Energy Storage Method for Suppressing Low-Frequency Oscillations Based on Damped and Resonant Controllers

Abstract

This invention discloses a method for suppressing low-frequency oscillations in energy storage based on a damped controller and a resonant controller, specifically relating to the field of low-frequency oscillation suppression. The method involves identifying the type of low-frequency oscillation and then adaptively selecting and online tuning the corresponding damped or resonant controller to generate compensation power. This method constructs a dual-discrimination path for negatively damped oscillations and forced oscillations, effectively identifying the type of low-frequency oscillation and overcoming the deficiency of lacking effective oscillation type identification capability. By adaptively selecting the controller according to the oscillation type, targeted suppression is applied to different types of oscillations, avoiding ineffective adjustment of a single control strategy, reducing resource waste in the energy storage system, improving suppression efficiency, slowing down the lifespan decay of the energy storage device, and curbing the propagation of forced oscillation energy.

Description

Technical Field

[0001] This invention relates to the field of low-frequency oscillation suppression technology, and more specifically, to a method for suppressing low-frequency oscillations by energy storage based on a damping controller and a resonant controller. Background Technology

[0002] With the development of large-scale new energy grid connection and regional power grid interconnection, ensuring the stable operation of the power system, especially suppressing low-frequency oscillations, is crucial. In traditional power systems, damping is usually provided by the power system stabilizer built into the synchronous generator unit. Its working principle is to suppress the relative sway between rotors by adjusting the excitation of the unit. However, the traditional solution has significant technical bottlenecks: the suppression effect of the power system stabilizer will be greatly reduced when the grid structure or operating conditions change; it cannot provide effective suppression of forced power oscillations caused by external periodic disturbance sources, thus leaving potential safety hazards.

[0003] To address the shortcomings of traditional power system stabilizers in terms of poor adaptability and limited suppression range, existing technologies utilize energy storage devices connected to key nodes of the power grid and equipped with additional damping controllers to form an active suppression scheme. By calculating the required damping power based on deviations in signals such as line power, frequency, or phase angle, the energy storage converter is instructed to directly inject or absorb a compensating power that is opposite to the power oscillation into the grid. This effectively suppresses negative damping oscillations in the system, thereby eliminating dependence on specific generator sets, resulting in faster response and more significant suppression effects.

[0004] However, in practical use, it still has some drawbacks. For example, it adopts a single additional damping control strategy and lacks the ability to effectively identify the type of oscillation. When encountering forced oscillations caused by external periodic disturbance sources, the controller cannot apply targeted suppression at the root. This leads to the controller continuously performing ineffective power regulation, resulting in a serious waste of energy storage system resources and a reduction in overall suppression efficiency. Ineffective power throughput accelerates the lifespan degradation of energy storage devices and increases the burden of operation and maintenance. Since it cannot effectively contain the energy of forced oscillations, it can propagate in the power grid, ultimately posing a direct threat to safe and stable operation. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides an energy storage method for suppressing low-frequency oscillations based on a damping controller and a resonant controller, which solves the problems mentioned in the background art through the following scheme.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] Energy storage methods for suppressing low-frequency oscillations based on damped controllers and resonant controllers include:

[0008] S1: Acquire the first oscillation signal of the target power grid area, and preprocess the first oscillation signal to obtain the second oscillation signal;

[0009] S2: Perform pattern recognition on the second oscillation signal to extract the first oscillation feature of the dominant oscillation mode at the target time, wherein the first oscillation feature includes at least the oscillation frequency and the damping ratio;

[0010] S3: Based on the first oscillation feature, determine the oscillation type of the dominant oscillation mode and output the second oscillation feature;

[0011] S4: Based on the second oscillation characteristic, adaptively select and enable the corresponding controller to generate a power suppression command:

[0012] When the second oscillation characteristic is a negative damped oscillation, the damping controller is activated, and the damping controller is tuned online using the first oscillation characteristic to generate a first suppression command for providing positive damping;

[0013] When the second oscillation characteristic is forced oscillation, the resonant controller is activated, and the resonant controller is frequency-locked using the first oscillation characteristic to generate a second suppression command to counteract periodic disturbances.

[0014] S5: Convert the first or second suppression command output by the selected controller into compensation power, and continuously execute S1 to S4 to perform closed-loop feedback adjustment of the suppression effect of low-frequency oscillation.

[0015] Preferably, step S2, extracting the first oscillation feature of the dominant oscillation mode at the target time, specifically includes:

[0016] The oscillation frequency in the first oscillation feature is obtained as follows:

[0017] ,

[0018] in, Represented as at a point in time The second oscillation signal at that time, Represented as the first The oscillation amplitude of each potential oscillation mode. Represented as the first The attenuation factor of each potential oscillation mode. Represented as the first The oscillation frequency of each potential oscillation mode. Represented as the first Phase shift of each potential oscillation mode.

[0019] Preferably, step S2, extracting the first oscillation feature of the dominant oscillation mode at the target time, specifically further includes:

[0020] Based on the The attenuation factor of each potential oscillation mode The first part is calculated by using its real part and imaginary part. Damping ratio of each potential oscillation mode Specifically, it is expressed as:

[0021] .

[0022] Preferably, in step S3, the second oscillation feature includes at least an oscillation type flag bit and its corresponding oscillation frequency value;

[0023] The oscillation type flag is used to trigger the corresponding control path, including the negative damped oscillation discrimination path and the forced oscillation discrimination path;

[0024] When the oscillation type flag is set to forced oscillation, the second oscillation feature includes the matched disturbance source.

[0025] Preferably, the discrimination logic of the negative damped oscillation discrimination path in S3 specifically includes:

[0026] The real-time damping ratio remains below the preset safety threshold.

[0027] The oscillation frequency is within a preset length of Within the time window, the fluctuation range is less than the preset fluctuation threshold, while the oscillation amplitude shows a continuous increasing trend.

[0028] Preferably, the discrimination logic of the forced oscillation discrimination path in S3 specifically includes:

[0029] The oscillation frequency matches the frequencies in the characteristic frequency library of periodic disturbance sources;

[0030] Analyze and determine the strong correlation between the oscillation phase and the operating state of the suspected disturbance source.

[0031] Preferably, step S4, which involves online tuning of the damping controller, specifically includes:

[0032] The gain of the damping controller is adaptively adjusted based on the difference between the damping ratio and the target damping ratio in the first oscillation characteristic.

[0033] Based on the oscillation frequency in the first oscillation characteristic, the lead time constant and lag time constant of the damping controller are adaptively calculated and adjusted to provide optimal phase compensation at the currently dominant oscillation frequency.

[0034] Preferably, step S4, frequency locking of the resonant controller, specifically includes:

[0035] The oscillation frequency in the first oscillation characteristic is directly set as the center frequency of the resonant controller, so that the resonant controller generates a cancellation signal with the same amplitude and opposite phase as the forced oscillation.

[0036] The technical effects and advantages of this invention are as follows:

[0037] 1. This invention achieves refined time-frequency domain analysis of oscillation signals by quantizing and extracting low-frequency oscillation signals, avoiding misjudgments caused by multimodal interference in traditional methods, and ensuring focus on the oscillation modes that pose the greatest threat;

[0038] 2. This invention achieves effective identification of low-frequency oscillation types by constructing a dual discrimination path of negative damped oscillation and forced oscillation, overcoming the deficiency of lacking effective oscillation type identification capability;

[0039] 3. This invention adaptively selects the controller based on the oscillation type, applies targeted suppression to different types of oscillations, avoids ineffective adjustment by a single control strategy, reduces the waste of energy storage system resources, improves suppression efficiency, slows down the lifespan decay of energy storage devices, and curbs the propagation of forced oscillation energy. Attached Figure Description

[0040] Figure 1 This is a flowchart illustrating the steps of an energy storage method for suppressing low-frequency oscillations based on a damping controller and a resonant controller, according to an embodiment of this application. Detailed Implementation

[0041] 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 skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to and includes any or all possible combinations of one or more of the listed items.

[0043] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0044] As attached Figure 1 The energy storage method for suppressing low-frequency oscillations based on damping and resonant controllers, as shown, identifies the type of low-frequency oscillations and then adaptively selects and tunes the corresponding damping or resonant controller online to generate compensation power; specifically, it includes the following steps:

[0045] S1: Acquire the first oscillation signal of the target power grid area, and preprocess the first oscillation signal to obtain the second oscillation signal;

[0046] S2: Perform pattern recognition on the second oscillation signal to extract the first oscillation feature of the dominant oscillation mode at the target time, wherein the first oscillation feature includes at least the oscillation frequency and the damping ratio;

[0047] S3: Based on the first oscillation feature, determine the oscillation type of the dominant oscillation mode and output the second oscillation feature;

[0048] S4: Based on the second oscillation characteristic, adaptively select and enable the corresponding controller to generate a power suppression command:

[0049] When the second oscillation characteristic is a negative damped oscillation, the damping controller is activated, and the damping controller is tuned online using the first oscillation characteristic to generate a first suppression command for providing positive damping;

[0050] When the second oscillation characteristic is forced oscillation, the resonant controller is activated, and the resonant controller is frequency-locked using the first oscillation characteristic to generate a second suppression command to counteract periodic disturbances.

[0051] S5: Convert the first or second suppression command output by the selected controller into compensation power, and continuously execute S1 to S4 to perform closed-loop feedback adjustment of the suppression effect of low-frequency oscillation.

[0052] Specifically, in S1, a first oscillation signal that can characterize the low-frequency oscillation state is acquired in real time by phasor measurement units deployed at key nodes in the target power grid area. The first oscillation signal is preferably the active power signal of the key tie line or the frequency signal of the important bus. A preprocessing process is performed on the acquired first oscillation signal to effectively filter out high-frequency noise, power frequency harmonic interference and abnormal values ​​in the measurement data, thereby eliminating measurement errors. The output of the preprocessing process is the second oscillation signal.

[0053] In this embodiment, regarding the specific implementation of the first oscillation signal acquisition, the selection of key nodes is used to capture and reflect the overall operating status and potential low-frequency oscillation information of the target power grid to the greatest extent. The scope includes, but is not limited to, key tie lines connecting different regional power grids, transmission lines of large power plants, and hub substation busbars of important load centers. Preferably, the active power signal of the tie line and the frequency signal of the hub substation busbar are collected as the first oscillation signal. All deployed phasor measurement units must strictly comply with the IEEE C37.118.1 standard and use the Global Positioning System to achieve high-precision time synchronization to ensure that measurement data from different geographical locations have a unified and accurate time scale.

[0054] Further, the preprocessing procedure for the first oscillation signal is as follows: Low-pass filtering, preferably using a digital low-pass filter in this embodiment, including but not limited to Butterworth and Chebyshev filters, to process the acquired first oscillation signal; the frequency band of the first oscillation signal is limited to the low-frequency oscillation band of 0.1Hz to 2.5Hz by low-pass filtering, so as to deterministically filter out high-frequency noise, power frequency and its harmonic components, and irrelevant signals such as subsynchronous oscillation caused by power electronic equipment, communication interference, etc.; noise reduction processing, preferably using wavelet transform noise reduction or adaptive Kalman filtering in this embodiment to process the filtered signal, so as to distinguish the characteristics of signal and noise at different scales. By setting an appropriate threshold to remove the wavelet coefficients corresponding to noise, random noise can be smoothed to a great extent while the key abrupt change characteristics in the oscillation waveform are completely preserved, thereby significantly improving the signal-to-noise ratio of the signal.

[0055] Furthermore, a verification mechanism is introduced into the preprocessing process of the first oscillation signal to ensure that the output second oscillation signal always meets high-quality standards. The verification mechanism includes: internal quality verification, which performs real-time time-domain and frequency-domain characteristic analysis on the preprocessed second oscillation signal and calculates its key quality indicators. In this embodiment, the key quality indicators include, but are not limited to, signal-to-noise ratio and total harmonic distortion rate. If the key quality indicators are better than the preset quality threshold, the preprocessing is deemed effective. Otherwise, if the key quality indicators fail to meet the standards, an alarm log is automatically generated. Collaborative cross-verification, in this embodiment, is based on the consistency of multiple PMU data for logical judgment. That is, by comparing signals from multiple key nodes that are geographically adjacent or electrically closely connected in real time, when the dynamic characteristics of the key node show significant uncorrelation with the surrounding signals, the dynamic characteristics include, but are not limited to, dominant frequency and phase change trends, it is determined that its node data may have deep-level anomalies, and corresponding fault-tolerant processing measures such as temporarily removing the signal, reducing its weight, or triggering a deep diagnostic procedure are taken.

[0056] Specifically, in S2, the bandwidth of the second oscillation signal output in S1 is limited to the low-frequency oscillation band of 0.1 to 2.5 Hz, and converted into a set of core feature parameters that can quantify the low-frequency oscillation dynamic characteristics of the target power grid area. The second oscillation signal is processed by the identification algorithm to extract the first oscillation feature of the dominant oscillation mode. The key quantitative indicators represented by the first oscillation feature include at least the oscillation frequency and the damping ratio.

[0057] In one possible implementation, pattern recognition of the second oscillation signal includes: truncating a preset length of data through a first-in-first-out (FIFO) data buffer. The current time window data sequence is used as the identification input; a pattern recognition model is established based on the data sequence and the parameters are estimated to obtain a set of potential oscillation modes; the potential oscillation modes are screened and verified, and the one with the highest amplitude is selected as the dominant oscillation mode to extract the first oscillation feature.

[0058] It should be noted that, in this embodiment, the preferred first oscillation feature includes oscillation frequency, damping ratio, and oscillation amplitude; the pattern recognition model preferably employs an adaptive Prony algorithm or a recursive subspace recognition algorithm. The adaptive Prony algorithm dynamically adjusts the model order to fit the second oscillation signal with a combination of decaying sine functions, and directly parses the damping ratio and oscillation frequency, which characterize the oscillation decay characteristics, from the fitted function as the first oscillation feature; the recursive subspace recognition algorithm recursively updates the pattern recognition model and performs eigenvalue decomposition on the identified system matrix to accurately obtain the first oscillation feature, thereby simultaneously identifying multiple dominant oscillation modes.

[0059] Furthermore, the oscillation frequency and amplitude in the first oscillation characteristic are obtained by superimposing multiple decaying sine functions, specifically expressed as follows:

[0060] ,

[0061] in, Represented as at a point in time The second oscillation signal at that time, Represented as the first The oscillation amplitude of each potential oscillation mode. Represented as the first The attenuation factor of each potential oscillation mode. Represented as the first The oscillation frequency of each potential oscillation mode. Represented as the first Phase shift of the potential oscillation mode; based on the first The attenuation factor of each potential oscillation mode The first part is calculated by using its real part and imaginary part. Damping ratio of each potential oscillation mode Specifically, it is expressed as:

[0062] ,

[0063] in, Represented as the first The oscillation frequency of each potential oscillation mode.

[0064] It should be noted that the first Oscillation amplitude of each potential oscillation mode It reflects the initial intensity of the oscillation; No. The attenuation factor of each potential oscillation mode determines whether the oscillation gradually weakens or continues to grow, and when it is determined to be a gradually weakening oscillation, the oscillation amplitude is used as the basis for adaptively tuning the gain of the additional damping controller; Oscillation frequency of each potential oscillation mode This reflects the speed of the oscillation, which is related to the swing frequency of the generator rotor; the... Phase shift of each potential oscillation mode This indicates the initial phase of the oscillation.

[0065] Furthermore, the screening and verification of the potential oscillation modes includes at least: eliminating all oscillation frequencies. Modes not within the 0.1Hz to 2.5Hz range; and oscillation amplitude excluded. Modes below which can be considered computational noise or physically unreasonable; among all valid oscillation modes that passed the screening, based on the oscillation amplitude of each mode. The patterns are sorted, and the one with the largest amplitude is identified as the dominant oscillation pattern that poses the greatest threat.

[0066] Specifically, in S3, based on the first oscillation feature extracted in S2, which includes oscillation frequency, damping ratio, and oscillation amplitude, intelligent diagnosis is performed on the dominant oscillation mode in the target power grid area. This distinguishes between negative damping oscillation caused by the deterioration of its own damping characteristics and forced oscillation caused by energy injected by external periodic disturbance sources. A structured second oscillation feature is then output. The second oscillation feature includes an oscillation type flag and key parameters corresponding to the oscillation type flag. That is, when it is identified as a negative damping oscillation, its oscillation frequency and damping ratio are output, or when it is identified as a forced oscillation, its oscillation frequency and the matched disturbance source identifier are output.

[0067] It should be noted that the step of intelligently identifying the oscillation type is achieved by executing two parallel discrimination paths based on different physical mechanisms. The discrimination paths include: a negative damping oscillation discrimination path for identifying oscillations caused by insufficient self-damping; and a forced oscillation discrimination path for identifying oscillations caused by external periodic disturbance sources.

[0068] In one possible implementation, the second oscillation feature includes at least an oscillation type flag and its corresponding oscillation frequency value. The oscillation type flag is used to trigger a corresponding control path. When the negative damped oscillation discrimination path is triggered, its discrimination logic includes: real-time damping ratio. The frequency remains below a preset safety threshold; in this embodiment, the preferred safety threshold is 0.03; oscillation frequency At the preset length Within the time window, it remains relatively stable, with its fluctuation range being less than the preset fluctuation threshold, while the oscillation amplitude... It shows a trend of continuous growth or remaining at a high level.

[0069] It should be noted that the safety threshold is set based on the minimum requirement for damping ratio in power system stability standards to ensure a minimum positive damping margin. If the damping ratio is lower than the safety threshold, it indicates that the current oscillation mode exhibits negative damping characteristics, posing a risk of instability, and intervention is necessary. The time window length... The setting must include at least the oscillation frequency. The corresponding 3 to 5 complete cycles are used to ensure that the amplitude growth trend and frequency stability can be reliably observed. In this embodiment, for 0.5Hz oscillation, The oscillation frequency is set to at least 6 seconds. The fluctuation threshold is usually set to the oscillation frequency. 1% to 2% is used to distinguish between frequency-stable forced oscillations and negatively damped oscillations that may drift slightly; for oscillation amplitude The determination of a growth trend can be made by calculating a time window. The linear regression slope of the internal oscillation amplitude sequence is determined as follows: if the slope is greater than zero and exceeds the minimum growth sensitivity threshold, it is determined to be continuously growing; if the mean of the amplitude sequence is continuously higher than the baseline amplitude in its static steady state by more than 50%, it is determined to be maintained at a high level; the above multiple criteria must be satisfied simultaneously to form a necessary and sufficient condition for negative damped oscillation.

[0070] In one possible implementation, the second oscillation feature includes at least an oscillation type flag and its corresponding oscillation frequency value. The oscillation type flag is used to trigger a corresponding control path. When the forced oscillation discrimination path is triggered, the second oscillation feature contains a matched disturbance source. Its discrimination logic includes: the oscillation frequency matches the frequency in a pre-stored periodic disturbance source feature frequency library; and the lock-in or strong correlation between the oscillation phase and the operating state of the suspected disturbance source is analyzed and determined.

[0071] It should be noted that the frequency matching is defined as the identified oscillation frequency. The characteristic frequencies of known periodic disturbance sources pre-entered into the database absolute difference between Less than a preset allowable error band The allowable error band The accuracy of PMU measurements, system frequency fluctuations, and mode identification algorithm errors need to be considered, and are typically set to 0.01–0.05 Hz. The disturbance sources may include, but are not limited to, the operating frequency of specific variable frequency loads, periodic impact loads of large industrial equipment, or the shaft torsional vibration frequency of generator sets. The oscillation phase correlation criterion is achieved by analyzing the synchronous phasor measurement signals of multiple key nodes in the target power grid. This is done by comparing the phase of the suspected disturbance source access point with the phase of the oscillation signal at the distant observation point. If both are within the oscillation frequency range... Maintaining a stable phase difference, i.e., phase locking; in this embodiment, it is preferred to quantify the judgment by calculating the coherence coefficient of the two signals near the oscillation frequency point. If the coherence coefficient is consistently higher than 0.9 and the phase difference is stable, it is determined that there is a strong correlation.

[0072] Specifically, in S4, by receiving and parsing the second oscillation characteristics, the suppression technology path is adaptively selected, and the core parameters of the selected controller are configured based on real-time operating conditions to generate instructions for suppressing power.

[0073] It should be noted that the adaptive selection of the suppression technology path and the activation of the corresponding controller are as follows: when the second oscillation characteristic is determined to be a negative damped oscillation, the resonant controller is locked and the damping controller is activated; when the second oscillation characteristic is determined to be a forced oscillation, the damping controller is locked and the resonant controller is activated.

[0074] In one possible implementation, when the second oscillation characteristic is determined to be a negatively damped oscillation, the process of online tuning of the damping controller includes: adaptively adjusting the gain of the damping controller based on the difference between the damping ratio and the target damping ratio in the first oscillation characteristic; the larger the difference in the damping deficiency ratio, i.e. the higher the degree of deficiency, the larger the gain is set, so that the controller can provide stronger damping torque; and adaptively calculating and adjusting the lead time constant and lag time constant of the damping controller based on the oscillation frequency in the first oscillation characteristic, so as to provide optimal phase compensation at the currently dominant oscillation frequency, so that the phase of the suppression signal finally output by the controller is precisely out of phase with the oscillation mode, thereby producing the maximum positive damping effect.

[0075] It should be noted that the target damping ratio can be preset according to the power grid safety operation regulations, and is usually a fixed value greater than zero; the preferred target damping ratio in this embodiment is 0.03 or 0.05; the adaptive adjustment process relies on a built-in preset tuning rule library, which contains parameter mapping relationships established based on phase compensation method, pole placement method, or a large amount of offline simulation and online learning data; the gain adjustment is used to provide a suppression strength that matches the degree of damping deficiency, while the adjustment of the lead and lag time constants ensures that at the current dominant oscillation frequency point, The controller provides precise phase compensation. In this embodiment, the preferred objective is to make the output suppression signal out of phase with the oscillation mode signal, thereby synthesizing the maximum positive damping torque. The damping controller preferably adopts a lead-lag compensation network structure with a transfer function. Preferably, the tuning process is a continuous dynamic optimization process. After the damping controller is put into operation, it continues to monitor the change in the damping ratio of the oscillation mode to form a closed-loop feedback. If the damping ratio is improved and tends to stabilize, the gain can be gradually reduced to avoid over-adjustment. If the damping effect does not meet expectations, the time constant can be further optimized or the gain can be appropriately increased.

[0076] Furthermore, the transfer function Specifically, it is expressed as:

[0077] ,

[0078] in, This is represented as real-time gain, which is adaptively adjusted by the tuning rule base based on the difference between the current damping ratio and the target damping ratio. Represented as a lead time constant, it is calculated online based on the oscillation frequency in the first oscillation characteristic to ensure that... The controller provides the required lead phase so that the controller output is precisely out of phase with the oscillation mode. It is expressed as the lag time constant and the bandwidth of the fixed-phase compensation.

[0079] In one possible implementation, when the second oscillation feature is determined to be a forced oscillation, the process of locking the frequency of the resonant controller includes: directly setting the oscillation frequency in the first oscillation feature to the center frequency of the resonant controller, so that the resonant controller generates a cancellation signal with the same amplitude and opposite phase as the forced oscillation.

[0080] It should be noted that the resonant controller adopts a structure based on a second-order generalized integrator or a variant thereof to achieve zero steady-state error tracking and generation of sinusoidal components at a specific frequency; the term "direct setting" means that the oscillation frequency... With the center frequency of the resonant controller There exist instantaneous, deterministic mathematical relationships, that is... Its effectiveness depends entirely on the accuracy of the oscillation frequency. By locking the frequency, the controller accurately captures the frequency components of the forced oscillation. Through its internal closed-loop control, it automatically adjusts the amplitude and phase of its output signal. The goal is to match the output active power command signal with the power fluctuation component generated by the detected forced oscillation source at the grid connection point in amplitude and with a phase difference of 180 degrees, thereby suppressing the forced oscillation at its source. Preferably, the output amplitude of the resonant controller is not fixed. In this embodiment, an amplitude control loop is configured, whose initial reference value is set based on the amplitude of the identified oscillation mode. During operation, it is fine-tuned according to the attenuation of the monitored oscillation amplitude to achieve optimal cancellation and avoid overcompensation.

[0081] Specifically, in S5, the power suppression command is converted into an actual power command that the energy storage converter can execute, and a tight closed-loop feedback is formed by real-time monitoring of the suppression effect, thereby dynamically adjusting the control state to achieve low-frequency oscillation suppression.

[0082] Furthermore, the process of converting the power suppression command into compensation power includes: dynamically scaling the suppression command generated by the damping controller or resonant controller based on the rated power capacity of the energy storage and the current operating status parameters to generate an active power reference increment within a safe range; superimposing the active power reference increment with the power setpoint used to perform other grid services, and forming the final total active power reference value through a priority coordination algorithm; and sending the total active power reference value to the control layer of the energy storage converter to drive the converter to generate the corresponding compensation power.

[0083] It should be noted that the dynamic scaling conversion is used to linearly or nonlinearly map the dimensionless commands output by the controller to the current maximum allowable charge and discharge power range. The conversion coefficient is a linear function represented by the current operating status parameters. The current operating status parameters include, but are not limited to, current available capacity, health status, and temperature parameters. This ensures that command execution is always within the safe operating boundary, fundamentally avoiding equipment overload and accelerated aging caused by overpower commands. The priority coordination algorithm is used to arbitrate potential conflicts between oscillation suppression commands and other grid service commands. These other grid service commands include, but are not limited to, frequency regulation and peak shaving. By assigning the highest priority to oscillation suppression, it is ensured that when low-frequency oscillations are detected, the suppression power demand can be prioritized and fully met. The power commands of other grid service commands are then allocated within the remaining capacity after the oscillation suppression demand is met.

[0084] Furthermore, the closed-loop feedback adjustment includes: continuously monitoring the real-time amplitude of the dominant oscillation mode; comparing the real-time amplitude with a first preset threshold; when the real-time amplitude is suppressed to below the first preset threshold, determining the type of controller currently in use; if the currently in use is a damping controller, gradually reducing the gain of the damping controller until it is completely deactivated; if the currently in use is a resonant controller, directly blocking the output of the resonant controller to immediately stop its operation; continuously monitoring the real-time amplitude of the dominant oscillation mode; comparing the monitored real-time amplitude with a second preset threshold; when the real-time amplitude exceeds the second preset threshold again, restarting the closed-loop feedback adjustment process; otherwise, continuing to continuously monitor the real-time amplitude of the dominant oscillation mode.

[0085] It should be noted that the first preset threshold is a safety threshold value for oscillation suppression, which is set based on the maximum oscillation amplitude allowed for stable operation of the target power grid; the second preset threshold is equal to or slightly higher than the first preset threshold to prevent frequent switching of the controller when the measurement signal fluctuates near the threshold, and to further reduce the number of invalid actions of energy storage; the gradual reduction of the damping controller gain is to avoid secondary impact caused by the sudden exit of the controller; and the direct-locked resonant controller does not require a gradual exit process because the cancellation of specific frequency disturbances has been completed.

[0086] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0087] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for suppressing low-frequency oscillations by energy storage based on a damping controller and a resonant controller, characterized in that, include: S1: Collect the first oscillation signal of the target power grid area and preprocess the first oscillation signal to obtain the second oscillation signal; wherein, the preprocessing includes logical judgment based on the consistency of multiple PMU data, that is, by comparing signals from multiple key nodes that are geographically adjacent or electrically closely connected in real time, when there is an anomaly in the key node, take measures such as temporarily removing the signal, reducing its weight, or triggering a deep diagnostic program. S2: Perform pattern recognition on the second oscillation signal to extract the first oscillation feature of the dominant oscillation mode at the target time, wherein the first oscillation feature includes at least the oscillation frequency and the damping ratio; S3: Based on the first oscillation feature, determine the oscillation type of the dominant oscillation mode and output a second oscillation feature; the second oscillation feature includes at least an oscillation type flag and its corresponding oscillation frequency value, the oscillation type flag is used to trigger the corresponding control path, including a negative damped oscillation discrimination path and a forced oscillation discrimination path; the discrimination logic of the forced oscillation discrimination path includes: The oscillation frequency matches the frequencies in the characteristic frequency library of periodic disturbance sources; The strong correlation between the oscillation phase and the operating status of the suspected disturbance source is analyzed and judged. This is achieved by analyzing the synchronous phasor measurement signal in the target power grid. The judgment is made by comparing the phase of the suspected disturbance source access point with the phase of the oscillation signal at the distant observation point and by calculating the coherence coefficient of the two signals near the oscillation frequency point. S4: Based on the second oscillation characteristic, adaptively select and enable the corresponding controller to generate a power suppression command: When the second oscillation characteristic is a negative damped oscillation, the damping controller is activated, and the damping controller is tuned online using the first oscillation characteristic to generate a first suppression command for providing positive damping; When the second oscillation characteristic is forced oscillation, the resonant controller is activated, and the resonant controller is frequency-locked using the first oscillation characteristic to generate a second suppression command to counteract periodic disturbances. S5: Convert the first or second suppression command output by the selected controller into compensation power, and continuously execute S1 to S4 to perform closed-loop feedback adjustment on the suppression effect of low-frequency oscillation; wherein, when the oscillation is suppressed, the closed-loop feedback adjustment includes: if a damping controller is engaged, gradually reduce the gain of the damping controller until it is completely deactivated; if a resonant controller is currently engaged, directly block the output of the resonant controller to make it stop working immediately.

2. The method for suppressing low-frequency oscillations based on a damping controller and a resonant controller according to claim 1, characterized in that: S2 extracts the first oscillation feature of the dominant oscillation mode at the target time, specifically including: The oscillation frequency in the first oscillation feature is obtained as follows: in, Represented as at a point in time The second oscillation signal at that time, Represented as the first The oscillation amplitude of each potential oscillation mode. Represented as the first The attenuation factor of each potential oscillation mode. Represented as the first The oscillation frequency of each potential oscillation mode. Represented as the first Phase shift of each potential oscillation mode.

3. The method for suppressing low-frequency oscillations based on a damping controller and a resonant controller according to claim 2, characterized in that: S2, extracting the first oscillation feature of the dominant oscillation mode at the target time, specifically includes: Based on the The attenuation factor of each potential oscillation mode The first part is calculated by using its real part and imaginary part. Damping ratio of each potential oscillation mode Specifically, it is expressed as: 。 4. The method for suppressing low-frequency oscillations based on a damping controller and a resonant controller according to claim 1, characterized in that: In S3, when the oscillation type flag is forced oscillation, the second oscillation feature includes the matched disturbance source.

5. The method for suppressing low-frequency oscillations based on a damping controller and a resonant controller according to claim 1, characterized in that: The discrimination logic of the negative damped oscillation discrimination path, S3, specifically includes: The real-time damping ratio remains below the preset safety threshold. The oscillation frequency is within a preset length of Within the time window, the fluctuation range is less than the preset fluctuation threshold, while the oscillation amplitude shows a continuous increasing trend.

6. The method for suppressing low-frequency oscillations based on a damping controller and a resonant controller according to claim 1, characterized in that: S4 involves online tuning of the damping controller, specifically including: The gain of the damping controller is adaptively adjusted based on the difference between the damping ratio and the target damping ratio in the first oscillation characteristic. Based on the oscillation frequency in the first oscillation characteristic, the lead time constant and lag time constant of the damping controller are adaptively calculated and adjusted to provide optimal phase compensation at the currently dominant oscillation frequency.

7. The method for suppressing low-frequency oscillations based on a damping controller and a resonant controller according to claim 1, characterized in that: S4, which involves frequency locking of the resonant controller, specifically includes: The oscillation frequency in the first oscillation characteristic is directly set as the center frequency of the resonant controller, so that the resonant controller generates a cancellation signal with the same amplitude but opposite phase as the forced oscillation.

Images