Multi-infeed series-parallel new energy station stability evaluation and optimization method based on equivalent external characteristics of network construction type converter

Abstract

The invention discloses a stability evaluation and optimization method for a multi-infeed series-parallel new energy station based on equivalent external characteristics of a network construction type converter, and belongs to the technical field of electric power automation. Comprising the following steps: constructing a new energy multi-infeed hybrid system stability analysis closed-loop dynamic model, and simplifying a network construction type admittance transfer function; based on a stability analysis closed-loop dynamic model of the new energy multi-infeed series-parallel system, simplifying the new energy multi-infeed series-parallel system into an equivalent single-infeed system, and evaluating the stability of the equivalent single-infeed system through a generalized operation short-circuit ratio and a critical generalized operation short-circuit ratio; analyzing and quantifying the influence of key parameters of the network construction type equipment on the stability of the system through a parameter sensitivity analysis method; a node participation factor is used as an evaluation index, and control parameters are optimized based on a stability optimization method of key parameters of network construction type equipment of system weak nodes. The method provided by the invention has good effectiveness and reliability, and can realize gradual improvement and refined regulation and control of the system stability.

Description

Technical Field

[0001] This invention relates to the field of power automation technology, and in particular to a method for stability assessment and optimization of multi-infeed hybrid renewable energy power plants based on the equivalent external characteristics of grid-connected converters. Background Technology

[0002] With the increasingly severe environmental pollution caused by traditional fossil fuel power generation, new energy power generation technologies, represented by wind power and photovoltaic power generation, have developed rapidly and been widely applied. However, the grid-connected operation of large-scale new energy converters has significantly altered the dynamic characteristics of traditional power systems, posing new challenges to the safe and stable operation of the power grid. On the one hand, new energy power plants mostly adopt grid-following (GFL) control methods for grid connection. As the connection ratio increases, the system strength decreases, easily leading to subsynchronous oscillation (SSO) instability risks dominated by phase-locked loops (PLLs). On the other hand, the output of new energy is affected by environmental conditions, exhibiting significant randomness and volatility. In actual operation, the output level of equipment and port voltage often deviate from the rated values, further threatening grid stability. To address these issues, converters using grid-forming (GFM) control, with their synchronous power control characteristics, can effectively suppress subsynchronous oscillations, enhance adaptability to weak grids, and provide active support for grid voltage and frequency, becoming an important technical direction for improving grid stability.

[0003] Currently, for heterogeneous grid-connected / parallel grid systems, it has been clearly established that grid-connected control converters can effectively suppress subsynchronous oscillations and enhance system stability. Existing technologies focus on the capacity allocation of grid-connected equipment, site selection optimization, and factors affecting system stability: Regarding capacity allocation, optimization criteria based on the short-circuit ratio have been established, and methods for solving the optimal capacity of grid-connected units have been proposed; regarding site selection optimization, energy storage site selection and capacity determination models considering various scenarios and optimization models based on intelligent algorithms have been constructed, and optimized site selection of grid-connected equipment has been achieved based on the generalized short-circuit ratio. However, existing technologies mostly focus on equipment capacity analysis, with insufficient analysis of the interaction mechanism between grid-connected equipment and system strength. While the impact of some parameters on the stability of the system under small disturbances has been qualitatively analyzed in terms of the mechanism of system parameters and control parameters on stability, and synchronous stability studies have been conducted in conjunction with the short-circuit ratio, quantitative analysis of key parameters and system stability and stability margin is lacking, and existing analysis methods are difficult to apply to multi-feedback systems of new energy sources. Meanwhile, considering the high cost of adding grid-type control units and modifying grid-type units, a stability assessment and optimization method for multi-infeed hybrid new energy power plants based on the equivalent external characteristics of grid-type converters is needed. Under the premise of ensuring system stability margin, the lower limit of grid-type converter ratio should be clarified, and the synergistic effect and interaction mechanism of system parameters and control parameters on stability should be revealed. Summary of the Invention

[0004] The purpose of this invention is to propose a stability assessment and optimization method for multi-infeed hybrid renewable energy power plants based on the equivalent external characteristics of grid-connected converters, comprising the following steps:

[0005] Considering the grid-connecting equipment in the feed-in system of large-scale new energy power plants, a closed-loop dynamic model for stability analysis of multi-feed hybrid new energy systems is constructed, and the grid-type admittance transfer function is simplified based on the external characteristics of the grid-connecting equipment.

[0006] Based on the closed-loop dynamic model of stability analysis of new energy multi-infeed hybrid system, the new energy multi-infeed hybrid system is simplified into an equivalent single-infeed system. The stability of the equivalent single-infeed system is evaluated by the generalized operating short-circuit ratio and the critical generalized operating short-circuit ratio.

[0007] Based on the generalized operating short-circuit ratio and the critical generalized operating short-circuit ratio, the impact of key parameters of grid-type equipment on system stability is analyzed and quantified by parameter sensitivity analysis.

[0008] Using node participation factor as an evaluation index, the stability optimization method of key parameters of network-type equipment based on weak nodes in the system is used to optimize control parameters when the system operating conditions change.

[0009] Furthermore, the simplified network admittance transfer function is:

[0010]

[0011] Among them, Y GFM (s) is the admittance transfer function matrix of the network equipment, Y DQ (s) is the admittance matrix of the voltage source branch of the network equipment, Y θ (s) represents the virtual ground branch admittance matrix of the network equipment, Y dq.i (s) is the admittance matrix of the voltage source branch of the i-th grid-connected converter, Z θ.i (s) represents the virtual ground branch equivalent impedance of the network equipment.

[0012] Furthermore, the generalized operating short-circuit ratio satisfies:

[0013]

[0014] Where, η gOSCR The short-circuit ratio is the generalized operating ratio, and n is the number of equivalent single-infeed systems. Represents the Kronecker product; y min.j x is the left eigenvector corresponding to the smallest eigenvalue; min.j Y is the right eigenvector corresponding to the smallest eigenvalue. O.j F(s) is the equivalent device admittance matrix; F(s) is the admittance transformation matrix.

[0015] Furthermore, the critical generalized operating short-circuit ratio satisfies:

[0016]

[0017] Where, η CgOSCR For the critical generalized operating short-circuit ratio, s d These are the dominant characteristic roots of the system.

[0018] Furthermore, the minimum critical generalized operating short-circuit ratio required by the system is determined solely by the phase-locked loop of the network equipment.

[0019] Furthermore, using the node participation factor as an evaluation index, the specific steps for optimizing control parameters based on the stability optimization method of key parameters of network-type equipment with weak nodes in the system when the system operating conditions change are as follows:

[0020] Step 1: Input the actual operating parameters of each device, the actual operating data of each node, and the network structure parameters under the actual operating conditions of the system;

[0021] Step 2: Calculate the equivalent nodal admittance matrix B on the network side under this operating condition. re Solve for the generalized operating short-circuit ratio η gOSCR And the corresponding left and right feature vectors;

[0022] Step 3: Using the left and right eigenvectors, the system is simplified to an equivalent single-feed system based on perturbation theory, and then the critical generalized operating short-circuit ratio η is solved.CgOSCR And the participation factors of each node;

[0023] Step 4: Optimize the network equipment parameters of the weakest node with the largest participation factor;

[0024] Step 5: Determine whether the system stability index meets the threshold requirement. If yes, end the process; otherwise, return to repeat steps 2 to 5 until the threshold requirement is met.

[0025] The beneficial effects of this invention are as follows:

[0026] 1. The stability quantification evaluation method proposed in this invention can accurately reflect the stability characteristics of the actual system under different network structures and operating conditions through its equivalent single-infeed system model. At the same time, the admittance transfer function matrix constructed by the equivalent multi-infeed system is suitable for modeling different types of new energy feed-in systems.

[0027] 2. The parameter optimization method proposed in this invention can accurately locate weak nodes in the system and efficiently improve system stability. Parameter optimization based on weak nodes is more effective when taking the same number of iterations. Attached Figure Description

[0028] Figure 1 This is a flowchart of a method for stability assessment and optimization of multi-infeed hybrid renewable energy power plants based on the equivalent external characteristics of grid-type converters.

[0029] Figure 2 To feed more new energy into the system.

[0030] Figure 3 This describes the closed-loop dynamic characteristics of a multi-feedback renewable energy system.

[0031] Figure 4 It is a GFM converter based on virtual synchronous machine control.

[0032] Figure 5 This is a schematic diagram of the voltage support characteristics of a GFM converter.

[0033] Figure 6 It is a 4-machine, 10-node system.

[0034] Figure 7 (a) and (b) represent the sensitivity of the GFM current control loop parameters.

[0035] Figure 8 (a) and (b) represent the sensitivity of the GFM voltage control loop parameters.

[0036] Figure 9 The sensitivity of the GFM current feedforward coefficient.

[0037] Figure 10 This is a stability optimization method based on network device parameters.

[0038] Figure 11 This is a comparison of the dominant eigenvalues ​​when the impedance coefficient changes.

[0039] Figure 12 This is a comparison of the dominant eigenvalues ​​when the power coefficient changes.

[0040] Figure 13 These are system stability indicators under different operating conditions.

[0041] Figure 14 (a)(b)(c)(d)(e)(f)(g)(h) represent different scenarios η gOSCR Results of two rounds of optimization.

[0042] Figure 15 This represents the active power of wind farm 1 in Example 3.

[0043] Figure 16 This represents the active power of wind farm 1 in Example 7. Detailed Implementation

[0044] This invention proposes a method for stability assessment and optimization of multi-infeed hybrid renewable energy power plants based on the equivalent external characteristics of grid-type converters. The invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0045] Figure 1 The flowchart for the stability assessment and optimization method of multi-infeed hybrid renewable energy power plants based on the equivalent external characteristics of grid-connected converters is as follows:

[0046] Step A. Considering the grid-connecting equipment in the feed-in system of large-scale new energy power plants, construct a closed-loop dynamic model for stability analysis of the multi-feed-in hybrid system of new energy, and simplify the grid-connecting admittance transfer function according to the external characteristics of the grid-connecting equipment;

[0047] like Figure 2 As shown, the multi-renewable energy power plant feed-in system consists of n new energy access nodes, m intermediate passive nodes, and an infinite busbar. Existing data shows that, to improve the same grid strength, the required grid-connecting equipment is lowest when retrofitting at the new energy power plant side compared to adding transformers to the high-voltage side. Therefore, it is assumed that all new energy power plants are equipped with a certain capacity of grid-connecting equipment to ensure a high stability margin under rated operating conditions, better meeting the actual operational needs of the project.

[0048] Based on the above conditions, establish Figure 2 The linearized model of the multi-feed system in the global xy coordinate system is shown below. Figure 3 As shown. Where, ΔU' xy ,ΔI' xyS represents an n-dimensional column vector consisting of the voltage and current increments on the access side of the new energy equipment; L S M The diagonal matrix consists of the device-side access network and the network capacity, respectively, satisfying... S M =αS L (i = 1, 2, ..., n), where α is the proportion of network equipment to the total capacity of the station; Y GFL Y GFM These are the admittance transfer function matrices of the network connection and network construction devices, respectively, B Grid To eliminate the network admittance matrix of intermediate nodes; Indicates the Kronecker product; I n Let F(s) be an n-dimensional identity matrix, and F(s) be the admittance transformation matrix, satisfying:

[0049]

[0050] Due to significant differences in actual operating conditions, the system closed-loop transfer function needs to be modified accordingly. When the two types of converters achieve approximate decoupling, their admittance transfer function matrices in the xy coordinate system both exhibit linear functions of voltage and do not contain power-related terms. Based on this, the characteristic differences between grid-connected and grid-connected equipment caused by actual operating conditions can be approximately ignored. Furthermore, to simplify the analysis process of the closed-loop characteristic equation under actual operating conditions, the equipment port voltage is approximated using a given operating voltage. Through the above processing, the closed-loop characteristic equation considering actual operating conditions can be reformulated as:

[0051]

[0052] In the formula: β=α / (1-α); V i P i These represent the port voltage and active power of the i-th new energy device, respectively.

[0053] The Virtual Synchronous Generator (VSG), as a typical grid-type converter control technology, can be equivalent to other types of grid-type control schemes in engineering applications. Without loss of generality, we take a grid-type converter with VSG control as an example to derive the GFM admittance transfer matrix. Figure 4 The diagram shown is a control block diagram of a VSG-based GFM converter, where V... abc I is the grid connection point voltage; abc I cabc These are the grid-side current and the filter inductor current, respectively; C f L f For filter inductors and capacitors; K i Kff These are the current feedforward coefficient and voltage feedforward coefficient, respectively; J and D are the virtual inertia and damping coefficients of the rocking equation. The parameters are represented in vector form (e.g., ...). At this time, the port voltage and current of the grid-connected inverter and their reference values ​​are as follows:

[0054]

[0055] In the formula: G i (s)=K pi +K ii / s is the transfer function of the current loop PI controller, K pi K ii These are the proportional and integral control coefficients of the transfer function, respectively; G v (s)=K pv +K iv / s is the transfer function of the voltage loop PI controller, K pv K iv These are the proportional and integral control coefficients of the transfer function, respectively; K ff (s)=1 / (1+T vf s) is the voltage feedforward transfer function, T vf K is the delay factor; i This refers to the current feedforward coefficient; As a reference voltage, in the small-signal model, Neglecting the delay of the PWM stage, linearizing the above equation yields:

[0056]

[0057] Y M (s)=(G v (s)G i (s)-H F (s)H L (s)-jωC f G i (s)

[0058] +(s+jω)C f H L (s)) / (H L (s)-K i G i (s))

[0059]

[0060] In the formula: These represent the small-signal offsets. The above equation can be further simplified, and the voltage feedforward transfer function within the frequency band K ff (s)≈1; and (s+jω)C fThe effect mainly occurs in the high-frequency range and is negligible in the low-frequency range. After simplification, Y M The matrix expression in the dq coordinate system is as follows:

[0061] -ΔI dq =Y dq (s)ΔV dq

[0062] Y dq (s)=B dq (s)F(s)

[0063]

[0064] H V (s)=G i (s)(G v (s)+sC f )

[0065]

[0066] In the formula: ξ is the current feedforward equivalent conversion coefficient, T L (s) is the transformation coefficient matrix. It is worth noting that when the system voltage feedforward coefficient is 1, B dq (s) will be simplified to a one-dimensional function in the frequency domain of s.

[0067] Therefore, the external characteristics of a grid-type converter can be equivalently represented as the grid-side admittance controlled by parameters. The equivalent part, Bdq, is in two-dimensional matrix form. To simplify the subsequent analysis, a dimensionality reduction strategy is adopted, transforming the admittance matrix into a single typical value: the maximum singular value of the dq impedance within the frequency band of interest is used for equivalent substitution, and the maximum value within the subsynchronous frequency band is selected as the final equivalent impedance. Its specific expression is as follows:

[0068]

[0069] In summary, the admittance transfer matrix of the grid-type converter can be simplified as follows:

[0070]

[0071] From the admittance transfer matrix Y dq (s) Analysis shows that the voltage support characteristics of GFM are similar to... Figure 5 The synchronous generator shown has an equivalent structure. According to the constant flux linkage model theory, a synchronous generator can be equivalent to a circuit structure with a voltage source and a synchronous inductor connected in series. The synchronous inductor can adjust the flux linkage through an automatic voltage regulator, thereby achieving terminal voltage control. The admittance model Y of the GFM... dq (s) It possesses a similar voltage support capability. As can be seen from the above formula, this capability mainly stems from the frequency domain transfer function of the filter inductor L.f The equivalent regulatory mechanism.

[0072] The admittance matrix of the above-mentioned grid-type converter is derived from the local dq coordinate system. When modeling the new energy feed-in system, it needs to be transformed to the global xy coordinate system. Given:

[0073]

[0074] Substituting, we get:

[0075]

[0076] In the formula: I0 and V0 are the port current and voltage vectors before the disturbance, respectively. In summary, the admittance transfer function matrix of a multi-infeed grid converter in the global coordinate system can be expressed as:

[0077]

[0078] Among them, Y GFM (s) is the admittance transfer function matrix of the network equipment, Y DQ (s) is the admittance matrix of the voltage source branch of the network equipment, Y θ (s) represents the virtual ground branch admittance matrix of the network equipment, Y dq.i (s) is the admittance matrix of the voltage source branch of the i-th grid-connected converter, Z θ.i (s) represents the virtual ground branch equivalent impedance of the network equipment.

[0079] In summary, the admittance transfer function matrix of network-connected equipment can be simplified into two parts: the voltage source branch and the virtual ground branch. Although the simplification of the GFM admittance matrix is ​​achieved from the perspective of equipment support characteristics, the dimensionality of the original system admittance transfer function is not reduced. Directly solving and analyzing the small-signal stability of the system still faces problems such as high dimensionality and large computational load. In addition, the traditional generalized short-circuit ratio criterion cannot be directly applied to the stability analysis of heterogeneous converters. Therefore, to quantify the relationship between network-connected equipment parameters and system stability, a small-signal stability assessment method suitable for this system is needed.

[0080] Step B. Based on the closed-loop dynamic model of the stability analysis of the multi-infeed hybrid system of new energy, the multi-infeed hybrid system of new energy is simplified into an equivalent single-infeed system. The stability of the equivalent single-infeed system is evaluated by the generalized operating short-circuit ratio and the critical generalized operating short-circuit ratio, mainly as follows:

[0081] The above analysis shows that the impedance admittance matrix of the network equipment consists of two parts, which can be rewritten as:

[0082] det(L(s)+M(s))=0

[0083] in:

[0084]

[0085] As shown in the equation, the small-signal stability of this system is jointly determined by L(s) and M(s). L(s) consists of the admittance matrix of the network-connected equipment and the equivalent network-side admittance matrix, while M(s) consists of the virtual ground-to-ground branches of the network-connected equipment, which can be considered as an additive perturbation of the L(s) system by M(s). For the original system, based on the eigenspace perturbation theory, the equivalent single-feed system characterizing the system stability is solved. Specifically: based on the system eigenvalues, the equivalent network-side admittance matrix... There are n eigenvalues ​​λ i (i = 1, 2, ..., n), construct the two sets of feature subspaces E(X) corresponding to the expanded dimension. i ), E(Y i ).in, Where y i x i These are the equivalent admittance matrices of the network side with respect to the eigenvalue λ. i The left and right eigenvectors. Based on the obtained eigensubspace, the system can be decoupled into n single-feed systems with different short-circuit ratios, as follows:

[0086]

[0087] Y O.j (s)=Y GFL.j (s)+βY θ.j

[0088] In the formula: y i.j x i.j Y represents the j-th element of the eigenvector corresponding to the i-th eigenvalue. GFL.j Y θ.j These are the admittance matrix of the j-th grid-connected converter and the virtual branch admittance matrix of the grid-connected converter, respectively.

[0089] The original system is decoupled into n equivalent single-feed systems. Each equivalent single-feed system consists of a weighted device admittance matrix and equivalent network admittance eigenvalues. The small-disturbance stability of the original system depends on the minimum value λ of the eigenvalues ​​of the decoupled equivalent single-feed systems. min That is, the generalized operational short-circuit ratio (gOSCR), denoted by η. gOSCR This indicates that the following expression is satisfied:

[0090]

[0091] Where, η gOSCR The short-circuit ratio is the generalized operating ratio, and n is the number of equivalent single-infeed systems. Represents the Kronecker product; ymin.j x is the left eigenvector corresponding to the smallest eigenvalue; min.j Y is the right eigenvector corresponding to the smallest eigenvalue. O.j F(s) is the equivalent device admittance matrix; F(s) is the admittance transformation matrix.

[0092] The minimum eigenvalue is derived from the grid-side equivalent admittance matrix. Its physical meaning is to quantify the sensitivity of the converter port voltage to the system injected current, thus characterizing the degree of fluctuation of the system port voltage under current disturbances. To comprehensively evaluate the system stability margin, it is necessary to further investigate the critical value required to maintain stable system operation, namely the critical generalized operational short-circuit ratio (CgOSCR). This parameter satisfies the following condition:

[0093]

[0094] In the formula: s d The dominant characteristic root of the system; the critical operating short-circuit ratio η CgOSCR This is the generalized running short-circuit ratio when the real part of the dominant eigenvalue of the weakest system is zero.

[0095] Unlike traditional short-circuit ratio analysis methods, this approach comprehensively considers the dual impact of grid-connected equipment on system stability when formulating the system's closed-loop characteristic equations: First, grid-connected equipment integrates into the grid-side structure through equivalent admittance branches, providing voltage and power support to the system; second, virtual ground branches act as additive disturbances on the equipment side, affecting the system's dynamic characteristics. Based on the above analysis, the system stability margin is defined as follows:

[0096]

[0097] Step C. Based on the generalized operating short-circuit ratio and the critical generalized operating short-circuit ratio, analyze and quantify the impact of key parameters of network-type equipment on system stability using parameter sensitivity analysis methods, including:

[0098] Based on the above analysis, the admittance transfer matrix of the network equipment is divided into two parts in the system closed-loop matrix: one part is the equivalent voltage source branch, and the other part is the virtual ground branch. Based on this, the quantitative relationship between the control parameters and various indicators of the network equipment is analyzed.

[0099] Voltage source branch equivalent admittance B eq The system's η is determined by the parameters of each control loop. By adjusting the key parameters of the control loops, the equivalent impedance of the network equipment is changed, thereby affecting the system's η. gOSCR Taking a 10-node, 4-machine feed-in system as an example ( Figure 6All four wind farms upgraded 8% of their equipment to grid-connected configurations. Equipment and line parameters are shown in Tables 1-3. Sensitivity analysis was used to assess the impact of key grid-connected equipment parameters on η. gOSCR The impact.

[0100] Table 1 GFM Equipment Parameters

[0101]

[0102] Table 2 GFL Equipment Parameters

[0103]

[0104] Table 3 Line parameters of the 4-machine 10-node system

[0105]

[0106] Figure 7 (a)(b) Figure 8 (a)(b) and Figure 9 The graphs show the sensitivity curves of the system's generalized short-circuit ratio to the key control parameters of the network equipment under different given values. The horizontal and vertical axes represent the system's generalized short-circuit ratio sensitivity curves when the system's η is adjusted to a fixed value based on the given baseline values. gOSCR The change in K. pi K ii K pv K iv For η gOSCR The influence is relatively high, belonging to a highly sensitive parameter, K i The adjustment capability is low, and the ability to regulate the system is weak. High-sensitivity parameters show good results in the initial adjustment phase, but the sensitivity decreases as the parameter increases, and the effect of adjusting the parameter on η becomes less significant. gOSCR The ability to improve the system gradually decreases, and there is a certain sensitivity threshold. Furthermore, different node device parameters have different impacts on the system's generalized short-circuit ratio, influenced by the network-side equivalent admittance matrix. The impact of characteristics.

[0107] Additionally, the parameter for η CgOSCR The influence of the indicator, η CgOSCR Essentially, this is the solution when the real part of the dominant eigenvalue of the system is zero. Analysis shows that in a grid-connected dominant new energy system with a certain proportion of grid-connected equipment, the virtual grounding branch of the grid can be approximated as a perturbation of the dynamics of the grid-connected equipment. η CgOSCR The parameters are determined by the phase-locked loop coefficients of the network-connected equipment and are independent of the parameters of the network-building equipment, as proven below:

[0108] In the admittance transfer matrix of network equipment, the dynamics of the high-frequency inner current loop and the outer loop control (with bandwidth lower than that of the phase-locked loop) are ignored; only the subsynchronous frequency band dominated by the phase-locked loop is considered. Multiplying the right side of the equation by the inverse of F(s) and simplifying, it can be rewritten as:

[0109]

[0110]

[0111] In the formula: V eq Equivalent voltage; K PLLp K PLLi These are the proportional and integral coefficients of the phase-locked loop, respectively; substituting the equations, we can obtain η. CgOSCR The expression is as follows:

[0112]

[0113] As can be seen from the formula, for the subsynchronous frequency band, the network-following equipment and the network-building equipment have a relationship with η. CgOSCR The impact can be considered as two decoupled parts, and is caused by two maximum η values ​​within the dominant frequency band. gOSCR Decision. And βY θ.11 (s) term subject to J j D j The influence of parameter tuning values ​​on βY within the dominant eigenvalue frequency band. θ.11 (s)≈0, which is considered as a perturbation of the dynamics of the network-connected equipment. The critical point of the system is determined by Y. GFL.22 (s) determines that the above formula can be rewritten as:

[0114]

[0115] Solving for the latter term yields the expression for the dominant eigenvalue of the system:

[0116]

[0117] Δ=(K PLLi V eq -η gOSCR K PLLp V eq ω0) 2 +4η gOSCR K PLLi V eq ω0(K PLLp V eq -η gOSCR ω0)

[0118] When the real part of the eigenvalue is zero, it corresponds to η gOSCR That is, the critical value η CgOSCR ,satisfy:

[0119]

[0120] In summary, for a grid-connected renewable energy power station with a certain proportion of grid-connected equipment, the minimum required η of the system is... CgOSCR It is determined by the phase-locked loop of the network equipment and is basically independent of other parameters.

[0121] Step D. Using node participation factor as an evaluation index, optimize control parameters based on the stability optimization method of key parameters of network-type equipment with weak nodes in the system when the system operating conditions change.

[0122] Considering the large number of system equipment parameters and the uncertainty of the sensitivity relationships among these parameters, how can we adjust these parameters to ensure good system stability under different operating conditions? To address this, a parameter optimization method for improving system stability is proposed. Considering that system stability is determined by the weakest equivalent single-feed system, when selecting adjustment equipment, η is used as the parameter optimization factor. gOSCR Using the participation factor of nodes as the standard, priority is given to optimizing weak nodes in the system, η gOSCR The participation factor expression for each node can be represented as:

[0123]

[0124] In the formula: p min.j For η gOSCR The indicator corresponds to the participation factor of each node j, B re.j For B re The j-th diagonal element. Define the node with the largest participating factor as the weak node. First, optimize the equipment parameters of this node to more efficiently improve system stability. Furthermore, when the system has many new energy devices, considering the small number of parameters to be optimized in each round, the number of optimization iterations can be appropriately reduced. After completing the adjustment of the current round, promptly recalculate the weak nodes for the next round of optimization to ensure that all optimized nodes maintain high sensitivity. The proposed optimization method flow is as follows: Figure 10 :

[0125] The specific steps are as follows:

[0126] 1) Input the actual operating parameters of each device, the actual operating data of each node, and the network structure parameters under the actual operating conditions of the system;

[0127] 2) Calculate the equivalent nodal admittance matrix B on the network side under this operating condition. re Solve for the minimum eigenvalue, i.e., the generalized operating short-circuit ratio η. gOSCR And the corresponding left and right feature vectors.

[0128] 3) Using the left and right eigenvectors, the system is simplified to an equivalent single-feed system based on perturbation theory, and then the critical generalized operating short-circuit ratio η is solved. CgOSCR And the participation factors of each node.

[0129] 4) Optimize the network equipment parameters of the weakest node with the largest participation factor to improve system η. gOSCR With η CgOSCR Parameters to improve system stability.

[0130] 5) Determine whether the system stability index meets the threshold requirements. If not, return to repeat steps (2)-(5).

[0131] The following is a detailed description of the experimental results of this embodiment:

[0132] (1) Verify the correctness of the decoupling characteristics of the equivalent single-feed system characterizing system stability, to prove that the decoupled single-feed characteristic equation can accurately reflect the system stability under different operating conditions and system parameters. Impedance coefficient k is set separately. z With power coefficient k p The grid-side impedance and the power of the new energy power station are adjusted proportionally, and the dominant characteristic values ​​of the original system and the equivalent single-infeed system are compared under different conditions.

[0133] Figure 11 and Figure 12 The dominant eigenvalues ​​of the actual new energy hybrid system and the eigenvalue distribution of the equivalent single-infeed system under different coefficient variations are presented. Table 4 details the dominant eigenvalue values ​​for each parameter. The analysis results show that the dominant eigenvalues ​​of the actual new energy hybrid system and the equivalent single-infeed system are highly consistent under different operating conditions. Specifically, regardless of changes in impedance and power coefficients, the eigenvalues ​​of the equivalent single-infeed system and the dominant eigenvalues ​​of the original system are well-aligned, with a maximum relative error of only 4.20%. This result fully verifies that the equivalent single-infeed system model can accurately reflect the stability characteristics of the actual system under different network structures and operating conditions.

[0134] Table 4 Comparison of System Dominant Characteristic Root Errors

[0135]

[0136] (2) Analysis of optimization examples of network equipment parameters under different operating conditions:

[0137] To further verify the applicability of the optimized parameters under different operating conditions, eight typical scenarios from the actual operation of a certain renewable energy power station were selected for analysis. The renewable energy operating port voltage was based on the output voltage of the wind farm's grid-connected equipment; specific parameters are detailed in Table 5. Since the impact of actual operating condition variations on the phase-locked loop parameters of the wind farm's grid-connected equipment is negligible, the critical operating short-circuit ratio of the wind farm remained constant under different operating conditions. The distribution of the system's generalized operating short-circuit ratio for each scenario is shown below. Figure 13 As shown.

[0138] Table 5 Data under different operating conditions

[0139]

[0140] At this point, the critical operating short-circuit ratio on the equipment side is 1.56. When the system operating conditions change, the system's ηgOSCR index changes. Specifically, the operating short-circuit ratio ηgOSCR in examples 1, 3, and 4... gOSCR Less than η CgOSCR The system is unstable. It can be seen that when the output of new energy sources increases and the port voltage level is low, the system η... gOSCR The smaller the value, the more unstable the system tends to be. The stability margin under various operating conditions and η gOSCR The indicators are shown in Table 6:

[0141] Table 6 Results of optimization indicators under different operating conditions

[0142]

[0143] Table 7. Node Participation Factors in Two Rounds of Optimization under Different Operating Conditions

[0144]

[0145] For stable system operation, the parameters of network equipment at weak nodes are prioritized for optimization to improve system stability margin. The stability margin threshold can be flexibly set based on engineering experience; for example, 20% is often used as a reference threshold in conventional engineering. For unstable operation, the parameters of network equipment at weak nodes are first adjusted. After obtaining the system stability indicators after parameter adjustment, a comprehensive judgment is made based on the actual operating status of the system to determine whether a second round of optimization is needed, thereby achieving gradual improvement and refined control of system stability.

[0146] Differential evolution (DE) algorithm was used to optimize the network equipment parameters under different scenarios. The proportional-integral coefficients of the voltage outer loop and current inner loop of the network equipment were selected as optimization parameters. To ensure that the system has a high stability margin under each operating condition, the algorithm was set to iterate 60 times per round, and two rounds of location optimization were carried out. The node participation factors and optimized system indicators for each scenario are detailed in Table 7.

[0147] Figure 14 Figures (a)(b)(c)(d)(e)(f)(g)(h) present the optimization results of two rounds under different scenarios. Considering the uncertainty in the algorithm optimization process, representative compromise results from multiple iterations are selected for display. The red area in the figures represents the region enclosed by the convergence curves of the objective function in each iteration. The analysis results show that each example basically meets the system stability threshold requirements after the first iteration, and the system stability margin is significantly improved after the second round of optimization.

[0148] Figure 15 and Figure 16The active power oscillation characteristics of node 1 are shown in Examples 3 and 7 after initial state and parameter optimization. In Example 3, the system is unstable in the initial state, exhibiting divergent oscillations and gradually becoming unstable after being disturbed. After the first round of optimization, its generalized operating short-circuit ratio... η gOSCR The value was increased to 1.60, exceeding the critical value, at which point the system could converge after a disturbance; after the second round of optimization, η gOSCR The value was further increased to 2.36, significantly enhancing the system's stability margin. From Figure 15 As can be seen, compared with the first round of optimization, the dynamic response performance of the system after the second round of optimization is significantly improved, the convergence time is shortened, and the oscillation amplitude is reduced. For example 7, after two rounds of parameter optimization, the oscillation characteristics of the system are significantly improved, and the dynamic response performance is significantly enhanced. The above results fully verify the effectiveness and reliability of the parameter optimization method proposed in this invention.

Claims

1. A method for stability assessment and optimization of multi-infeed hybrid renewable energy power plants based on the equivalent external characteristics of grid-connected converters, characterized in that, Includes the following steps: Considering the grid-connecting equipment in the feed-in system of large-scale new energy power plants, a closed-loop dynamic model for stability analysis of multi-feed hybrid new energy systems is constructed, and the grid-type admittance transfer function is simplified based on the external characteristics of the grid-connecting equipment. Based on the closed-loop dynamic model of stability analysis of new energy multi-infeed hybrid system, the new energy multi-infeed hybrid system is simplified into an equivalent single-infeed system. The stability of the equivalent single-infeed system is evaluated by the generalized operating short-circuit ratio and the critical generalized operating short-circuit ratio. Based on the generalized operating short-circuit ratio and the critical generalized operating short-circuit ratio, the impact of key parameters of grid-type equipment on system stability is analyzed and quantified by parameter sensitivity analysis. Using node participation factor as an evaluation index, the stability optimization method of key parameters of network-type equipment based on weak nodes in the system is used to optimize control parameters when the system operating conditions change.

2. The method for stability assessment and optimization of multi-infeed hybrid renewable energy power plants based on the equivalent external characteristics of grid-type converters according to claim 1, characterized in that, The simplified network admittance transfer function is: Among them, Y GFM (s) is the admittance transfer function matrix of the network equipment, Y DQ (s) is the admittance matrix of the voltage source branch of the network equipment, Y θ (s) represents the virtual ground branch admittance matrix of the network equipment, Y dq.i (s) is the admittance matrix of the voltage source branch of the i-th grid-connected converter, Z θ.i (s) represents the virtual ground branch equivalent impedance of the network equipment.

3. The method for stability assessment and optimization of multi-infeed hybrid renewable energy power plants based on the equivalent external characteristics of grid-type converters according to claim 1, characterized in that, The generalized operating short-circuit ratio satisfies: Where, η gOSCR The short-circuit ratio is the generalized operating ratio, and n is the number of equivalent single-infeed systems. Represents the Kronecker product; y min.j x is the left eigenvector corresponding to the smallest eigenvalue; min.j Y is the right eigenvector corresponding to the smallest eigenvalue. O.j F(s) is the equivalent device admittance matrix; F(s) is the admittance transformation matrix.

4. The method for stability assessment and optimization of multi-infeed hybrid renewable energy power plants based on the equivalent external characteristics of grid-type converters according to claim 3, characterized in that, The critical generalized operating short-circuit ratio satisfies: Where, η CgOSCR For the critical generalized operating short-circuit ratio, s d These are the dominant characteristic roots of the system.

5. The method for stability assessment and optimization of multi-infeed hybrid renewable energy power plants based on the equivalent external characteristics of grid-type converters according to claim 1 or 4, characterized in that, The minimum critical generalized operating short-circuit ratio required by the system is determined solely by the phase-locked loop of the network equipment.

6. The method for stability assessment and optimization of multi-infeed hybrid renewable energy power plants based on the equivalent external characteristics of grid-type converters according to claim 1, characterized in that, Using node participation factors as evaluation indicators, the specific steps for optimizing control parameters based on the stability optimization method for key parameters of network-type equipment with weak nodes in the system when operating conditions change are as follows: Step 1: Input the actual operating parameters of each device, the actual operating data of each node, and the network structure parameters under the actual operating conditions of the system; Step 2: Calculate the equivalent nodal admittance matrix B on the network side under this operating condition. re Solve for the generalized operating short-circuit ratio η gOSCR And the corresponding left and right feature vectors; Step 3: Using the left and right eigenvectors, the system is simplified to an equivalent single-feed system based on perturbation theory, and then the critical generalized operating short-circuit ratio η is solved. CgOSCR And the participation factors of each node; Step 4: Optimize the network equipment parameters of the weakest node with the largest participation factor; Step 5: Determine whether the system stability index meets the threshold requirement. If yes, end the process; otherwise, return to repeat steps 2 to 5 until the threshold requirement is met.

Images