A method for suppressing ultra-low frequency oscillation based on oscillation source positioning

Abstract

The present invention provides a method for suppressing ultra-low frequency oscillations based on oscillation source positioning, which relates to the technical field of power system stability and control. The method includes performing oscillation mode decomposition on each hydropower generator in a wind-water storage system; determining the generator set as the oscillation source based on the calculated oscillation energy; selecting the acceleration power of the oscillation source as the output of the wind-water storage system, and using the wind storage output as the input; identifying the transfer function matrix of the wind-water storage system; setting an additional controller on the wind storage side, and using the acceleration power of the oscillation source as the input of the additional controller, and calculating the transfer function matrix of the additional controller based on the transfer function matrix of the wind-water storage system; adding the output calculated based on the transfer function matrix of the additional controller to the fixed power side of the wind storage to achieve ultra-low frequency oscillation suppression. This method can fully utilize the flexible and adjustable performance advantages of wind storage, improve the utilization rate of auxiliary service resources and the suppression effect of ultra-low frequency oscillations.

Description

Technical Field

[0001] The present invention relates to the technical field of power system stabilization and control, and in particular to a method for suppressing ultra-low frequency oscillation based on oscillation source positioning. Background Art

[0002] Large power grids with a high proportion of hydropower, such as isolated hydropower DC systems or asynchronously interconnected systems, are at serious risk of ultra-low frequency oscillations. For example, during an asynchronous interconnection experiment at a local power grid, an ultra-low frequency oscillation with a period of approximately 20 seconds, lasted for half an hour. These ultra-low frequency oscillations, unlike traditional low-frequency oscillations, are primarily mechanical power oscillations. Existing measures to suppress ultra-low frequency oscillations primarily target hydropower units, such as adjusting the speed regulator parameters of the units and optimizing speed regulation parameters using deep learning or particle swarm algorithms to increase damping. However, these approaches reduce the primary frequency regulation performance of the hydropower units, increasing the incidence of frequency limit violations under large disturbances.

[0003] During the large-scale integration of wind power, the random fluctuations in wind turbine output further exacerbate grid frequency stability issues. The difference in inertia between wind turbines and traditional synchronous generators further reduces grid inertia, increasing the risk of ultra-low frequency oscillations. Energy storage integration can address the unstable wind turbine output. Wind-storage systems offer rapid dynamic regulation capabilities, and adding appropriate additional damping control on the wind-storage side can suppress ultra-low frequency oscillations. However, existing ultra-low frequency oscillation suppression measures often use frequency as a reference signal, without incorporating the oscillation source signal, resulting in low utilization of ancillary service resources. Summary of the Invention

[0004] In order to solve the above-mentioned technical problems existing in the prior art, for a high-proportion hydropower system with wind-storage access, the present invention aims to provide a method for suppressing ultra-low frequency oscillations by involving a wind-storage system and combining it with oscillation source positioning.

[0005] Specifically, the present invention provides an oscillation source-localized ultra-low frequency oscillation suppression method for suppressing ultra-low frequency oscillations in a wind-water storage system, wherein the wind-water storage system includes a wind turbine generator set connected by a busbar, energy storage, and at least two hydropower generator sets.

[0006] The technical solution specifically includes the following steps:

[0007] Step S1: performing oscillation mode decomposition on each hydropower unit in the wind-water storage system, and decomposing the speed curve data or mechanical power curve data of the corresponding generator;

[0008] Select the ultra-low frequency oscillation mode to calculate the oscillation energy, and determine the unit as the oscillation source based on the oscillation energy;

[0009] Step S2: Using the acceleration power of the oscillation source as the output of the wind water storage system, i.e., the first output; using the wind turbine output and the energy storage output as the input of the wind water storage system; wherein the acceleration power is specifically the acceleration power of each of the two oscillation sources with the largest oscillation energy;

[0010] The transfer function matrix of the Feng Shui storage system is identified, namely the first transfer function matrix;

[0011] Step S3: setting an additional controller on the wind-storage side and taking the acceleration power as input; calculating a transfer function matrix of the additional controller based on hybrid robust control theory and the first transfer function matrix, i.e., a second transfer function matrix;

[0012] The output of the additional controller, i.e., the second output, is obtained by calculation according to the second transfer function matrix;

[0013] The second output is added to the fixed power side of the wind turbine and the fixed power side of the energy storage respectively, the output of the wind turbine is adjusted to the additional power instruction value of the wind turbine, and the output of the energy storage is adjusted to the additional power instruction value of the energy storage, so as to realize ultra-low frequency oscillation suppression.

[0014] Preferably, step S1 specifically includes:

[0015] Step S11: Create a sampling signal Y based on the generator's speed deviation curve or mechanical power curve.

[0016] Y=[x(0),x(1),...x(i),...,x(N-1)] T ;

[0017] Where i is the sampling point number, x(i) is the data value of the i-th sampling point, N is the number of sampling points, i = 0, 1, ..., N-1, T represents transposition;

[0018] Step S12: constructing an augmented Hankel matrix according to the sampled signal Y; performing singular value decomposition on the augmented Hankel matrix to obtain a signal subspace;

[0019] Step S13: Find the K eigenvalues ​​of the optimal matrix of the signal subspace by the least square method to generate the optimal matrix λ.

[0020]

[0021] Where, is the i-th power of the k-th eigenvalue of the optimal matrix, k = 1, 2, ..., K;

[0022] For the oscillation mode, we have

[0023]

[0024] c=[c1 c2 … c k ] T =(λ T λ) -1 λ T Y;

[0025] Where, ω k is the angular frequency of the oscillation mode, T s is the period of the sampling signal Y, σ k is the exponential decay factor of the oscillation mode, c is c k The transposed matrix of the set, c k is the product of the amplitude of the oscillation mode signal and the phase rotation factor;

[0026] Step S14: Select an oscillation mode with an oscillation frequency lower than a preset frequency, i.e., an ultra-low frequency oscillation mode, and calculate the oscillation energy.

[0027]

[0028] ΔE m (k)=∫ΔP mk Δω k dt;

[0029] Where, ΔP mk is the mechanical power signal under the ultra-low frequency oscillation mode, Δω k is the speed deviation signal of the generator in the ultra-low frequency oscillation mode, n is the sampling point number, n=1,2,...,N, ΔE m (k) is the oscillation energy in the ultra-low frequency oscillation mode;

[0030] If ΔE m (k)>0, the corresponding unit is the oscillation source.

[0031] Preferably, the preset frequency is 0.1 Hz.

[0032] Preferably, step S2 specifically includes:

[0033] Step S21: applying a power step disturbance to each of the two oscillation sources with the largest oscillation energy to perform oscillation mode decomposition; performing a Laplace transform on the decomposition result, and dividing it by the Laplace transform result of the power step disturbance to obtain the corresponding transfer function;

[0034] Step S22: combine all transfer functions into a first transfer function matrix,

[0035]

[0036] Where s is the Laplace operator, y1(s) and y2(s) are the outputs of the wind-water storage system, i.e., the acceleration power of the two oscillating sources; u1(s) and u2(s) are the inputs of the wind-water storage system, i.e., the wind turbine additional power command value and the energy storage additional power command value; G fg (s) is u g (s) to y f The transfer function of (s), f∈[1,2], g∈[1,2].

[0037] Preferably, the speed curve data and the mechanical power curve data of the generator are decomposed based on a least squares-rotational invariance technique algorithm.

[0038] Preferably, the transfer function matrix of the Feng Shui storage system is identified based on the least squares-rotational invariance algorithm.

[0039] The technical solution provided by the present invention can select the acceleration power corresponding to different oscillation sources as the feedback signal of additional control on the basis of ensuring the primary frequency regulation performance of the hydropower unit, and introduce the fan output and energy storage output as the input of the hydropower unit. It can fully utilize the mechanical power oscillation characteristics of ultra-low frequency oscillation and the flexible and adjustable performance advantages of fans and energy storage, and improve the suppression effect of ultra-low frequency oscillation on the basis of improving the utilization rate of auxiliary service resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a schematic diagram of the structure of the additional controller on the wind storage side in one embodiment of the present invention.

[0041] Figure 2 The figure is a flow chart of ultra-low frequency oscillation suppression based on oscillation source positioning in the present invention.

[0042] Figure 3 Schematic diagram of the structure of a Feng Shui storage system in one embodiment of the present invention.

[0043] Figure 4 This is an oscillation energy curve diagram of each unit in a hydropower system in one embodiment of the present invention.

[0044] Figure 5 1 is a comparison diagram of frequency curves of wind-hydropower systems under different control methods in one embodiment of the present invention. DETAILED DESCRIPTION

[0045] The present invention designs an additional damping controller based on hybrid robust control. The input signal of the additional controller is the acceleration power signal of the oscillation source unit, and the output signal is added to the constant power side of the wind storage system, so as to realize dynamic adjustment of the wind storage output with the acceleration power of the oscillation source unit, increase the ultra-low frequency damping of the system, and realize the suppression of ultra-low frequency oscillation.

[0046] Hereinafter, the technical solution provided by the present invention will be further elaborated in combination with specific embodiments and drawings.

[0047] Example 1

[0048] This embodiment provides a method for suppressing ultra-low frequency oscillation based on oscillation source positioning. The method involves the wind storage system, such as Figure 1 As shown, the specific steps include:

[0049] Step 1: Perform oscillation mode decomposition using the Total Least Squares-Estimation on Signal Parameters via Rotational Technique (TLS-ESPRIT) algorithm, breaking down the generator speed and mechanical power curves. Select the ultra-low frequency oscillation mode and calculate the oscillation energy. Based on this oscillation energy, determine whether the generator set is the oscillation source.

[0050] The sampling signal Y is established based on the generator speed deviation curve or mechanical power curve.

[0051] Y=[x(0),x(1),...x(i),...,x(N-1)] T ;

[0052] Where i is the sampling point number, x(i) is the data value of the i-th sampling point, N is the number of sampling points, i = 0, 1, ..., N-1, and T represents the transpose. The augmented Hankel matrix is ​​constructed based on the sampled signal Y; the augmented Hankel matrix is ​​subjected to singular value decomposition to obtain the signal subspace and noise subspace. The K eigenvalues ​​of the optimal matrix of the signal subspace found by the TLS (Total Least Squares) method generate the matrix λ.

[0053]

[0054] Where, is the i-th power of the k-th eigenvalue of the optimal matrix, k = 1, 2, ..., K.

[0055] For the oscillation mode, we have

[0056]

[0057] a k =2|c k |;

[0058]

[0059] Where, ω k is the angular frequency of the oscillation mode, Ts is the period of the sampling signal Y, σ k is the exponential decay factor of the oscillation mode, a k is the amplitude of the oscillation mode, c k is the product of the oscillation mode signal amplitude and the phase rotation factor, is the phase angle of the oscillation mode.

[0060] in,

[0061] c=[c1 c2 … c k ] T =(λ T λ) -1 λ T Y.

[0062] Select ultra-low frequency oscillation, that is, the mode with oscillation frequency less than 0.1Hz, and the mechanical power signal ΔP under the ultra-low frequency oscillation mode mk Or generator speed deviation signal Δω k for:

[0063]

[0064] Where n is the sampling point number, n = 1, 2, ..., N;

[0065] The oscillation energy is:

[0066] ΔE m (k)=∫ΔP mk Δω k dt;

[0067] Where, ΔP mk is the mechanical power variation under ultra-low frequency oscillation mode;

[0068] If the oscillation energy is positive, the corresponding unit is the oscillation source.

[0069] Step 2: Select the acceleration power of the two oscillation source signals with the largest oscillation energy as the output of the water-wind storage system, and the wind storage output as the input of the wind-water storage system. Based on the TLS-ESPRIT algorithm, the partial transfer function of the system is identified and the complete transfer function matrix of the multi-input and output system is formed.

[0070] Specifically: the actual power grid is a multi-input multi-output model:

[0071]

[0072] Where s is the Laplace operator, y1(s) and y2(s) are the outputs of the wind-water storage system, i.e., the acceleration power of the two oscillating sources; u1(s) and u2(s) are the inputs of the wind-water storage system, i.e., the wind turbine additional power command value and the energy storage additional power command value.11 (s) is the transfer function of u1(s)-y1(s); G 12 (s) is the u2(s)-y1(s) transfer function; G 21 (s) is the transfer function of u1(s)-y2(s); G 22 (s) is the transfer function of u2(s)-y2(s). G 11 (s), G 12 (s), G 21 (s) and G 22 (s) can be obtained through identification. The identification process is to apply a small power step disturbance only at the fan, and through the oscillation mode decomposition of step one, the oscillation mode decomposition result of the acceleration power of the oscillation source unit G1 is obtained. After the Latent transformation, the Latent transformation of the step disturbance is divided to obtain G 11 (s), similarly, according to the acceleration power decomposition result of the oscillation source unit G2, we can get G 21 (s). Only a small power step disturbance is applied at the energy storage. Through the oscillation mode decomposition of step one, the oscillation mode decomposition result of the acceleration power of the oscillation source unit G1 is obtained. After the Lapse transformation, it is divided by the Lapse transformation of the step disturbance to obtain G 12 (s), similarly, according to the acceleration power decomposition result of the oscillation source unit G2, we can get G 22 (S).

[0073] The above-mentioned selection of two oscillation source signals with the largest oscillation energy as the output of the Feng Shui storage system and the feedback signal of the additional controller is an optimal method. It can reduce the difficulty of calculation while ensuring the accuracy of the model and improve the rate of ultra-low frequency oscillation suppression.

[0074] Step 3: Based on the hybrid robust control theory and the transfer function matrix identified in step 2, the transfer function matrix K(s) of the additional controller on the wind-storage side is calculated. The acceleration power of the oscillation source of the feedback signal of the additional controller is added, and the output of the additional controller is added to the constant power side of the wind storage.

[0075] Specifically: According to the transfer function G obtained by identification 11 (s), G 12 (s), G 21 (s) and G 22 (s), based on the robust control algorithm, the corresponding calculation controller transfer function K 11 (s), K 12 (s), K 21 (s) and K 22 (s), the specific structure of the additional controller on the wind-storage side is as follows Figure 1 As shown. Among them, P1 is the actual output of energy storage, P ref1 is the energy storage output command value, P2 is the actual output of the fan, P ref2is the wind turbine output command value, and the oscillation source acceleration power is input to the additional controller on the wind-storage side and is added to the wind-storage output command value after passing through the controller and the limiter link. The wind turbine and energy storage output are dynamically adjusted with the unbalanced power of the oscillation source, thereby achieving the purpose of enhancing the system damping and suppressing ultra-low frequency oscillation.

[0076] The TLS-ESPRIT algorithm is a preferred algorithm due to its strong noise and interference resistance. Alternatively, other algorithms, such as the prony algorithm, can be used to implement its functionality. PI stands for Proportional-Integral control.

[0077] Example 2

[0078] The process of the ultra-low frequency oscillation suppression method based on oscillation source positioning provided in this embodiment is as follows: Figure 2 Combining this method with Figure 3 As shown, a wind turbine system with a busbar connection, energy storage and at least two hydropower units is simulated and verified on PSCAD (Power Systems Computer Aided Design, electromagnetic transient simulation software). Wind turbine system 1) applies a small power disturbance on the wind turbine side, and uses the TLS-ESPRIT algorithm to perform oscillation mode decomposition, decomposing the generator speed curve and mechanical power curve data. The ultra-low frequency oscillation mode is selected to calculate the oscillation energy. The oscillation energy of each unit is as follows: Figure 4 As shown, generators G1 and G3 have high oscillation energy, representing the oscillation source. Generator G2 has near-zero oscillation energy, providing almost no oscillation energy. Generator G4 has negative oscillation energy, representing a positive damping unit. Ultra-low-frequency oscillations are often accompanied by electromagnetic power oscillations, so the acceleration power of generators G1 and G3 can be selected as feedback signals.

[0079] 2) The acceleration power of the oscillation source generators G1 and G3 is selected as the system output, and the wind storage system output is selected as the system input. The transfer function obtained based on the TLS-ESPRIT algorithm is:

[0080]

[0081] 3) The transfer function of the additional controller on the wind-storage side is calculated as:

[0082]

[0083] The feedback signal of the additional controller is the acceleration power of the oscillation source generators G1 and G3, and the additional controller output is on the wind storage constant power side. When the load is reduced by 2500MW, the effects of different controllers are compared as follows: Figure 5As shown, the robust controller based on an oscillating source signal achieves better suppression than the non-oscillating source signal, demonstrating the effectiveness of this approach. Under the additional robust control based on the oscillating source signal, the system frequency peaks at 50.1 Hz, while the frequency peak based on the non-oscillating source signal is 50.2 Hz, the same as the peak value of traditional PI control. 50.2 Hz is the upper frequency limit for normal operation, and the system frequency under non-oscillating source control is prone to exceeding this limit.

[0084] In summary, the technical solution provided by the present invention can, on the basis of ensuring the primary frequency regulation performance of the hydropower unit, select the acceleration power corresponding to different oscillation sources as the feedback signal of the additional control, design the additional robust damping controller on the wind-storage side, introduce the wind turbine output and energy storage output as the input of the hydropower unit, the controller input signal is the acceleration power signal of the oscillation source, and the output signal is attached to the fixed power side of the wind turbine and energy storage unit. It can fully utilize the mechanical power oscillation characteristics of ultra-low frequency oscillation and the flexible and adjustable performance advantages of wind turbines and energy storage, and improve the suppression effect of ultra-low frequency oscillation on the basis of improving the utilization rate of auxiliary service resources;

[0085] Furthermore, each preferred solution has achieved the following beneficial effects on the basis of achieving the above beneficial effects: processing data through the least squares-rotational invariance technology can more effectively reduce noise interference, making the decomposition result or the established sampling signal more accurate.

Claims

1. A method for suppressing ultra-low frequency oscillation based on oscillation source positioning, for suppressing ultra-low frequency oscillation of a wind-water storage system, wherein the wind-water storage system comprises a wind turbine generator set, energy storage, and at least two hydropower generator sets connected by a busbar; It is characterized in that The following steps are involved: Step S1: performing oscillation mode decomposition on each hydropower unit in the wind-water storage system, and decomposing the speed curve data or mechanical power curve data of the corresponding generator; Select the ultra-low frequency oscillation mode to calculate the oscillation energy, and determine the unit as the oscillation source based on the oscillation energy; Step S2: Using the acceleration power of the oscillation source as the output of the wind water storage system, i.e., the first output; using the wind turbine output and the energy storage output as the input of the wind water storage system; wherein the acceleration power is specifically the acceleration power of each of the two oscillation sources with the largest oscillation energy; The transfer function matrix of the Feng Shui storage system is identified, namely the first transfer function matrix; Step S3: setting an additional controller on the wind-storage side and taking the acceleration power as input; calculating a transfer function matrix of the additional controller based on hybrid robust control theory and the first transfer function matrix, i.e., a second transfer function matrix; The output of the additional controller, i.e., the second output, is obtained by calculation according to the second transfer function matrix; The second output is added to the fixed power side of the wind turbine and the fixed power side of the energy storage respectively, the output of the wind turbine is adjusted to the additional power instruction value of the wind turbine, and the output of the energy storage is adjusted to the additional power instruction value of the energy storage, so as to realize ultra-low frequency oscillation suppression.

2. The method for suppressing ultra-low frequency oscillation based on oscillation source positioning according to claim 1, characterized in that: Step S1 specifically includes: Step S11: Create a sampling signal Y based on the generator's speed deviation curve or mechanical power curve. Y=[x(0),x(1),...x(i),...,x(N-1)] T ; Where i is the sampling point number, x(i) is the data value of the i-th sampling point, N is the number of sampling points, i = 0, 1, ..., N-1, T represents transposition; Step S12: constructing an augmented Hankel matrix according to the sampled signal Y; performing singular value decomposition on the augmented Hankel matrix to obtain a signal subspace; Step S13: Find the K eigenvalues ​​of the optimal matrix of the signal subspace by the least square method to generate the optimal matrix λ. Where, is the i-th power of the k-th eigenvalue of the optimal matrix, k = 1, 2, ..., K; For the oscillation mode, we have c=[c1 c2 … c k ] T =(λ T l) -1 l T Y; Where, ω k is the angular frequency of the oscillation mode, T s is the period of the sampling signal Y, σ k is the exponential decay factor of the oscillation mode, c is c k The transposed matrix of the set, c k is the product of the amplitude of the oscillation mode signal and the phase rotation factor; Step S14: Select an oscillation mode with an oscillation frequency lower than a preset frequency, i.e., an ultra-low frequency oscillation mode, and calculate the oscillation energy. D.E. m (k)=∫ΔP mk Give k dt; Where, ΔP mk is the mechanical power signal under the ultra-low frequency oscillation mode, Δω k is the speed deviation signal of the generator in the ultra-low frequency oscillation mode, n is the sampling point number, n=1,2,...,N, ΔE m (k) is the oscillation energy in the ultra-low frequency oscillation mode; If ΔE m (k)>0, the corresponding unit is the oscillation source.

3. The method for suppressing ultra-low frequency oscillation based on oscillation source positioning according to claim 2, characterized in that: The preset frequency is 0.1 Hz.

4. The method for suppressing ultra-low frequency oscillation based on oscillation source positioning according to claim 1, characterized in that: Step S2 specifically includes: Step S21: applying a power step disturbance to each of the two oscillation sources with the largest oscillation energy to perform oscillation mode decomposition; performing a Laplace transform on the decomposition result, and dividing it by the Laplace transform result of the power step disturbance to obtain the corresponding transfer function; Step S22: combine all transfer functions into a first transfer function matrix, Where s is the Laplace operator, y1(s) and y2(s) are the outputs of the wind-water storage system, i.e., the acceleration power of the two oscillating sources; u1(s) and u2(s) are the inputs of the wind-water storage system, i.e., the additional power command value of the wind turbine and the additional power command value of the energy storage; G fg (s) is u g (s) to y f The transfer function of (s), f∈[1,2], g∈[1,2].

5. The method for suppressing ultra-low frequency oscillation based on oscillation source positioning according to any one of claims 1 to 4, characterized in that: The generator speed curve data and mechanical power curve data are decomposed based on the least squares-rotational invariance technology algorithm.

6. The method for suppressing ultra-low frequency oscillation based on oscillation source positioning according to any one of claims 1 to 4, characterized in that: The transfer function matrix of the Feng Shui storage system is obtained based on the least squares-rotational invariance technique algorithm.

Images