A phase space euclidean distance driven VSG double threshold fault ride-through control method

Abstract

This invention discloses a VSG dual-threshold fault ride-through control method driven by phase space Euclidean distance. It uses Hessian quadratic approximation of the analytical stable manifold and trajectory arc length topology algorithm to solve the instability critical intersection point. Using phase space Euclidean distance as a unified driving index, it builds a two-stage collaborative control architecture of first-stage inertial damping adaptive and second-stage saturated phase angle current limiting, and is equipped with smooth exit hysteresis logic to achieve adaptive fault ride-through under all operating conditions from light disturbances to extreme depth short circuits, thereby improving the transient stability of new energy grid-connected systems.

Description

Technical Field

[0001] This invention relates to a phase-space Euclidean distance driven VSG dual-threshold fault ride-through control method, which can be widely applied to the large disturbance fault stability control of grid-connected new energy systems such as wind power, photovoltaic, and energy storage, and belongs to the transient stability control technology of new energy power electronic converters. Background Technology

[0002] In recent years, the grid connection penetration rate of wind power and photovoltaic new energy sources has continued to increase. Traditional grid-connected inverters lack physical inertia and damping, significantly reducing the grid's ability to withstand disturbances. Virtual synchronous generators (VSGs), by simulating the rotor motion equations of synchronous machines, endow converters with active frequency and voltage regulation capabilities, making them a core grid-building equipment for new power systems. However, when a deep three-phase short-circuit fault occurs in the grid, VSGs are highly susceptible to transient power angle instability, becoming a core bottleneck for large-scale applications. Existing technologies are divided into two categories: stability analysis methods and fault ride-through control strategies, both of which have significant drawbacks. Existing transient stability analysis methods typically include: The equal area rule is only applicable to simple single-machine models and cannot be adapted to VSG multi-ring coupling and time-varying nonlinear systems with parameters, resulting in low analysis accuracy. Lyapunov energy function method: It is difficult to construct a strict global energy function for high-order power electronic systems, and the stability domain boundary estimation is too conservative, which cannot support real-time online control; Traditional phase plane analysis can only simulate fixed offline operating conditions, cannot dynamically track the fault evolution trajectory, and lacks unified, real-time quantifiable instability indicators; Current research on manifold theory only provides offline qualitative mapping of stability boundaries, lacks analytical solution schemes, cannot capture fault critical boundary points online, and the theory cannot be applied to closed-loop control.

[0003] Meanwhile, existing VSG fault-crossing control typically includes: Single inertia damping adaptive control: It only adjusts parameters and cannot reconstruct the stable manifold boundary of the system, and its stabilization effect is weak in the face of deep voltage drop faults; Fixed threshold current limiting control: The saturation current phase angle is fixed, and the threshold cannot be adaptively adjusted with the fault depth. Light faults cause excessive damping and power loss, while extreme faults have insufficient stability margin. The lack of hierarchical collaborative architecture means that a single control logic cannot cover all fault conditions, and hard mode switching is prone to power angle and secondary power oscillation impact. The control logic lacks stable boundary quantization input and relies solely on voltage amplitude to determine faults, making it unable to predict instability risks and thus constituting passive lag control.

[0004] Therefore, current VSG control lacks a complete three-in-one solution integrating online analytical modeling of global stability boundaries, real-time instability quantification indicators, and hierarchical adaptive collaborative control. This prevents the simultaneous implementation of inverter hardware overcurrent safety protection, fault transient stability domain widening, and rapid synchronous recovery after a fault, thus limiting its engineering applications. Therefore, there is an urgent need to provide a dual-threshold hierarchical fault ride-through control method based on phase space Euclidean distance driving. Summary of the Invention

[0005] This invention aims to solve the critical intersection point of instability by approximating the analytical stable manifold with Hessian quadratic approximation and using the trajectory arc length topology algorithm. It uses the Euclidean distance in phase space as a unified driving index to build a two-stage collaborative control architecture with first-stage inertial damping adaptive and second-stage saturated phase angle current limiting, and is equipped with smooth exit hysteresis logic to achieve adaptive fault ride-through under all working conditions from light disturbances to extreme depth short circuits, thereby improving the transient stability of new energy grid-connected systems.

[0006] To achieve the above objectives, the technical solution of this invention discloses a VSG dual-threshold fault ride-through control method driven by phase space Euclidean distance, comprising the following steps: Step 1: Real-time acquisition of the operating status of the power grid and converter, and calculation to obtain the current time. Real-time virtual power angle of virtual synchronous generator Real-time virtual angular velocity and real-time output electromagnetic power ; Step 2: Monitor the voltage drop status of the power grid in real time. If the drop exceeds the set threshold, determine that a short circuit fault has occurred in the power grid and proceed to the next step. Step 3: Based on the second-order rotor motion equation with active frequency, obtain the stable equilibrium point SEP and the unstable saddle point UEP, and calculate the virtual internal potential amplitude according to the reactive power-voltage regulation equation. Among them, the stable equilibrium point SEP is the steady-state operating point of the system during normal operation, and the unstable saddle point UEP is the critical saddle point of the stability domain boundary; Step 4: Employ a priority saturation current limiting strategy, utilize the current-limited power equation, and simultaneously solve the power balance constraints before and after the fault to obtain the optimal saturation phase angle analytically. ; Step 5: Calculate the system Jacobian matrix at the unstable saddle point UEP, solve for the positive and negative eigenvalues, and extract the normalized left eigenvector corresponding to the stable manifold. Combined with the second-order symmetric coefficient matrix of the stable manifold, an implicit quadratic approximation function for the stable manifold is obtained to fully characterize the boundary of the global stable domain in the all-phase plane. ,in, For the two-dimensional state variables of the system, , For virtual power angle, This refers to virtual angular velocity; Step 6: Using the initial operating state of the fault as the initial value for integration, perform forward numerical integration on the nonlinear ordinary differential equation of the system with a fixed step size to generate a set of discrete state trajectory points for the fault evolution. Retrieve the discrete state trajectory point set of fault evolution For the first time, the implicit quadratic approximation function of the stable manifold was derived. The trajectory points of the formed stable polygon are used to obtain the corresponding critical intersection points. , The virtual work angle corresponding to the trajectory point. The virtual angular velocity corresponding to the trajectory point; Step 7: Calculate the current running point Intersection with the critical point Real-time Euclidean distance between , ; Step 8: Based on real-time Euclidean distance Compared with the first-level warning threshold and Level II warning threshold The comparison results are used for dual-threshold hierarchical collaborative control and smooth recovery.

[0007] Preferably, in step 2, the stable equilibrium point (SEP) and the unstable saddle point (UEP) are obtained through the following steps: Step 201: Establish the second-order rotor motion equation with active frequency:

[0008] In the formula: For virtual power angle, for The first derivative; , These are the virtual angular velocity and the grid synchronous rated angular velocity, respectively. for The first derivative; Provided for active power; To output electromagnetic power to the system; This is virtual inertia; For virtual damping; Step 202: Let the second-order rotor motion equation with active frequency be... , In the range of work angles The inner solution yields: Stable Equilibrium Point (SEP): , The limiting electromagnetic power in the unrestricted flow state. The virtual power angle for the stable equilibrium point SEP; Unstable saddle point UEP: , The virtual power angle for the unstable saddle point UEP.

[0009] Preferably, in step 2, the virtual internal potential amplitude It is obtained by calculation using the following formula:

[0010] In the formula: This is the no-load voltage; , These are the reactive power and voltage regulation coefficients, respectively. , The reactive power is given by the reactive power and the reactive power output by the VSG. , These are the rated voltage and the actual terminal voltage, respectively.

[0011] Preferably, step 4 includes the following steps: To make the target equilibrium point converge to the current current-limited state ,Will Substituting into the current-limited power equation shown below, The initial active power of the system:

[0012] In the formula, This represents the limiting electromagnetic power under current-limited conditions. Differentiating the current-limited power equation, we have:

[0013] The above inequality constraints require That is, take Then we have: .

[0014] Preferably, in step 5, the normalized left feature vector is obtained using the following steps. : Taking the first-order partial derivative of the system dynamics equations at the unstable saddle point UEP yields the Jacobian matrix. : ; The positive eigenvalues ​​of the corresponding unstable manifold direction are obtained by solving the characteristic equation. and the negative eigenvalues ​​corresponding to the stable manifold direction Among them, positive eigenvalues Satisfy the following formula:

[0015] Suppose and The corresponding left feature vector is ,satisfy , , , The first element of the left eigenvector. This is the second element of the left eigenvector; For left eigenvectors Normalization is performed.

[0016] Preferably, in step 5, the second-order symmetric coefficient matrix of the stable manifold is obtained using the following steps: Introducing the second derivative Hessian matrix, the non-zero components of the system's Hessian matrix exist only in In the following formula:

[0017] The second-order symmetric coefficient matrix of the stable manifold is expressed as: Then it satisfies the following Sylvester-type matrix equation:

[0018] In the formula, It is a second-order identity matrix; Solving the above Sylvester-type matrix equations yields the second-order symmetric coefficient matrix of the stable manifold. The analytical solution for each element is shown in the following equation:

[0019]

[0020] .

[0021] Preferably, in step 5, the implicit quadratic approximation function of the stable manifold The second approximation obtained through Taylor expansion is:

[0022] In the formula, Let be the two-dimensional state vector corresponding to the unstable saddle point UEP.

[0023] Preferably, in step 6, the critical intersection point is obtained using the following steps. : The implicit quadratic approximation function of the stable manifold Discretized into counterclockwise ordered vertices, by vertices Orderly connected line segments form a closed safe polygon , ; For the generated set of discrete state trajectory points of the fault evolution Along the closed safety polygon Trajectory arc length search direction introduces topological inclusion indicator function For the set of discrete state trajectory points of the fault evolution The point-by-point spatial location is verified using the following algebraic topology: For the set of discrete state trajectory points of the fault evolution Any point within Calculation points Compared to the closed safety polygon sequentially connected a directed boundary line segment The sum of the directed angles subtended, the wandering point number of rings The algebraic analytical expression is defined as:

[0024] In the formula, Indicates As vertices, they point to the vertices of the two adjacent polygons respectively. , The directed angle formed by two vectors; According to the properties of topological homology groups, if and only if point Located in the closed security polygon When it is inside, the algebraic sum of the angles subtended by its sides is ,but If point Located in the closed security polygon If the algebraic sum of the angles subtended by its sides is 0, then... ; Obtain the first breaking of the closed security polygon The discrete time containing constraints, and the set of discrete state trajectory points corresponding to the fault evolution at that discrete time. The index of the midpoint is defined as Then we have:

[0025] In the formula, It is a set of positive integers; The set of discrete state trajectory points of the fault evolution With index Corresponding points Determined as the critical intersection point That is, the point The set of discrete state trajectory points for the fault evolution China's first to meet point.

[0026] Preferably, in step 8, the first-level early warning threshold and the secondary warning threshold It is dynamically updated using the following method: For the first-level warning threshold : ,and ,in: This is an infimum operation; This is the set of phase space distances from the operating point to the stability boundary when the system is subjected to the maximum permissible disturbance under normal operation. This is the minimum safe distance under normal disturbance conditions; A safety dead zone margin is reserved for hardware sampling noise and numerical calculation errors; The first-level warning threshold With the aforementioned secondary early warning threshold The phase space buffer depth between them; For the aforementioned secondary warning threshold :

[0027] In the formula: The transient power angle at which the system triggers the hardware current limiting critical state is calculated by the following formula:

[0028] in, For hardware security thresholds, For equivalent connection reactance, This is the voltage drop. The set of discrete state trajectory points for the fault evolution Zhongyu The corresponding angular velocity deviation.

[0029] Preferably, in step 8, the dual-threshold hierarchical collaborative control includes the following steps: like Then, first-level adaptive control is performed based on the following formula: Dynamically and synchronously increase virtual inertia Virtual damping The real-time Euclidean distance The smaller the virtual inertia Virtual damping The greater the increase, the more continuously the rotor's acceleration kinetic energy is consumed, suppressing the divergence of the power angle; like Then, secondary collaborative control will be implemented: Maintain the virtual inertia The virtual damping While implementing first-level adaptive control, the optimal saturation phase angle is simultaneously applied. Current limiting control actively shifts the unstable saddle point UEP, reconstructs the fault stability domain, and limits the output current to no more than the hardware threshold. like Then, normal control measures will be implemented: Run basic virtual synchronous generator control without any transient optimization; The smooth recovery includes the following steps: If the virtual power angle is within the frequency modulation dead zone of the stable equilibrium point (SEP), and the real-time Euclidean distance... satisfy The virtual inertia then decays linearly. The virtual damping The saturation current limiting is gradually removed until the initial value is reached, without sudden switching, thus eliminating the impact of secondary power and power angle oscillations.

[0030] Compared with existing technical solutions, the present invention has the following beneficial effects: 1) The Jacobi and Hessian matrix quadratic approximation method is used to analyze the globally stable manifold, breaking through the limitations of traditional offline qualitative analysis; the trajectory arc length and topological loop number algorithms can solve the instability critical intersection point online, and the Euclidean distance in phase space is used as a unified quantitative index to realize real-time perception of instability risk throughout the fault process, providing accurate input for closed-loop control; 2) The first-stage inertia damping adaptively handles moderate voltage drops, consuming only transient kinetic energy and not triggering hardware current limiting; the second-stage saturation phase angle current limiting works in concert to handle extreme deep short circuits, simultaneously widening the stability domain and protecting power devices; it solves the pain points of traditional single control, such as "over-control in minor faults and instability in severe faults"; simulation verification shows that under the same extreme fault conditions, this invention can maintain system stability, while the traditional fixed current limiting strategy completely diverges and disconnects from the grid. 3) The first-level control dynamically matches transient kinetic energy, shortening the fault convergence time; the second-level control reconstructs the stable manifold boundary through the optimal saturation phase angle, significantly extending the critical fault clearing time (CCT); under the extreme fault condition of the simulated 10kVA VSG system, the peak value of the first swing of the power angle is reduced by 15.8%, the fault recovery time is shortened by 38%, and the inverter current is always limited to within 1.2 times the rated hardware threshold. 4) Design a composite hysteresis criterion of power angle steady-state constraint and space safety distance, and linearly and gradually cancel the control parameters to avoid secondary oscillations in power and voltage caused by hard switching, and meet the dynamic performance requirements of IEEE547 and the national standard for low voltage ride-through of new energy. 5) It can be adapted to various VSG converters for wind power, photovoltaic, and energy storage by only modifying basic parameters such as rated capacity, connection reactance, and hardware current limiting; the algorithm has a small amount of computation and can be directly embedded into DSP and FPGA digital controllers to run in real time without the need for offline large-scale time-domain simulation, and can be adapted to the hardware modification of existing new energy converters. Attached Figure Description

[0031] Figure 1 The flowchart shows the overall execution process of the multi-level transient collaborative control strategy. Figure 2 Phase plane diagram for critical intersection point prediction using the trajectory arc length method; Figure 3 For extreme depth voltage drop ( Phase space evolution trajectory diagram. Detailed Implementation

[0032] The present invention will be further illustrated below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Furthermore, it should be understood that after reading the teachings of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent forms also fall within the scope defined by the appended claims.

[0033] The following detailed description, in conjunction with the accompanying drawings and a preferred embodiment of a 10kVA three-phase VSG, provides a complete reproduction process of the phase-space Euclidean distance-driven VSG dual-threshold fault ride-through control method disclosed in this invention. Those skilled in the art can reproduce the method by modifying the system capacity, impedance, and current-limiting parameters. The scope of protection of this invention is not limited to this embodiment.

[0034] In this embodiment of the invention, the terms used are uniformly interpreted as follows: VSG—Virtual Synchronous Generator, SEP—Stable Equilibrium Point, UEP—Unstable Saddle Point, CSA—Current Saturation Algorithm —The Euclidean distance from the phase space running point to the critical intersection of instability. —virtual inertia —Virtual damping, —The critical intersection of the fault trajectory and the stable manifold.

[0035] Based on the above terminology definitions, such as Figure 1 As shown, the method disclosed in this embodiment of the invention specifically includes the following steps: Step S1, System Operation and Real-time Sampling: The system collects the three-phase voltage at the grid connection point and the three-phase current output by the inverter in real time. It then obtains the grid synchronization angular velocity through a phase-locked loop. After calculation by the power calculation module and the virtual rotor motion mechanism, it obtains the real-time virtual power angle, virtual angular velocity, and output electromagnetic power of the VSG. In a preferred embodiment of this invention, the sampling period is matched with the control period to ensure millisecond-level refresh of the state variables.

[0036] Step S2, Fault Detection and Setting: Based on the real-time sampled voltage amplitude at the grid connection point, a short-circuit fault is determined to have occurred in the power grid when the voltage drop exceeds a set threshold. At this time, the time of the fault occurrence and the per-unit value of the voltage drop are recorded, and the electromagnetic power in the system dynamic equation is switched to the fault condition expression simultaneously to configure the fault boundary conditions for subsequent trajectory prediction.

[0037] Step S3: Transient dynamic modeling and equilibrium point solution of the current-limiting constraint VSG.

[0038] In a preferred embodiment of the present invention, step S3 may specifically include the following sub-steps: Step S3-1: Establish the basic dynamic equations of the VSG: Ignoring the high-frequency transients of the IGBT switch, the second-order rotor motion equation at the active frequency is expressed as follows (1): (1) In the formula: For VSG virtual power angle, for The first derivative; , These are the virtual angular velocity and the grid synchronous rated angular velocity, respectively. for The first derivative; Provided for active power; To output electromagnetic power to the system; This is virtual inertia; For virtual damping.

[0039] The reactive power-voltage regulation equation is established as shown in equation (2) below: (2) In the formula: The virtual internal potential amplitude, virtual internal potential amplitude The inverter output current amplitude and electromagnetic power limit are directly determined and are the core input parameters for current limiting constraint calculation; This is the no-load voltage; , These are the reactive power and voltage regulation coefficients, respectively. , The reactive power is given by the reactive power and the reactive power output by the VSG. , These are the rated voltage and the actual terminal voltage, respectively.

[0040] Step S3-2, Solve for SEP and UEP: In formula (1) , The system reaches a steady-state equilibrium point, and substituting the values, we get... , The limiting electromagnetic power in the unrestricted current state, within the power angle range. The inner solution yields: Stable Equilibrium Point (SEP): The steady-state operating point of the system during normal operation; Unstable saddle point UEP: The critical saddle point at the boundary of the stability region.

[0041] Step S3-3, Derivation of Saturation Current Limitation and Optimal Phase Angle: To ensure that the target equilibrium point under the current flow-limited state converges to the initial position , need to Substituting into the current-limited power equation, The initial active power of the system satisfies: (3) In equation (3), This represents the limiting electromagnetic power under current-limited conditions. This is the saturation phase angle.

[0042] According to the synchronization stability criterion in nonlinear dynamics, a necessary condition for a stable equilibrium point is that its synchronization power coefficient must be greater than zero, i.e. Differentiating the current-limited power equation reveals that: (4) This inequality constraint requires That is, take .

[0043] In summary, the analytical expression for the optimal saturation phase angle can be obtained as shown in equation (5) below.

[0044] (5) Step S3-4, Solving the quadratic analytical approximation of the stable manifold, further includes the following steps: 1) Obtain the first-order partial derivative of the system dynamics equations at the unstable saddle point UEP to obtain the Jacobian matrix. : By solving the characteristic equation, two real eigenvalues ​​are obtained: positive eigenvalues. ( Corresponding to the direction of the unstable manifold, negative eigenvalues ( This corresponds to the stable manifold direction. Because... This matrix must have a positive real eigenvalue. Corresponding to an unstable manifold and a negative eigenvalue Corresponding to a stable manifold. Unstable positive real eigenvalues. satisfy: (6) Suppose and The corresponding left feature vector is ,satisfy .make , can be obtained ,right Normalization processing Left eigenvector This determines the linear tangent hyperplane of the stable manifold at the saddle point.

[0045] 2) Introduce the second-order derivative Hessian matrix. The non-zero components of the system's Hessian matrix exist only in the second equation: (7) Let the two-dimensional state variables of the system be... According to nonlinear dynamics theory, the implicit scalar function of a stable manifold in the phase plane The result can be approximated by Taylor expansion as follows: (8) In the formula, The two-dimensional state vector and second-order coefficient matrix corresponding to the system saddle point UEP. It is a symmetric matrix that satisfies the following Sylvesterian matrix equations: (9) In equation (9), It is a second-order identity matrix.

[0046] Solving this system of equations and substituting the results, we can obtain the matrix. Analytical solutions for each element: (10) (11) (12) Step S4: Prediction of critical intersection points and Euclidean distance Real-time calculation.

[0047] In a preferred embodiment of the present invention, step S4 may specifically include the following sub-steps: Step S4-1: During the fault, based on the voltage drop... The nonlinear ordinary differential equation of the system can be expressed as: (13) In equation (13), Two-dimensional state variables The first derivative. Equations (7) to (13) together constitute the implicit equations for the analytic quadratic approximation of the stable manifold.

[0048] by As the initial value, a fixed step size is used. By performing forward numerical integration on equation (13), the discretized predicted trajectory point set during the fault period can be obtained. As shown in equation (14).

[0049] (14) In the formula, This represents a continuous time variable in the fault evolution process.

[0050] Step S4-2: The stable manifold obtained in step S3-4 Discretize the trajectory into a closed polygon with counterclockwise ordered vertices, and calculate the number of points encircling the trajectory. Number of rings An equal value of 1 represents a point. Within the stable region, the number of orbitals 0 represents a point Outside the instability region, the specific steps include: 1) Construct a closed safety polygon on a two-dimensional phase plane The closed safety polygon Depend on vertices Composed of ordered connected line segments: (15) In the formula, This represents the stability region of the system.

[0051] 2) For the generated discretized predicted trajectory point set A topological inclusion indicator function is introduced along the trajectory arc length search direction. A rigorous algebraic topological check is performed on the point-by-point spatial position of the trajectory: For any wandering point on the predicted trajectory Calculate the point relative to the polygon. sequentially connected a directed boundary line segment The sum of the directed angles subtended, its number of turns The algebraic analytical expression is defined as: (16) In equation (16), Indicates Let be a vertex, and let be the directed angle formed by two adjacent vertices of the polygon. According to the properties of topological homology groups, is true if and only if [the vertex is not specified]. When located inside a stable polygon, the algebraic sum of the angles subtended by its sides is: (Right now If it is located on the outside, the algebraic sum of the angles subtended by its sides is 0 (i.e., ).

[0052] Step S4-3: The critical condition for the system to experience transient instability and escape is the discrete moment when the forward-stepped trajectory first breaks the polygon containment constraint. Let the trigger index for this sudden change be... Then it can be expressed rigorously as: (17) In equation (17), It is a set of positive integers.

[0053] Step S4-4, Critical Intersection Point of the System It is uniquely identified as: Retrieve the first matching discrete index Interpolation yields the critical intersection point. .

[0054] Step S4-5, Current running status point The real-time sampling module acquires the three-phase voltage and current at the inverter terminals. Through a phase-locked loop and a power calculation module, the real-time virtual power angle and virtual angular velocity are obtained, forming the current operating state point. .

[0055] Step S4-6, Real-time Euclidean Distance calculate: (18) Step S5: Dual threshold tuning, hierarchical control, and smooth exit logic.

[0056] In a preferred embodiment of the present invention, step S5 may specifically include the following sub-steps: Step S5-1, Two-level threshold calculation: 1) Calculation of Level 1 Early Warning Threshold: ,and ,in: Operations for the infimum (minimum); This is the set of phase space distances from the operating point to the stability boundary when the system is subjected to the maximum permissible disturbance under normal operation. That is, the minimum safe distance under normal disturbances, which is obtained through offline simulation calibration; The safety dead zone margin reserved for hardware sampling noise and numerical calculation errors is typically taken as 5% to 10% of the minimum normal distance in engineering practice. To provide a buffer depth in the phase space between the two threshold levels, ensuring that the first-level control has a sufficient path to consume transient kinetic energy, in engineering, 30% to 50% of the second-level threshold is used, or it can be calibrated through simulation.

[0057] 2) Calculation of Level 2 Emergency Threshold: Pick During the fault, the grid connection point voltage drops to... Ignoring line resistance, the amplitude of the VSG transient output current Its internal virtual power angle There is a direct coupling relationship: (19) In the formula, The amplitude of the VSG virtual internal potential. This is the equivalent connection reactance.

[0058] As the fault trajectory evolves with the acceleration power, it continues until the output current reaches the hardware safety threshold. At this time, the inverter's CSA will be forcibly activated. (This is followed by a seemingly unrelated instruction: "Let...") It can inversely solve for the transient power angle when the system triggers the hardware current limiting critical state. : (20) The calculated critical power angle Mapping to Discretized Predicted Trajectory Point Set In this process, numerical interpolation can be used to extract the trajectory evolution to the critical work angle. The corresponding angular velocity deviation coordinates This is the geometric mapping of the hardware overcurrent limit on the transient dynamic trajectory.

[0059] Calculate the current limiting point to Distance, that is : (twenty one) Step S5-2, if If so, then secondary collaborative control will be implemented; if If so, then first-level adaptive control will be implemented; if If so, then routine control measures will be implemented.

[0060] 1) First-order adaptive control is based on the following formula: (twenty two) (twenty three) In the formula, , These are the initial virtual inertia and the initial virtual damping, respectively. , For adaptive gain, the value is greater than 0; The smaller, , Increase synchronously.

[0061] 2) Second-level coordinated control: Maintain the above , The first-level adaptive control simultaneously switches the electromagnetic power equation to the current-limiting optimal phase angle form, constraining the safety threshold. During the current-limiting period, the transient power equation is reconstructed as follows: (twenty four) Step S5-3: Update system status in real time: The input for this step is: the virtual inertia output by the hierarchical control. Virtual damping saturation phase angle Parameters, as well as real-time voltage and current data sampled from the power grid; The specific update process used in this step is as follows: Substitute the control parameters into the VSG dynamic equations, and refresh the current system's power angle by combining the real-time sampling data. angular velocity Output electromagnetic power ; The output of this step is: the updated real-time state vector. This serves as the input for the next round of Euclidean distance calculation and hysteresis exit determination.

[0062] Step S5-4, Composite Hysteresis Exit Criterion : (25) In the formula, Real-time state vector Intersection with critical point Real-time Euclidean distance between them This is the frequency modulation dead zone power angle threshold.

[0063] When both conditions shown in equation (25) are met simultaneously, the linear reduction occurs. , Once the initial value is reached, the saturation current limiting control is gradually withdrawn, and the system smoothly switches back to normal VSG control.

[0064] The embodiments of the present invention simulate the above scheme. The simulation platform used is MATLAB / Simulink; 10kVA three-phase VSG grid-connected system. The core simulation parameters are shown in the table below.

[0065]

[0066] Table 1 Simulation Parameters The fault conditions are set as follows: a three-phase short-circuit fault occurs in the power grid, the fault duration is 0.1s, and the per-unit values ​​of the voltage drop at the generator terminals are 0.21pu (moderate fault), 0.20pu (deep fault), and 0.18pu (extreme depth fault).

[0067] The obtained experimental simulation results are as follows: (1) Algorithm verification conditions: Offline and online verification is carried out based on the trajectory arc length integral combined with the topological wrap number determination logic. The system performs verification without applying control. The calculated critical intersection points and high-precision simulation results can pinpoint the instability boundary points, providing a reliable quantitative basis for the subsequent dynamic calculation of the two-level thresholds, such as... Figure 2 .

[0068] (2) Moderate fault Only the first-level adaptive parameter control was activated, and the entire trajectory converged within the safe and stable domain. The peak value of the power angle oscillation was 1.06 rad, and the oscillation of the entire operating condition completely dissipated in 1.15 seconds.

[0069] (3) Deep Fault The two-level control mechanism works in tandem, expanding the system's stability boundary and ensuring that the converter's output current never touches the hardware current limit.

[0070] (4) Limit failure The original stability boundary of the fault is breached by a short-circuit disturbance. The system's stability is ensured by relying on the safety range after current limiting reconstruction to completely encompass the fault evolution trajectory. Figure 3 .

[0071] This invention is not limited to a single 10kVA VSG grid-connected architecture: In a multi-unit parallel VSG grid-connected system, a single converter can independently calculate the real-time instability distance and autonomously execute the two-level hierarchical control logic; for direct-drive wind turbine generators and grid-connected energy storage converters, only the two basic parameters of line impedance and installed rated capacity need to be replaced, and the closed-loop logic of the entire hierarchical control does not need to be readjusted, making it highly reusable across equipment categories.

[0072] Those skilled in the art will understand that the features described in the various embodiments and / or claims of this invention can be combined or combined in various ways, even if such combinations or combinations are not explicitly described in this disclosure. In particular, the features described in the various embodiments and / or claims of this invention can be combined or combined in various ways without departing from the spirit and teachings of this disclosure. All such combinations and / or combinations fall within the scope of this disclosure.

[0073] The embodiments of the present invention have been described above. However, these embodiments are merely illustrative and not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. The scope of the present invention is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of the present invention, and all such substitutions and modifications should fall within the scope of this disclosure.

Claims

1. A VSG dual-threshold fault ride-through control method driven by phase space Euclidean distance, characterized in that, Includes the following steps: Step 1: Real-time acquisition of the operating status of the power grid and converter, and calculation to obtain the current time. Real-time virtual power angle of virtual synchronous generator Real-time virtual angular velocity and real-time output electromagnetic power ; Step 2: Monitor the voltage drop status of the power grid in real time. If the drop exceeds the set threshold, determine that a short circuit fault has occurred in the power grid and proceed to the next step. Step 3: Based on the second-order rotor motion equation with active frequency, obtain the stable equilibrium point SEP and the unstable saddle point UEP, and calculate the virtual internal potential amplitude according to the reactive power-voltage regulation equation. Among them, the stable equilibrium point SEP is the steady-state operating point of the system during normal operation, and the unstable saddle point UEP is the critical saddle point of the stability domain boundary; Step 4: Employ a priority saturation current limiting strategy, utilize the current-limited power equation, and simultaneously solve the power balance constraints before and after the fault to obtain the optimal saturation phase angle analytically. ; Step 5: Calculate the system Jacobian matrix at the unstable saddle point UEP, solve for the positive and negative eigenvalues, and extract the normalized left eigenvector corresponding to the stable manifold. Combined with the second-order symmetric coefficient matrix of the stable manifold, an implicit quadratic approximation function for the stable manifold is obtained to fully characterize the boundary of the global stable domain in the all-phase plane. ,in, For the two-dimensional state variables of the system, , For virtual power angle, This refers to virtual angular velocity; Step 6: Using the initial operating state of the fault as the initial value for integration, perform forward numerical integration on the nonlinear ordinary differential equation of the system with a fixed step size to generate a set of discrete state trajectory points for the fault evolution. Retrieve the discrete state trajectory point set of fault evolution For the first time, the implicit quadratic approximation function of the stable manifold was derived. The trajectory points of the formed stable polygon are used to obtain the corresponding critical intersection points. , The virtual work angle corresponding to the trajectory point. The virtual angular velocity corresponding to the trajectory point; Step 7: Calculate the current running point Intersection with the critical point Real-time Euclidean distance between ; Step 8: Based on real-time Euclidean distance Compared with the first-level warning threshold and Level II warning threshold The comparison results are used for dual-threshold hierarchical collaborative control and smooth recovery.

2. The VSG dual-threshold fault ride-through control method driven by phase space Euclidean distance as described in claim 1, characterized in that, In step 2, the stable equilibrium point (SEP) and the unstable saddle point (UEP) are obtained through the following steps: Step 201: Establish the second-order rotor motion equation with active frequency: In the formula: For virtual power angle, for The first derivative; , These are the virtual angular velocity and the grid synchronous rated angular velocity, respectively. for The first derivative; Provided for active power; To output electromagnetic power to the system; This is virtual inertia; For virtual damping; Step 202: Let the second-order rotor motion equation with active frequency be... , In the range of work angles The inner solution yields: Stable Equilibrium Point (SEP): , The limiting electromagnetic power in the unrestricted flow state. The virtual power angle for the stable equilibrium point SEP; Unstable saddle point UEP: , The virtual power angle for the unstable saddle point UEP.

3. The VSG dual-threshold fault ride-through control method driven by phase space Euclidean distance as described in claim 2, characterized in that, In step 2, the virtual internal potential amplitude It is obtained by calculation using the following formula: In the formula: This is the no-load voltage; , These are the reactive power and voltage regulation coefficients, respectively. , The reactive power is given by the reactive power and the reactive power output by the VSG. , These are the rated voltage and the actual terminal voltage, respectively.

4. The VSG dual-threshold fault ride-through control method driven by phase space Euclidean distance as described in claim 3, characterized in that, Step 4 includes the following steps: To make the target equilibrium point converge to the current current-limited state ,Will Substituting into the current-limited power equation shown below, The initial active power of the system: In the formula, This represents the limiting electromagnetic power under current-limited conditions. Differentiating the current-limited power equation, we have: The above inequality constraints require That is, take Then we have: .

5. The VSG dual-threshold fault ride-through control method driven by phase space Euclidean distance as described in claim 4, characterized in that, In step 5, the normalized left feature vector is obtained using the following steps. : Taking the first-order partial derivative of the system dynamics equations at the unstable saddle point UEP yields the Jacobian matrix. : ; The positive eigenvalues ​​of the corresponding unstable manifold direction are obtained by solving the characteristic equation. and the negative eigenvalues ​​corresponding to the stable manifold direction Among them, positive eigenvalues Satisfy the following formula: Suppose and The corresponding left feature vector is ,satisfy , , , The first component of the left eigenvector. This is the second component of the left eigenvector; For left eigenvectors Normalization is performed.

6. The VSG dual-threshold fault ride-through control method driven by phase space Euclidean distance as described in claim 5, characterized in that, In step 5, the second-order symmetric coefficient matrix of the stable manifold is obtained using the following steps: Introducing the second derivative Hessian matrix, the non-zero components of the system's Hessian matrix exist only in In the following formula: The second-order symmetric coefficient matrix of the stable manifold is expressed as: Then it satisfies the following Sylvester-type matrix equation: In the formula, It is a second-order identity matrix; Solving the above Sylvester-type matrix equations yields the second-order symmetric coefficient matrix of the stable manifold. The analytical solution for each element is shown in the following equation: 。 7. The VSG dual-threshold fault ride-through control method driven by phase space Euclidean distance as described in claim 6, characterized in that, In step 5, the stable manifold implicit quadratic approximation function The second approximation obtained through Taylor expansion is: In the formula, Let be the two-dimensional state vector corresponding to the unstable saddle point UEP.

8. The VSG dual-threshold fault ride-through control method driven by phase space Euclidean distance as described in claim 7, characterized in that, In step 6, the critical intersection point is obtained using the following steps. : The implicit quadratic approximation function of the stable manifold Discretized into counterclockwise ordered vertices, by vertices Orderly connected line segments form a closed safe polygon , ; For the generated set of discrete state trajectory points of the fault evolution Along the closed safety polygon Trajectory arc length search direction introduces topological inclusion indicator function For the set of discrete state trajectory points of the fault evolution The point-by-point spatial location is verified using the following algebraic topology: For the set of discrete state trajectory points of the fault evolution Any point within Calculation points Compared to the closed safety polygon sequentially connected a directed boundary line segment The sum of the directed angles subtended, the wandering point number of rings The algebraic analytical expression is defined as: In the formula, Indicates As vertices, they point to the vertices of the two adjacent polygons respectively. , The directed angle formed by two vectors; According to the properties of topological homology groups, if and only if point Located in the closed security polygon When it is inside, the algebraic sum of the angles subtended by its sides is ,but If point Located in the closed security polygon If the algebraic sum of the angles subtended by its sides is 0, then... ; Obtain the first breaking of the closed security polygon The discrete time containing constraints, and the set of discrete state trajectory points corresponding to the fault evolution at that discrete time. The index of the midpoint is defined as Then we have: In the formula, It is a set of positive integers; The set of discrete state trajectory points of the fault evolution With index Corresponding points Determined as the critical intersection point That is, the point The set of discrete state trajectory points for the fault evolution China's first to meet point.

9. The VSG dual-threshold fault ride-through control method driven by phase space Euclidean distance as described in claim 8, characterized in that, In step 8, the first-level early warning threshold and the secondary warning threshold It is dynamically updated using the following method: For the first-level warning threshold : ,and ,in: This is an infimum operation; This is the set of phase space distances from the operating point to the stability boundary when the system is subjected to the maximum permissible disturbance under normal operation. This is the minimum safe distance under normal disturbance conditions; A safety dead zone margin is reserved for hardware sampling noise and numerical calculation errors; The first-level warning threshold With the aforementioned secondary early warning threshold The phase space buffer depth between them; For the aforementioned secondary warning threshold : In the formula: The transient power angle at which the system triggers the hardware current limiting critical state is calculated by the following formula: in, For hardware security thresholds, For equivalent connection reactance, This is the voltage drop. The set of discrete state trajectory points for the fault evolution Zhongyu The corresponding angular velocity deviation.

10. The VSG dual-threshold fault ride-through control method driven by phase space Euclidean distance as described in claim 8, characterized in that, In step 8, the dual-threshold hierarchical collaborative control includes the following steps: like Then, first-level adaptive control is performed based on the following formula: Dynamically and synchronously increase virtual inertia Virtual damping The real-time Euclidean distance The smaller the virtual inertia Virtual damping The greater the increase, the more continuously the rotor's acceleration kinetic energy is consumed, suppressing the divergence of the power angle; like Then, secondary collaborative control will be implemented: Maintain the virtual inertia The virtual damping While implementing first-level adaptive control, the optimal saturation phase angle is simultaneously applied. Current limiting control actively shifts the unstable saddle point UEP, reconstructs the fault stability domain, and limits the output current to no more than the hardware threshold. like Then, normal control measures will be implemented: Run basic virtual synchronous generator control without any transient optimization; The smooth recovery includes the following steps: If the virtual power angle is within the frequency modulation dead zone of the stable equilibrium point (SEP), and the real-time Euclidean distance... satisfy The virtual inertia then decays linearly. The virtual damping The saturation current limiting is gradually removed until the initial value is reached, without sudden switching, thus eliminating the impact of secondary power and power angle oscillations.

Images