New energy grid-connected system dense mode separation method based on gabor disc parameter optimization

Abstract

The present application relates to the technical field of new energy grid-connected system stability control, and particularly relates to a new energy grid-connected system dense mode separation method based on Gerschgorin disc parameter optimization. The present application establishes a harmonic state space model of a new energy field and a grid-connected system under multiple time scales, can separate the new energy grid-connected system of different time scales, and reduces the parameter optimization iteration calculation amount when separating the new energy grid-connected system dense mode subsequently. The present application realizes the physical machine decoupling of the dynamic interaction of multiple time scales, guarantees the oscillation mode separation while taking into account the overall control robustness of the system in the wide frequency domain, and is beneficial to improving the working stability and reliability of the power generation unit damping controller under actual operation condition fluctuations.

Description

Technical Field

[0001] This invention relates to the field of stability control technology for new energy grid-connected systems, and in particular to a method for separating dense modes in new energy grid-connected systems based on Gale disk parameter optimization. Background Technology

[0002] In recent years, the global power industry has gradually transitioned to low-carbon power, with new energy power generation, represented by wind power and photovoltaics, experiencing rapid development. Large-scale integration of new energy sources has become a trend in power system development. New energy power plants often have a large number of identical generating units. Because these units operate under similar conditions and are connected to the same electrical system, their converters interact with the grid, leading to a concentrated distribution of oscillation mode damping and frequency, resulting in similar or identical oscillation modes. Dense-mode resonance in new energy grid-connected systems suffers from weak robustness of the participation factor; that is, small changes in physical parameters can cause significant changes in the participation factor, leading to the failure or even adverse effect of the designed damping controller.

[0003] Existing methods for dense pattern separation and analysis still have significant limitations when dealing with large-scale renewable energy grid-connected systems. Traditional numerical algorithms based on the eigenvalue decomposition of the entire system suffer from huge computational loads and the "curse of dimensionality" when dealing with high-dimensional sparse state matrices, and even produce a large number of "missing roots" phenomena, making it difficult to fully reproduce the oscillation of the system. Summary of the Invention

[0004] To overcome the problem of broadband oscillation eigenvalue clustering in large-scale renewable energy grid-connected systems in the prior art, this invention proposes a dense mode separation method for renewable energy grid-connected systems based on Gale disk parameter optimization. Without relying on accurate eigenvalue sensitivity calculation, it achieves efficient separation of dense modes under strong coupling of dense modes at multiple time scales, thereby effectively improving the robustness of the dense mode participation factor in renewable energy grid-connected systems.

[0005] This invention proposes a method for separating dense modes in new energy grid-connected systems based on Gale disk parameter optimization, comprising the following steps: Step 1: Obtain the state variable vector x and linearized state-space model of the new energy grid-connected system. A detail ∈R n ×n Where n is the number of new energy power generation units, and x is the set of state variables of all power generation units in the grid-connected system; according to different time scales... A detail Block rearrangement is performed; the time scale is divided into three levels: microseconds, milliseconds, and seconds. Block arrangement involves rearranging the matrix... A detail Reorganize according to the speed of the time scale; St2, A after block arrangement detail Calculate the density mode coupling degree index of each new energy power generation unit; the density mode coupling degree index of the system is A after block arrangement. detail The summation of the absolute values ​​of the off-diagonal elements in a single row of a matrix along the row-count dimension, which is the ratio of the absolute values ​​of the diagonal elements in the same row. If the dense pattern coupling degree index > k th +c0; indicates that the new energy grid-connected system will experience intense oscillations. If the dense pattern coupling degree index < k th -c0; indicates that no dense oscillation mode has occurred. c0 is the set floating difference value, k th This is the oscillation coupling threshold.

[0006] Preferred: ; Among them, A p The state matrix of the new energy power grid; E d F represents the interface matrix for the transformation from DC power generation units to new energy power generation units, characterized by an n1-dimensional column vector of interface coefficients for the transformation of DC grid state variables to the state variables of each new energy power generation unit, where n1 is the number of DC power generation units; d The interface matrix for converting new energy power generation units to DC power generation units is represented by an n1-dimensional row vector of interface coefficients for the state variables of new energy power generation units to be converted to state variables of DC grids. E a This is the interface matrix for converting AC power generation units to new energy power generation units, representing the coupling relationship between DC grid state variables and new energy power generation units; F a This is the interface matrix for converting new energy power generation units to AC power generation units, representing the coupling relationship between new energy power generation units and the AC power grid.

[0007] Preferably, A after block rearrangement detail for: ; A pF A pM A pL A respectively detail The rearranged block matrix at various time scales of microseconds (µs), milliseconds (ms), and seconds (s); as described in the dense mode separation method for new energy grid-connected systems based on Gale disk parameter optimization, characterized in that if the dense mode coupling degree index is located in the interval (k... th -c0, k thIf the second-order sensitivity of the specified parameters of each new energy power generation unit is calculated, and if the second-order sensitivity of the parameters is greater than the set threshold, it is determined that the new energy power generation unit has a strong mode coupling risk. Let the set of specified parameters of the power generation unit be referred to as parameter set p; the parameters correspond one-to-one with the operating conditions of the new energy power grid, and are a one-dimensional vector composed of multiple target parameters; the h-th parameter p in p. h The second-order sensitivity is denoted as ; For p h Participation factors; let the state variables of the power generation unit be related to p h The derivative of the parameter p of the power generation unit is denoted as p. h Sensitivity S h ; Parameter p h The second-order sensitivity is: ; Among them, S j p is the j-th parameter in p. j The sensitivity, u is the total number of parameters in p; ω h and ω j p h and p j Weighted contribution to oscillation modes.

[0008] Preferably, when there is a risk of strong mode coupling in the new energy power generation unit, controllable parameters in the parameter set p of the new energy power generation unit are selected for optimization. The selection method for controllable parameters is as follows: calculate each parameter p h The first-order sensitivities are sorted in descending order, and the parameters corresponding to the top y first-order sensitivities are selected as key parameters. From the key parameters, adjustable parameters are selected as controllable parameters; parameter p h The first-order sensitivity is obtained by taking the derivative of its participation factor with respect to the parameter set p in vector form; The optimization objective of the controllable parameters is: ; in, Indicates the controllable parameter θ g A after time-division detail The element in the i-th row and j-th column, To represent the controllable parameter θ g A after time-division detail The element in the i-th row and i-th column; Indicates the controllable parameter θ g A measure of the degree of dense pattern coupling in a time system.

[0009] Preferably, the controllable parameter θ of the new energy power generation unit g The optimization method is as follows: ; , ; Where, θ g = , representing the controllable parameters after the current iteration m times; and Let represent the controllable parameters after m+1 and m-1 iterations, respectively; the initial value of m is 0. The initial value is 0; and Let them represent first-order terms and second-order terms, respectively. This indicates the first-order derivative. This indicates the second derivative.

[0010] Preferred, ; ; in, After block rearrangement The element in the i-th row and j-th column, After block rearrangement The element in the i-th row and i-th column; k m The dense mode coupling degree index of the system after the m-th iteration is the set of controllable parameters of the system.

[0011] Preferred, p h Participating factors The result is a normalized product of the absolute value of sensitivity and the weight contribution in the parameter dimension.

[0012] Preferably, the oscillation coupling threshold k of the new energy power generation unit th Take the average value of the damping coefficient of the new energy power generation unit. and average inertia coefficient The ratio of the square root of .

[0013] This invention proposes a system for implementing a method for separating dense modes in new energy grid-connected systems based on Gale disk parameter optimization, comprising: The state-space model building module is used to establish harmonic state-space models of new energy fields and grid-connected systems at different time scales. It analyzes the inertial equation expressions of each component considering damping effects to obtain the Gael disk block matrix at different time scales, i.e., the state-space model after block arrangement. A detail ; The coupling index calculation module is used for the state-space model based on the block arrangement. Adetail The coupling degree index of dense mode for each new energy power generation unit is calculated, and the boundary conditions of the coupling degree index of dense mode are constructed based on the Gale disk to achieve dense mode separation. The parameter iteration optimization module iteratively optimizes system parameters with strong coupling risks, taking the minimization of the Gehr circle radius and the stability of the new energy grid-connected system as constraints for different coupling degrees.

[0014] The present invention proposes a storage medium storing a computer program, which, when executed, is used to implement the aforementioned method for separating dense modes in new energy grid-connected systems based on Gale disk parameter optimization.

[0015] The advantages of this invention are: (1) This invention establishes a harmonic state-space model of the new energy field and grid-connected system at multiple time scales, which can separate the dense modes of the new energy grid-connected system at different time scales, reducing the amount of parameter optimization iteration calculations for subsequent separation of dense modes of the new energy grid-connected system. This invention achieves physical decoupling of the dynamic interaction at multiple time scales, ensuring the separation of oscillation modes while taking into account the overall control robustness of the system in a wide frequency domain, which is conducive to improving the stability and reliability of the damping controller of the power generation unit under actual operating condition fluctuations.

[0016] (2) This invention proposes a method for selecting the coupling degree index of a Gell-Hill disk by performing first-order and second-order sensitivity analysis on each participating factor based on the harmonic state space block matrix. Combining the minimization of the Gell-Hill circle radius and system stability constraints, the parameters of the dense mode system are iteratively optimized. This invention uses second-order sensitivity to assess the risk of strong coupling, which is beneficial to eliminating the computational distortion risk caused by the extremely strong nonlinear coupling characteristics exhibited by the first-order eigenvalues ​​in the "dense mode" region where eigenvalues ​​are highly clustered. The application of second-order sensitivity is more conducive to accurately assessing the modal drift risk caused by parameter perturbation.

[0017] (3) The present invention also proposes a dense mode separation parameter optimization method; the method effectively reduces the clustering coupling of dense mode feature values ​​while ensuring the stability margin of oscillation region and non-oscillation region, and significantly improves the dynamic stability performance of new energy grid-connected system.

[0018] (4) The present invention first establishes a linearized state-space model that takes into account multi-machine coupling. A detailThis invention introduces generalized damped-inertial analysis theory to quantify the dynamic response time constant of state variables, using this as a physical criterion to construct a multi-timescale block state matrix. Secondly, based on Gell's disk theorem, a dense mode coupling degree index is derived, and by analyzing the second-order sensitivity of control parameters, the key dominant parameters of the strongly coupled region are accurately identified. Furthermore, a nonlinear optimization model is constructed with the objectives of minimizing the overlap area of ​​the Gell's circle and maximizing the system stability margin, achieving effective separation of dense modes. Finally, Monte Carlo probabilistic sensitivity analysis is used to evaluate the robustness after separation. This invention significantly improves the multi-mode decoupling capability and operational stability of grid-connected systems while avoiding the difficulty of solving high-sensitivity eigenvalues. Attached Figure Description

[0019] Figure 1 This is a structural diagram of a new energy grid-connected system; Figure 2 Flowchart of a dense mode separation method for new energy flexible DC grid-connected systems based on Gale disk parameter optimization; Figure 3 Flowchart of the method for selecting controllable parameters. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0021] like Figure 1 As shown, the new energy grid-connected system includes a new energy grid composed of new energy power generation units (such as photovoltaic units and wind turbine units), a DC grid composed of DC power generation units, and an AC grid composed of AC power generation units.

[0022] In a renewable energy power grid, the number of renewable energy power generation units is n, and the total number of state variables in the renewable energy power grid is a. n, a represents the number of state variables in a single new energy power generation unit; In a DC power grid, the number of DC power generation units is n1, and the total number of state variables of the DC power grid is a1. n1 and a1 are the number of state variables in a single DC power generation unit; In an AC power grid, the number of AC power generation units is n2, and the total number of state variables of the AC power grid is a2. n2 and a2 are the number of state variables in a single AC power generation unit; n>n1, and n>n2.

[0023] This embodiment proposes a dense mode separation method for new energy grid-connected systems based on Gell disk parameter optimization (hereinafter referred to as the dense mode separation method). When it is determined that the new energy grid-connected system (hereinafter referred to as the system) will experience dense mode oscillations, the linearized state-space model of the new energy grid-connected system is used. A detail The sum of the absolute values ​​of the off-diagonal elements in each row of the matrix (i.e., the radius of the Gaelic disk) is used as the optimization objective to update the key parameters of the system and achieve separation of dense oscillation modes.

[0024] In this embodiment, the linearized state-space model A detail for: (1) in, A detail A is an n×n matrix; p For the state matrix of the new energy power grid, E d For the interface matrix of DC power generation unit to new energy power generation unit, F d E is the interface matrix for converting new energy power generation units to DC power generation units. a For the interface matrix of AC power generation unit to new energy power generation unit, F a This is the interface matrix for converting new energy power generation units to AC power generation units; specifically: ; ; ; ; ; Among them, A p1 A p2 and A pn These are the state matrices for the 1st, 2nd, and nth new energy power generation units, respectively; E d1 E d2 and E dn1 These are the interface matrices for transforming DC grid state variables to the state variables of the 1st, 2nd, and n1th new energy power generation units, respectively; F d1 F d2 and F dn1 These are the interface matrices for transforming the state variables of the 1st, 2nd, and n1th new energy power generation units to the state variables of the DC grid; E a1 E a2 and E an1 These are the interface matrices for transforming AC grid state variables to the state variables of the 1st, 2nd, and n1st new energy power generation units, respectively; F a1 F a2and F an1 These are the interface matrices for transforming the state variables of the 1st, 2nd, and n1th new energy power generation units to the state variables of the AC power grid, respectively. It is worth noting that the state variables of the power generation unit are vectors composed of multiple variables; the interface matrix is ​​a mapping matrix of the power grid state variables on both sides of the interface.

[0025] Specifically, in this scheme, the initial operating point (i.e., steady-state operating point) of the photovoltaic grid-connected system is first obtained through power flow calculation. x 0, and obtain using the small-signal linearization method. n Each new energy power generation unit is at its initial operating point. x A linearized state-space model at point 0 is used to solve the th... i State matrix of each new energy power generation unit A pi The formula is expressed as: ; In equation (2), d represents the differential. t For a moment, A pi For the first i The state matrix of each new energy power generation unit is derived from known circuit equations. x pi For the first i The state variables of the nth new energy power generation unit, i.e. the nth i The state value of a new energy power generation unit on specified indicators (such as integral link, current and voltage); Δ x pi for x pi The increment; B pi For the first i The input matrix of the grid-connected system when a new energy power generation unit is connected to the grid is the matrix composed of the target parameters given by the power grid dispatch plan; u pi For the first i The input variables of each new energy power generation unit, i.e., the vector composed of current state parameters, Δ u pi express u pi The increment; 1 ≤ i ≤ n .

[0026] Reference Figure 2 The dense pattern separation method includes the following steps: Step 1: Obtain the state variable vector x={x} of the new energy grid-connected system. r|1≤r≤a1×n1+a2×n2+a×n} and linearized state-space model A detail According to different time scales A detail After performing block rearrangement, we get: ; Among them, A pF A pM A pL The original A detail A block consisting of elements at various time scales, including microseconds (µs), milliseconds (ms), and seconds (s); the subscripts F, M, and L represent time scales of µs, ms, and s, respectively.

[0027] Using the above formulas, the centers of the Gale disks are qualitatively divided into three categories based on different time scales, thus allowing for a preliminary distinction of the weakly coupled regions of dense resonance modes.

[0028] St2, A after block arrangement detail Define an index for the degree of coupling in dense patterns; ; Where k is an indicator of the degree of dense mode coupling of the system (i.e., the new energy grid-connected system); After block rearrangement The element in the i-th row and j-th column, After block rearrangement The element in the i-th row and i-th column; the maximum value of row number i is n, which is the number of new energy power generation units.

[0029] Step 3: Determine if k > k th +c0; c0 is the set float value, which can be set to 0. <c0<0.1; k th This is the oscillation coupling threshold; If so, it is determined that the new energy grid-connected system will experience intense oscillations; No, proceed to step St4.

[0030] Let the damping coefficient D and inertia coefficient K of the i-th renewable energy generation unit in the renewable energy grid-connected system after m iterations of the renewable energy grid operation status be denoted as... and The initial value of m is 0. In this step, k th The calculation formula is: ; In the formula, , , and The damping coefficients of the new energy power generation unit after the m-th iteration of the system are respectively. and inertia coefficient The average value.

[0031] Specifically, when the system state variable vector x passes through the equilibrium point, because only the acceleration is equal to 0, but the velocity is not equal to 0, an overshoot phenomenon occurs, causing x to deviate from the equilibrium point; any change in any state of x r Overshoot creates deviation Δx r This incentivizes the readjustment of the coupled system, and thus inertia leads to the equilibrium point of the coupled system and the controlled object x. r The equilibrium points are out of sync, causing system oscillations; at this time, the state variable x r It has inertia, and its expression is: ; n1 and n2 are the total number of DC state variables, AC state variables, and new energy power generation unit state variables, respectively.

[0032] In complex grid-connected systems, the real part of the eigenvalues ​​corresponds to damping characteristics, and the imaginary part corresponds to the oscillation frequency. To predict system dynamics without fully solving for the eigenvalues, this invention modifies the state equations... Perform second-order physical equivalence. This is a transition term; for the r-th state variable x of the system r Its dynamic behavior is equivalent to a second-order forced oscillation system: ; In the above formula, K and D are x r The corresponding inertia coefficient and damping coefficient of the power generation unit are: ; This formula describes the state variable Δx r The equation describes the coupling relationship between acceleration, velocity, and state variables. The relationship between velocity and state variables reflects inertia and oscillation, while the relationship between acceleration and velocity reflects damping and stability. The solution to this equation reflects the dynamic response of the state variable increment Δx at the equilibrium point. The state variable x after the disturbance... r If the motion of a system is bounded, the system is stable; if it is unbounded, it is unstable and divergent; if it is bounded but does not converge, it is critically stable.

[0033] Step 4: Determine if k < k th -c0; Yes, it means the distance between the disks is greater than 1. The coupling between modes is weak, and dense oscillation modes do not occur by default; The inertia coefficient K during the iteration of the new energy power grid m The average value; No, then k th -c0 <k< k th +c0, i.e., k ≈ k th At this point, it is possible to further determine whether there is a risk of strong coupling between modes, so as to optimize parameters and reduce the risk of coupling between modes.

[0034] Specifically, this implementation defines the sensitivity of higher-order terms. If the sensitivity of higher-order terms is significant, then in step St4, it is determined that there is a risk of strong mode coupling, and further parameter optimization of the new energy power generation unit is required. That is, when k is satisfied... th -c0 <k< k th If +c0 and the higher-order terms have significant sensitivity, then the system is considered to have a risk of strong mode coupling.

[0035] The criterion for determining the significant sensitivity of higher-order terms is that the sensitivity index of all parameters is greater than the set threshold.

[0036] Reference Figure 3 Controllable parameter selection and each state variable x r The process of calculating the sensitivity of higher-order terms includes the following steps: S41. Define a parameter set p = [p1, p2, ..., p u ], p1, p2 and p u Let p be the 1st, 2nd, and uth parameter vectors, respectively, and each parameter vector corresponds to an operating condition of the power generation unit; define parameter vector p. h The sensitivity is S h = x r / p h p h Let u be the h-th parameter vector in p, where 1 ≤ h ≤ u; S42. Calculate the parameter vectors p in the parameter set p. h Participating factors ; ; In the formula, S h = x r / p h S j = x r / p j S h and S j p h and p j Sensitivity, p j Let ω be the j-th parameter vector in p;h and ω j p h and p j The weight contribution of the oscillation mode can be calculated using the existing eigenvalue-based mode weighting method; denominator For all parameters {p j The weighted sum of the absolute values ​​of the sensitivity of |1≤j≤u}; S43. Calculate the parameter vectors p. h First-order and second-order sensitivities, if all parameter vectors p h If the second-order sensitivity is greater than the set value, for example, 0.5, then it is judged that the higher-order term sensitivity is significant. Let the h-th parameter vector p h The first-order sensitivity is denoted as: h h = α h / p Where, α h The participation factor of the h-th parameter vector is the first-order sensitivity of the parameter vector, which is the derivative of its participation factor with respect to the parameter set p, specifically the first-order derivative. Let the h-th parameter vector p h The second-order sensitivity is denoted as That is, the second derivative of the participating factor with respect to the parameter set p, expressed by the formula: ; S44. Sort the first-order sensitivities of each parameter vector in descending order, and select the top y first-order sensitivities h. h The corresponding parameter vector serves as the key parameter; y is the set value; S45. Select controllable parameters from the key parameters, denoted as θ. g ;θ g This refers to the power generation unit (state variable x) in step St4. r The optimization object (corresponding power generation unit) is essentially a vector containing multiple parameters.

[0037] In this implementation, when St4 determines that there is a risk of strong mode coupling, the coupling degree of the key parameter k(θ) is minimized. g The controllable parameter θ of each new energy power generation unit is used as the objective function. g Optimize.

[0038] The objective function is expressed as: ; Indicates the controllable parameter θ g A after time-division detail The element in the i-th row and j-th column, To represent the controllable parameter θ g A after time-division detail The element in the i-th row and i-th column.

[0039] The controllable parameter θ of the i-th new energy power generation unit g The iterative optimization steps are as follows: S1. Obtain the controllable parameters θ of the i-th new energy power generation unit. g And calculate the controllable parameter θ. g A at that time detail , denoted as A detail (θ g );θ g The initial value is the filtering result of step S45.

[0040] S2, regarding the controllable parameter θ g The second-order correction update is performed, and the formula is expressed as: ; ; ; ; ; Where, θ g = , represents the current controllable parameters, i.e., the controllable parameters after m iterations; k is the coupling degree index vector of the system, k m This is an index of the degree of dense mode coupling after the m-th iteration of the overall controllable parameters of the new energy grid-connected system. It is worth noting that the iteration of the controllable parameters of any new energy power generation unit belongs to the iteration of the overall controllable parameters of the new energy grid-connected system. The learning rate can be set to η = 0.01.

[0041] This represents the controllable parameters of the new energy grid-connected system after m+1 iterations. This represents the controllable parameters after m iterations. This represents the controllable parameters after m-1 iterations; the initial value of m is 0. The initial value is 0; and Let them represent first-order terms and second-order terms, respectively. This indicates the first-order derivative. This represents second-order differentiation; Considering the higher-order nonlinear effects of the new energy grid-connected system, a second-order term is incorporated. It can accelerate convergence, second order 2 / θ2 g quantization parameter second perturbation ensures that optimization targets the strongest coupling; In new energy grid-connected systems, to ensure that the radius of the disk is reduced after optimization, it is necessary to separate dense eigenvalues, that is, to make the state matrix A... detail The sum of the absolute values ​​of the off-diagonal elements in each row is less than 1, until the controllable parameters meet the convergence condition.

[0042] S3. Determine whether the controllable parameters have converged iteratively; If yes, then execute the controllable parameters after the latest iteration; If not, update m to m+1 and then return to step S1.

[0043] The condition for convergence of the controllable parameter iteration is: ; Alternatively, the number of iterations may reach a set value, i.e., m = the set number of iterations K, specifically K = 100.

[0044] k m+1 Indicate execution The dense pattern coupling degree index k of the time system, k m+1 Indicate execution The dense mode coupling degree index k of the time system. Convergence threshold ε0 = 10 -4 .

[0045] This optimization process is based on the conservative bound of the Gaelic disk, which shrinks to a radius greater than [missing information] under high coupling. This achieves mode separation and avoids the "root loss" problem caused by accurate eigenvalue solving; low-coupling modes do not require this stage.

[0046] Of course, those skilled in the art will recognize that the present invention is not limited to the details of the exemplary embodiments described above, but also includes the same or similar structures that can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0047] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

[0048] The technologies, shapes, and structures not described in detail in this invention are all known technologies.

Claims

1. A method for separating dense modes in a new energy grid-connected system based on Gale disk parameter optimization, characterized in that, Includes the following steps: Step 1: Obtain the state variable vector x and linearized state-space model of the new energy grid-connected system. A detail ∈R n×n Where n is the number of new energy power generation units, and x is the set of state variables of all power generation units in the grid-connected system; according to different time scales... A detail Perform block rearrangement; The time scale is divided into three levels: microseconds, milliseconds, and seconds. The block arrangement is the matrix... A detail Reorganize according to the speed of the time scale; St2, A after block arrangement detail Calculate the density mode coupling degree index of each new energy power generation unit; the density mode coupling degree index of the system is A after block arrangement. detail The summation of the absolute values ​​of the off-diagonal elements in a single row of a matrix along the row dimension, which is the ratio of the absolute values ​​of the diagonal elements in the same row. If the dense pattern coupling degree index > k th +c0; indicates that the new energy grid-connected system will experience intense oscillations. If the dense pattern coupling degree index < k th -c0; indicates that no dense oscillation mode has occurred. c0 is the set float value, k th This is the oscillation coupling threshold.

2. The method for separating dense modes of new energy grid-connected systems based on Gale disk parameter optimization as described in claim 1, characterized in that: Among them, A p The state matrix of the new energy power grid; E d F is the interface matrix for the transformation from DC power generation units to new energy power generation units. It is represented by an n1-dimensional column vector of interface coefficients for the transformation of DC grid state variables to the state variables of each new energy power generation unit, where n1 is the number of DC power generation units. d The interface matrix for converting new energy power generation units to DC power generation units is represented by an n1-dimensional row vector of interface coefficients for the state variables of new energy power generation units to be converted to state variables of DC grids. E a This is the interface matrix for converting AC power generation units to new energy power generation units, representing the coupling relationship between DC grid state variables and new energy power generation units; F a This is the interface matrix for converting new energy power generation units to AC power generation units, representing the coupling relationship between new energy power generation units and the AC power grid.

3. The method for separating dense modes of new energy grid-connected systems based on Gale disk parameter optimization as described in claim 1, characterized in that, After block rearrangement, A detail for: A pF A pM A pL A respectively detail Rearranged block matrices at various time scales including microseconds (µs), milliseconds (ms), and seconds (s); The dense mode separation method for new energy grid-connected systems based on Gale disk parameter optimization as described in claim 1 is characterized in that, if the dense mode coupling degree index is located in the interval (k... th -c0, k th If the second-order sensitivity of the specified parameters of each new energy power generation unit is calculated, and if the second-order sensitivity of the parameters is greater than the set threshold, it is determined that the new energy power generation unit has a strong mode coupling risk. Let the set of specified parameters of the power generation unit be referred to as parameter set p; the parameters correspond one-to-one with the operating conditions of the new energy power grid, and are a one-dimensional vector composed of multiple target parameters; the h-th parameter p in p. h The second-order sensitivity is denoted as ; For p h Participation factors; let the state variables of the power generation unit be related to p h The derivative of the parameter p of the power generation unit is denoted as p. h Sensitivity S h ; Parameter p h The second-order sensitivity is: Among them, S j p is the j-th parameter in p. j The sensitivity, u is the total number of parameters in p; ω h and ω j p h and p j Weighted contribution to oscillation modes.

4. The method for separating dense modes of new energy grid-connected systems based on Gale disk parameter optimization as described in claim 3, characterized in that, When there is a risk of strong mode coupling in the new energy power generation unit, controllable parameters in the parameter set p of the new energy power generation unit are selected for optimization. The selection method for controllable parameters is as follows: calculate each parameter p. h The first-order sensitivities are sorted in descending order, and the parameters corresponding to the top y first-order sensitivities are selected as key parameters. From the key parameters, adjustable parameters are selected as controllable parameters; parameter p h The first-order sensitivity is obtained by taking the derivative of its participation factor with respect to the parameter set p in vector form; The optimization objective of the controllable parameters is: in, Indicates the controllable parameter θ g A after time-division detail The element in the i-th row and j-th column, To represent the controllable parameter θ g A after time-division detail The element in the i-th row and i-th column; Indicates the controllable parameter θ g A measure of the degree of dense pattern coupling in a time system.

5. The method for separating dense modes of new energy grid-connected systems based on Gale disk parameter optimization as described in claim 4, characterized in that, The controllable parameter θ of the new energy power generation unit g The optimization method is as follows: ， ； Where, θ g = , representing the controllable parameters after the current iteration m times; and Let represent the controllable parameters after m+1 and m-1 iterations, respectively; the initial value of m is 0. The initial value is 0; and Let them represent first-order terms and second-order terms, respectively. This indicates the first-order derivative. This indicates the second derivative.

6. The method for separating dense modes of new energy grid-connected systems based on Gale disk parameter optimization as described in claim 5, characterized in that, in, After block rearrangement The element in the i-th row and j-th column, After block rearrangement The element in the i-th row and i-th column; k m The dense mode coupling degree index of the system after the m-th iteration is the set of controllable parameters of the system.

7. The method for separating dense modes of new energy grid-connected systems based on Gale disk parameter optimization as described in claim 4, characterized in that, p h Participating factors The result is a normalized product of the absolute value of sensitivity and the weight contribution in the parameter dimension.

8. The method for separating dense modes of new energy grid-connected systems based on Gale disk parameter optimization as described in claim 1, characterized in that, Oscillation coupling threshold k of new energy power generation unit th Take the average value of the damping coefficient of the new energy power generation unit. and average inertia coefficient The ratio of the square root of .

9. A system for implementing the dense mode separation method for new energy grid-connected systems based on Gale disk parameter optimization as described in any one of claims 1-8, characterized in that, include: The state-space model building module is used to establish harmonic state-space models of new energy fields and grid-connected systems at different time scales. It analyzes the inertial equation expressions of each component considering the damping effect to obtain the Gael disk block matrix at different time scales, i.e., the state-space model after block arrangement. A detail ; The coupling index calculation module is used for the state-space model based on the block arrangement. A detail The coupling degree index of dense mode for each new energy power generation unit is calculated, and the boundary conditions of the coupling degree index of dense mode are constructed based on the Gale disk to achieve dense mode separation. The parameter iterative optimization module performs iterative optimization of system parameters with strong coupling risks, taking the minimization of the Gehr circle radius and the stability of the new energy grid-connected system as constraints for different coupling degrees.

10. A storage medium, characterized in that, The system contains a computer program that, when executed, is used to implement the dense mode separation method for new energy grid-connected systems based on Gale disk parameter optimization as described in any one of claims 1-8.

Images