-frequency mapping. PWM mapping changes. and However, it does not form a separate basic subsystem, coupled subsystem, or stability analysis layer.
[0045] The above control equations are combined, and the port current input and port voltage output of the network-type MMC and the port voltage input and port current output of the root-type MMC are extracted. , , and During the extraction process, the internal harmonic state, control dynamics, circulating current suppression control, and PWM multi-frequency coupling of the MMC are preserved. If two MMCs share an internal network, the common network is first assembled according to the actual connection relationship, and then their respective port models are formed.
[0046] S2 determines the stability of the power grid, the grid-type MMC, the grid-connected MMC, and the load base subsystem in the hybrid system, respectively.
[0047] The system takes the load as the starting point for layer-by-layer combination, and takes the power grid, the grid-type MMC, the grid-connected MMC and the load as basic subsystems. Based on the closed-loop port model of each basic subsystem, a corresponding return difference matrix is formed. The stability of each basic subsystem is judged by the generalized Nyquist criterion, and the basic subsystems that do not meet the stability conditions are output.
[0048] In this embodiment, the first layer is the basic subsystem layer. Taking load 4 as the starting point for subsequent combinations, the independent closed-loop stability of grid 1, grid-type MMC 2, grid-connected MMC 3, and load 4 are checked simultaneously. Load branches are defined using port absorption direction according to the input admittance model, and their port relationships are as follows: The specific dynamics of the load are determined by the actual system.
[0049] For each basic subsystem, a return difference matrix and a return difference determinant are established based on its actual closed-loop port boundaries. The generalized Nyquist criterion is applied to each return difference determinant, and internal dynamics, interconnect well-posedness, and boundary poles are checked. PWM multi-frequency relationships are already embedded in two types of MMC closed-loop models and do not require a separate analysis layer. The first layer decision passes and the process proceeds to the second layer only when all four basic subsystems meet the stability conditions.
[0050] S3 determines the stability of the coupled subsystem consisting of the network-type MMC and the load.
[0051] The first coupling system is formed by combining the grid-type MMC with the load impedance. The loop gain matrix is established by multiplying the input admittance of the grid-type MMC with the input impedance of the load according to the voltage relationship at the common coupling point and the current balance relationship under the unified port direction. The stability of the first coupling system is judged by the generalized Nyquist criterion, and the corresponding sensitive subsystem or interconnection is output with reduced stability margin.
[0052] Specifically, the second layer will form the first coupled system with the network-type MMC3 and the load. The port voltages of both will be PCC voltage disturbances. Established according to the direction of common feedback algebra:
[0053] in, This refers to the current disturbance at the external port of the first coupled system. To mitigate port current disturbances in mesh-type MMCs. For the port current disturbance of the load. To control the current source disturbance of the grid-type MMC.
[0054] For the MMC branch in the grid-connected configuration, the current is defined according to the direction injected from the branch to the common coupling point; for the load branch, the current is defined according to the port absorption direction of the input admittance model; the port voltages of the grid-connected MMC and the load are unified as the common coupling point voltage disturbance. Let the load input impedance matrix And establish the following formula:
[0055]
[0056]
[0057] in, To establish the loop gain matrix between the network-type MMC and the load, To return the difference matrix, For the corresponding dimension, To return the difference determinant, The load input admittance matrix, The determinant is represented by ; after the stability test of the first coupled system, its external input admittance is denoted as . ,in Let be the external input admittance matrix of the first coupled system. To establish the closed-loop input admittance matrix of the mesh-type MMC, The input admittance matrix is the load.
[0058] When irreversible, a compatible return difference matrix is formed based on the complete closed loop of the same interconnect boundary, without forcibly using it. Under the premise of first-level stability, GNC is used; if the stability condition is not met at this layer, the output stability margin is reduced for the corresponding sensitive subsystem or interconnection.
[0059] S4 determines the stability of the hybrid subsystem consisting of the grid-type MMC, the interconnected MMC, and the load.
[0060] In this process, a network-type MMC is added to the first coupling system to form a second coupling system. The controlled voltage source and output impedance relationship of the network-type MMC are retained. The loop gain matrix is obtained by multiplying the input admittance of the first coupling system with the output impedance of the network-type MMC. The stability of the second coupling system is judged by the generalized Nyquist criterion, and the corresponding sensitive subsystem or interconnection is reduced by outputting the stability margin.
[0061] Specifically, the third layer adds the network-type MMC to the first coupling system to form the second coupling system. The port voltage of the network-type MMC is set to the PCC voltage perturbation. And establish it according to the unified port direction:
[0062]
[0063]
[0064] Therefore, we get:
[0065]
[0066]
[0067]
[0068] in, The feature matrix of the hybrid source of the network-type MMC and the first coupled system. This is the gain matrix of the third-layer loop. To return the difference matrix, The return difference determinant of the second coupled system; by and The stability of the second coupled system is determined using the generalized Nyquist criterion.
[0069] Under the premise of stability of the first coupling system GNC will be used; if stability conditions are met, , and Pass it to the fourth level.
[0070] S5 determines the stability of the overall interconnected system consisting of the power grid, the grid-type MMC, the grid-connected MMC, and the load.
[0071] In this process, the power grid is added to the second coupling system to form an overall interconnected system. The power grid and the grid-type MMC are combined into a controlled voltage source and an equivalent output impedance. The loop gain matrix is obtained by multiplying the input admittance of the first coupling system with the equivalent output impedance. The overall stability is judged by using the generalized Nyquist criterion and the number of poles in the right half plane realized by compatible open loop. The corresponding sensitive subsystem or interconnection link with reduced stability margin is output.
[0072] Specifically, the power grid and the grid-type MMC are combined into a controlled voltage source and equivalent output impedance model, expressed as follows:
[0073]
[0074]
[0075]
[0076]
[0077]
[0078] in, and These are the combined source-side equivalent output impedance and the controlled voltage source disturbance, respectively. This is a disturbance from the equivalent voltage source of the power grid. For the power grid output impedance, The combined current disturbance injected into the common coupling point by the power grid and grid-type MMC. and These are the overall loop gain matrix and the return difference matrix, respectively. The return difference determinant of the overall interconnected system is given by and The generalized Nyquist criterion is used to determine the overall stability.
[0079] The above formula holds true only if the port coordinates, frequency sets, and matrix dimensions are compatible, the required inverse matrix exists, and the joint source side is internally stable and the interconnection is well-defined. Under these conditions, [the following is a continuation of the previous statement]. GNC is used to determine overall stability.
[0080] Specifically, the return difference matrix of each basic subsystem in the first layer, and the second layer... The third layer and the fourth floor Unified as ,in , When the value is 1, it corresponds to the basic subsystem. When the value is 34, it corresponds to the first coupled system. Taking 234 corresponds to the second coupled system. When 1234 is selected, the corresponding overall interconnected system is expressed as:
[0081] in, For the first The return difference determinant of the layer, For the first The return difference matrix of the layer.
[0082] Each layer adopts a common frequency set, compatible port coordinates, and matrix dimensions; when it is necessary to eliminate non-analyzed frequency components, the elimination is completed in the complete HSS model according to the actual port termination relationship, and the PWM frequency coupling is still retained in the corresponding converter closed-loop model.
[0083] In this embodiment, each layer takes the counterclockwise closed contour of the right half-plane of the complex frequency domain as the positive direction, and lets... , and The first The number of right-half-plane poles, the number of right-half-plane poles in the closed-loop implementation, and the number of signed windings of the return difference determinant around the origin satisfy the following:
[0084] Simultaneously, the internal dynamics, interconnection well-stability, and boundary pole conditions are checked, and the judgment results are recorded in the order of basic subsystem, first coupled system, second coupled system, and overall interconnection system. The sensitive subsystem or interconnection link corresponding to the decrease in stability margin is output.
[0085] Only when Furthermore, if there are no unprocessed boundary poles, the layer is determined to meet the stability condition; simultaneously, the output stability margin is reduced for the corresponding sensitive subsystem or interconnect. Local frequency axis scanning is used to display the results and does not replace the winding calculation of the complete profile.
[0086] To verify the effectiveness of this invention, simulation verification was performed. For example... Figure 3 As shown, under the frequency sweep condition with fixed disturbance amplitude and initial phase angle, the theoretical curves and frequency sweep results of the MMC positive and negative sequence port impedances are presented. Figure 3It can be seen that the frequency sweep results generally agree with the theoretical curves for both positive-sequence and negative-sequence port impedance amplitudes; similarly, the theoretical curves and frequency sweep results also show good consistency for both positive-sequence and negative-sequence port impedance phase angles. These results demonstrate that the present invention possesses good broadband port impedance description capabilities, providing reliable basic subsystem model support for positive-sequence and negative-sequence analysis, as well as subsequent layer-by-layer generalized Nyquist stability assessment. This helps in locating sensitive subsystems or interconnects corresponding to reduced stability margins.
[0087] In summary, the stability analysis method for MMC-based grid converter hybrid systems according to the above embodiments has the following beneficial effects: First, addressing the problem that existing technologies using a single average gain or equivalent control delay cannot explicitly describe the PWM multi-frequency disturbance mapping, this invention establishes a PWM modulation multi-frequency model based on the actual MMC modulation method and embeds the PWM frequency coupling into the closed-loop harmonic state-space model of two types of MMCs. Therefore, it can explicitly preserve the multi-frequency coupling relationship between modulation reference disturbance, switching timing disturbance, switching function disturbance, and bridge arm voltage disturbance. Second, addressing the problem that the specific circuit equivalents and return ratios of existing systems cannot be directly used for grid-type MMCs, follow-up grid-type MMCs, and grid and load connected in parallel at the same common coupling point, this invention establishes controlled voltage source and output impedance models for grid-type MMCs, controlled current source and input admittance combination models for follow-up grid-type MMCs, load input admittance models, and grid input models, respectively. Therefore, it preserves the source-type port relationships and port boundaries when the four types of branches are connected in parallel at the common coupling point. Furthermore, addressing the issue of locating sensitive components in layer-by-layer analysis, steps S2 to S5 sequentially combine components layer by layer in the following order: the first coupled system consisting of the load, the grid-connected MMC and the load; the second coupled system consisting of the grid-connected MMC and the load; and the overall hybrid system consisting of the power grid, the grid-connected MMC, and the load. Each layer establishes a corresponding loop gain matrix based on the stability conditions of the previous layer and employs the generalized Nyquist criterion. Therefore, it can distinguish the newly added interconnection effects of the basic subsystem, the first coupled system, the second coupled system, and the overall interconnected system, and output the corresponding sensitive subsystem or interconnection link with reduced stability margin. Finally, this invention combines the power grid and the grid-connected MMC into a controlled voltage source and equivalent output impedance, and uses the number of poles in the right half-plane achieved by compatible open-loop control to determine overall stability. Therefore, it can complete the overall small-signal stability judgment without changing the physical relationship between the grid-connected voltage source and the grid-connected current source.
[0088] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.
Claims
1. A stability analysis method for a hybrid MMC-grid converter system, characterized in that, The hybrid system includes a power grid connected in parallel to an AC common coupling point, a grid-type MMC, a grid-connected MMC, and loads. The method is as follows: S1. Establish basic subsystem models of the power grid, grid-type MMC, grid-connected MMC, and load in the hybrid system: Based on the internal circuits, control relationships, and actual PWM modulation methods of each multilevel converter, form a closed-loop small-signal model including PWM frequency coupling. Under the common port reference and frequency coordinates, obtain the controlled voltage source and output impedance model of the grid-type MMC, the controlled current source and input admittance combination model of the grid-connected MMC, and establish the input models of the power grid and load. S2, determine the stability of the power grid, grid-type MMC, grid-connected MMC and load basic subsystem in the hybrid system respectively: take the load as the starting point of layer-by-layer combination, take the power grid, grid-type MMC, grid-connected MMC and load as basic subsystems respectively, form the corresponding return difference matrix according to the closed-loop port model of each basic subsystem, use the generalized Nyquist criterion to judge the stability of the basic subsystem one by one, and output the basic subsystem that does not meet the stability condition; S3, Determine the stability of the coupled subsystem consisting of the grid-type MMC and the load: The grid-type MMC and the load impedance form the first coupled system. According to the voltage relationship at the common coupling point and the current balance relationship under the unified port direction, establish the loop gain matrix obtained by multiplying the input admittance of the grid-type MMC and the input impedance of the load. Use the generalized Nyquist criterion to determine the stability of the first coupled system and output the corresponding sensitive subsystem or interconnection link with reduced stability margin. S4. Determine the stability of the hybrid subsystem consisting of the network-type MMC, the interconnected network-type MMC, and the load: Add the network-type MMC to the first coupling system to form the second coupling system. Retain the controlled voltage source and output impedance relationship of the network-type MMC. Obtain the loop gain matrix by multiplying the input admittance of the first coupling system with the output impedance of the network-type MMC. Use the generalized Nyquist criterion to determine the stability of the second coupling system and output the stability margin to reduce the corresponding sensitive subsystem or interconnection. S5, determine the stability of the overall interconnected system consisting of the power grid, the grid-type MMC, the grid-connected MMC, and the load: add the power grid to the second coupling system to form the overall interconnected system, combine the power grid and the grid-type MMC into a controlled voltage source and an equivalent output impedance form, obtain the loop gain matrix by multiplying the input admittance of the first coupling system with the equivalent output impedance, use the generalized Nyquist criterion and combine it with the number of poles in the right half plane realized by compatible open loop to determine the overall stability, and output the sensitive subsystem or interconnection link corresponding to the decrease in stability margin.
2. The stability analysis method for MMC-based grid converter hybrid systems according to claim 1, characterized in that, In step S1, common-mode and differential-mode variables are used to describe the arm current, submodule capacitor voltage, and modulation amount of the MMC. Small-signal linearization is performed near the operating trajectory, and the coupling between the retained frequency components is described using the harmonic state-space form. The harmonic state-space model of the same MMC power circuit is applicable to both grid-type and follow-grid-type MMCs. The difference between the two is formed by the control model and the source port relationship. The PWM frequency coupling is embedded in the closed-loop harmonic state-space model of the strain gauge current transformer and is not treated as an independent analysis subsystem.
3. The stability analysis method for MMC-based grid converter hybrid systems according to claim 2, characterized in that, The grid-type MMC employs active-frequency and reactive-voltage droop control, transforming active and reactive power disturbances into controlled voltage sources through power measurement, droop control, phase angle dynamics, and port voltage mapping, while satisfying the following: in, For controlled voltage source disturbances in network-type MMC. For the dynamic mapping matrix of the controlled voltage source, For complex frequency variables, and These are active power disturbances and reactive power disturbances, respectively. Indicates transpose. and These are the port voltage disturbance and port current disturbance of the network-type MMC, respectively. It is a closed-loop output impedance matrix that includes the internal dynamics of the network-type MMC, droop control, phase angle dynamics, circulating current suppression control, actual virtual impedance control and / or current inner loop, and PWM frequency coupling.
4. The stability analysis method for MMC-based grid converter hybrid systems according to claim 3, characterized in that, The grid-type MMC employs a phase-locked loop, an outer DC voltage loop, an inner AC current loop, and circulating current suppression control to form controlled current sources from DC voltage disturbances, active power disturbances, and reactive power disturbances, while satisfying the following: in, To accommodate the controlled current source disturbance of the grid-type MMC. The dynamic mapping matrix of the controlled current source. This is a DC voltage disturbance. and These are the port current disturbance and port voltage disturbance of the mesh-type MMC, respectively. It is a closed-loop input admittance matrix that includes a phase-locked loop, an outer loop of DC voltage, an inner loop of AC current, circulating current suppression control, internal dynamics of the MMC, and PWM frequency coupling.
5. The stability analysis method for a hybrid MMC-grid converter system according to claim 4, characterized in that, In step S1, a PWM modulation multi-frequency model is established based on the actual MMC modulation method. For the modulation process that uses continuous modulation reference and carrier comparison and has non-degenerate steady-state intersection, the switching time disturbance is determined based on the slope difference between the modulation reference and carrier at the steady-state intersection, the switching function disturbance is determined based on the switching jump variable, and the multi-frequency mapping matrix is formed by the switching function disturbance, so that the PWM modulation dynamics are embedded into the converter closed-loop harmonic state space model in the order of modulation reference disturbance, actual modulation response disturbance and bridge arm voltage disturbance.
6. The stability analysis method for a hybrid MMC-grid converter system according to claim 5, characterized in that, For the MMC branch in the grid-connected configuration, the current is defined according to the direction injected from the branch to the common coupling point; for the load branch, the current is defined according to the port absorption direction of the input admittance model; the port voltages of the grid-connected MMC and the load are unified as the common coupling point voltage disturbance. Let the load input impedance matrix And establish the following formula: in, To establish the loop gain matrix between the network-type MMC and the load, To return the difference matrix, For the corresponding dimension, To return the difference determinant, The load input admittance matrix, The determinant is represented by ; after the stability test of the first coupled system, its external input admittance is denoted as . ,in Let be the external input admittance matrix of the first coupled system. This is the closed-loop input admittance matrix of the mesh-type MMC.
7. The stability analysis method for a hybrid MMC-grid converter system according to claim 6, characterized in that, In step S4, the voltage at the network-type MMC port and the voltage at the common coupling point are unified as follows: And establish based on current balance: in, The feature matrix of the hybrid source of the network-type MMC and the first coupled system. This is the gain matrix of the third-layer loop. To return the difference matrix, The return difference determinant of the second coupled system; by and The stability of the second coupled system is determined using the generalized Nyquist criterion.
8. The stability analysis method for a hybrid MMC-grid converter system according to claim 7, characterized in that, In step S5, the power grid and the grid-type MMC are combined into a controlled voltage source and an equivalent output impedance model, expressed as: in, and These are the combined source-side equivalent output impedance and the controlled voltage source disturbance, respectively. This is a disturbance from the equivalent voltage source of the power grid. For the power grid output impedance, The combined current disturbance injected into the common coupling point by the power grid and grid-type MMC. and These are the overall loop gain matrix and the return difference matrix, respectively. The return difference determinant of the overall interconnected system is given by and The generalized Nyquist criterion is used to determine the overall stability.
9. The stability analysis method for a hybrid MMC-grid converter system according to claim 8, characterized in that, Each layer takes the counterclockwise closed contour of the right half-plane of the complex frequency domain as the positive direction, and lets , and The first The number of right-half-plane poles, the number of closed-loop right-half-plane poles, and the number of signed windings of the return difference determinant around the origin in a layer-compatible open-loop implementation are given. , When the value is 1, it corresponds to the basic subsystem. When the value is 34, it corresponds to the first coupled system. Taking 234 corresponds to the second coupled system. When 1234 is selected, the overall interconnected system satisfies: Simultaneously, the internal dynamics, interconnection well-stability, and boundary pole conditions are checked, and the judgment results are recorded in the order of basic subsystem, first coupled system, second coupled system, and overall interconnection system. The sensitive subsystem or interconnection link corresponding to the decrease in stability margin is output.
