Network-configuration type energy storage control method and system based on adaptive adjustment

Abstract

This invention discloses a grid-based energy storage control method and system based on adaptive regulation, belonging to the field of energy storage control technology. The method includes: acquiring wideband electrical quantities of three-phase voltage and current at the grid connection point of the energy storage converter, and extracting wideband oscillation feature vectors; combining a wideband stability domain model to calculate the stable and safe range of control parameters online; and using a hierarchical collaborative controller containing multi-band regulation units to adaptively regulate and constrain parameters in different segments, and summarizing the parameters according to priority and inputting them into a dual closed-loop controller to achieve adaptive suppression of wideband oscillations. This invention solves the technical problems in existing technologies where insufficient mid-to-high frequency stability margin and lack of wideband collaborative regulation capabilities lead to poor wideband oscillation suppression effects and an inability to adapt to complex wideband power grid scenarios.

Description

Technical Field

[0001] This invention relates to the field of energy storage control technology, and in particular to a grid-type energy storage control method and system based on adaptive adjustment. Background Technology

[0002] With the deepening construction of a new power system based on new energy sources, grid-connected energy storage technology is regarded as a key means to improve the stability of systems with a high proportion of new energy sources due to its ability to provide inertia and voltage support for the power grid. At present, grid-connected energy storage control based on adaptive regulation has become a research hotspot in this field. By dynamically adjusting the output characteristics of energy storage converters (such as virtual inertia, damping coefficient, and virtual impedance), grid-connected energy storage has achieved significant results in improving the dynamic stability of the system in the low-frequency range (such as 0.1Hz to 2Hz), and is gradually becoming an important evolution direction for grid-connected energy storage control under weak power grids.

[0003] In existing technologies, adaptive adjustment-based grid-type energy storage control methods have demonstrated initial application value. Their dynamic parameter adaptation capability enables converters to maintain good power synchronization characteristics under different grid intensities, effectively suppressing low-frequency oscillations and improving the small-signal stability of the system. However, in the process of practical engineering implementation and promotion in broadband grid scenarios, the core technical issues surrounding stability in the mid-to-high frequency band (such as 10Hz to 1000Hz) and broadband oscillation suppression have not been effectively resolved, severely limiting the large-scale application of this technology and its adaptability to the entire grid frequency band.

[0004] On the one hand, existing adaptive regulation methods have inherent limitations in stability margin control at mid-to-high frequencies. The output impedance of traditional grid-based control (such as virtual synchronous machine control) exhibits alternating inductive and resistive transitions at mid-to-high frequencies, resulting in complex impedance amplitude and phase characteristics. Furthermore, most existing adaptive regulation strategies focus on parameter optimization at low-frequency electromechanical timescales, lacking the ability to actively sense and suppress resonant peaks in the tens to hundreds of hertz range. When grid-based energy storage is connected to a power grid containing multiple grid-connected converters or weak transmission lines, it is prone to series and parallel resonances with line impedance, filters, passive components, and other components at mid-to-high frequencies, leading to continuous broadband oscillations. In existing methods, a few studies have attempted to introduce virtual resistors or additional filters to suppress mid-to-high frequency oscillations, but these measures are mostly designed with fixed parameters, making it difficult to adapt to dynamic changes in grid topology and operating conditions. They often come at the cost of sacrificing dynamic response speed, resulting in mismatched suppression effects or control lag. Therefore, current adaptive regulation-based grid-based energy storage control methods suffer from poor broadband oscillation suppression due to insufficient stability margin at mid-to-high frequencies.

[0005] On the other hand, existing adaptive regulation frameworks lack integrated sensing and coordinated suppression capabilities for wide-band oscillation characteristics. Oscillations in real-world power grids often simultaneously contain multi-mode components such as low-frequency electromechanical oscillations, subsynchronous oscillations, and mid-to-high-frequency resonances. Existing methods often treat low-frequency adaptive parameter adjustment and mid-to-high-frequency filtering or impedance reshaping as independent modules, lacking a unified wide-band information extraction and parameter coordinated adjustment mechanism. This fragmented design can lead to control conflicts in the dynamic behavior of converters at different frequency bands. For example, low-frequency adaptive regulation may enhance the inertial response but inadvertently worsen the impedance characteristics at mid-to-high frequencies, inducing new oscillation risks. Furthermore, existing methods generally assume that the grid impedance characteristics change slowly and are only inductive. However, in actual distribution networks or microgrids with complex feeders and multiple conversion devices, the grid impedance exhibits nonlinear and time-varying characteristics over a wide frequency range, making it difficult for existing adaptive methods to achieve real-time tracking and dynamic matching across the entire frequency range's stable region. Therefore, existing adaptive regulation-based grid-type energy storage control methods suffer from a lack of wide-band coordinated sensing and regulation capabilities, resulting in the difficulty of simultaneously and effectively suppressing oscillations at different frequency bands.

[0006] In summary, how to provide a grid-based energy storage control method that can achieve wideband oscillation characteristic sensing and collaborative adaptive suppression, in order to solve the problem of poor wideband oscillation suppression effect caused by insufficient mid-to-high frequency stability margin and lack of wideband collaborative mechanism in the existing technology, has become a key technical problem that urgently needs to be solved in this field. Summary of the Invention

[0007] In view of this, the purpose of the present invention is to provide a grid-type energy storage control method and system based on adaptive adjustment, which aims to solve the problem of poor accuracy of grid-type energy storage control based on adaptive adjustment in the prior art.

[0008] This invention proposes a grid-type energy storage control method based on adaptive adjustment, the method comprising: Obtain the broadband electrical quantity signal at the grid connection point of the grid-connected energy storage converter. The broadband electrical quantity signal includes three-phase voltage and three-phase current. Extract broadband oscillation features from the broadband electrical quantity signal to obtain a broadband oscillation feature vector. Based on the broadband oscillation characteristic vector and combined with the pre-established broadband stability domain model, the safe range of parameters required to ensure stability of virtual inertia, damping coefficient and virtual impedance transfer function in the virtual synchronous machine control at the current operating point is calculated online. A hierarchical and collaborative adaptive controller is constructed, which includes adaptive adjustment units for different frequency bands. Based on the frequency band where the current oscillation occurs, the corresponding adaptive adjustment unit is activated to perform corresponding adjustments, and the output parameters of each adjustment unit are limited to the parameter safety range. The final virtual inertia, damping coefficient, and virtual impedance transfer function output by each regulation unit are summarized according to a preset priority and then output to the voltage and current dual closed-loop controller of the grid-type energy storage converter to complete the adaptive suppression of wideband oscillation.

[0009] Furthermore, in the above-mentioned adaptive adjustment-based grid-type energy storage control method, the step of extracting broadband oscillation features from the broadband electrical quantity signal to obtain a broadband oscillation feature vector includes: A set of adaptive bandpass filter banks with dynamically shiftable center frequency is used to perform parallel filtering on wideband electrical signals, and the signals are divided into five analysis subbands according to low frequency band, subsynchronous frequency band, fundamental frequency neighborhood, mid frequency band and high frequency band. For each sub-band, the oscillation amplitude, oscillation frequency, damping ratio and phase difference of that frequency band are calculated in real time using a recursive least squares identification algorithm, and a wideband oscillation feature vector is output. The broadband oscillation eigenvector is represented as:

[0010] in, Representing the first The oscillation amplitude, oscillation frequency, damping ratio, and phase difference of each sub-band.

[0011] Furthermore, in the aforementioned adaptive adjustment-based grid-type energy storage control method, after the step of activating the corresponding adaptive adjustment unit to perform corresponding adjustments based on the frequency band where the current oscillation occurs, the method further includes: After each adjustment cycle, calculate the change in damping ratio of all sub-bands before and after the adjustment; If the damping ratio of all sub-bands does not decrease and no new oscillation mode appears, then lock the current parameters; If the damping ratio of any sub-band decreases or a new oscillation characteristic appears, the execution parameters are rolled back to the previous effective parameter set, and the participation weight of the corresponding adjustment unit is reduced.

[0012] Furthermore, in the above-mentioned adaptive adjustment-based grid-type energy storage control method, the expression for the parameter safety range is:

[0013] in, For virtual inertia, The damping coefficient is... This is the virtual impedance transfer function. For the first Damping ratio of the belt, This is the preset minimum allowable damping ratio; The expression for the virtual impedance transfer function is:

[0014] in, For virtual resistance, For virtual inductance, For the differential operator, the third term represents multiple dynamically configurable notch / peak compensation units. For the center frequency, To compensate for the gain, The quality factor is used to optimize the configuration in real time based on the stability boundary of the mid-frequency band. This represents the total number of compensation units currently in use by the system. For the first Branch number of each independent notch / peak compensation unit.

[0015] Furthermore, in the above-mentioned adaptive adjustment-based grid-type energy storage control method, the step of activating the corresponding adaptive adjustment unit to perform corresponding adjustments according to the frequency band where the current oscillation occurs includes: If the oscillation is in the low-frequency band or subsynchronous frequency band, adjust the virtual inertia and damping coefficient within the safe range of parameters; If it is a mid-frequency oscillation, the amplitude-frequency characteristics of the virtual impedance are adjusted within the parameter safety range. Specifically, notch compensation units are dynamically configured in the virtual impedance to compensate for the resonant peak value in reverse. If it is a high-frequency oscillation, then increase the virtual resistance and reduce the current loop control bandwidth within the parameter range; Among them, the adjustment of virtual inertia and damping coefficient adopts incremental fuzzy PID control law, so that the damping ratio in the low frequency band or subsynchronous frequency band rises back to the upper limit of the parameter safety range. The virtual resistance automatically increases to a preset upper limit value in the high-frequency band, and the current loop control bandwidth decreases from the first preset value to the second preset value.

[0016] Furthermore, in the above-mentioned adaptive adjustment-based grid-type energy storage control method, the step of rolling back the execution parameters to the previous valid parameter set includes: Construct a circular parameter sequence buffer of length L to store the set of controller parameters and their corresponding damping ratio satisfaction function values ​​that are within the parameter safety range in the past L effective control cycles. When a rollback is triggered, the following strategy is executed: Level 1 rollback: Roll back to the original saved set of valid parameters from the previous cycle; if the number of rollbacks exceeds a preset value within a preset number of consecutive control cycles, it is determined that an unrecoverable continuous conflict state has been entered. Secondary rollback: If a rollback is triggered again after the primary rollback, the circular parameter sequence buffer is traversed, and the set of parameters that form a weak response relationship with the current oscillation sub-band, i.e., the set with the smallest absolute value of the damping ratio sensitivity factor, is selected as the new target rollback point. The feedforward compensation channel is forcibly activated, and the current loop control bandwidth is simultaneously reduced. Level 3 safety rollback: If the rollback is triggered multiple times after the Level 2 rollback strategy is enabled, the rollback will be reverted to the pre-specified absolute safety parameters, and a device-level shutdown protection request will be issued and reported to the power grid dispatch. After the rollback control takes effect, set the satisfaction enhancement boundary function for restoring the adjustment weights of each conflicting sub-band. Only when satisfy The first set weight is restored only when the value is greater than or equal to the preset threshold; otherwise, the second set weight is maintained.

[0017] Furthermore, the above-mentioned adaptive adjustment-based grid-type energy storage control method further includes, after the step of calculating online the parameter safety range required to ensure stability of virtual inertia, damping coefficient, and virtual impedance transfer function in the virtual synchronous machine control over a wide frequency range at the current operating point, the following: A broadband discrete closed-loop control state-space model of a grid-type energy storage system is established and constructed as an augmented state-space matrix. Two consecutive first-order eigenvalue perturbation analyses are performed on the augmented state-space matrix to solve the perturbation sensitivity factor of the damping ratio of the current controller parameters along the coordinate axis and along a specific Jacobian descent direction. After normalizing each sensitivity factor, construct the parameter-damping ratio Jacobian matrix at the current operating point, and calculate the projection of the current parameter adjustment direction onto this matrix to generate the feedforward control quantity in a directional manner. The feedforward control quantity is directly superimposed on the output of each hierarchical control unit, so that before the feedback loop in each control unit responds, the feedforward component actively adjusts the parameters to counteract the foreseeable trend of decreasing broadband damping ratio.

[0018] Another object of the present invention is to provide a grid-type energy storage control system based on adaptive adjustment, the system comprising: The acquisition module is used to acquire the broadband electrical quantity signal at the grid connection point of the grid-connected energy storage converter. The broadband electrical quantity signal includes three-phase voltage and three-phase current. The broadband oscillation feature is extracted from the broadband electrical quantity signal to obtain the broadband oscillation feature vector. A module is established to calculate online the parameter safety range required to ensure stability of virtual inertia, damping coefficient and virtual impedance transfer function in virtual synchronous machine control at the current operating point within a wide frequency range, based on the wide frequency oscillation characteristic vector and a pre-established wide frequency stability domain model. The startup module is used to build a hierarchical and collaborative adaptive controller. The adaptive controller includes adaptive adjustment units for different frequency bands. Based on the frequency band where the current oscillation occurs, the corresponding adaptive adjustment unit is started to perform corresponding adjustments, and the output parameters of each adjustment unit are limited to the parameter safety range. The control module is used to summarize the final virtual inertia, damping coefficient and virtual impedance transfer function output by each regulation unit according to the preset priority and output them to the voltage and current dual closed-loop controller of the grid-type energy storage converter to complete the adaptive suppression of wideband oscillation.

[0019] Another object of the present invention is to provide a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0020] Another object of the present invention is to provide an electronic device including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the program to implement the steps of the method described above.

[0021] This invention acquires wideband electrical signals of three-phase voltage and three-phase current at the grid connection point of a grid-connected energy storage converter, and extracts wideband oscillation features from these signals to obtain wideband oscillation feature vectors. This allows for comprehensive capture of multi-band oscillation mode characteristics of the power grid, including low-frequency, subsynchronous, and medium-to-high-frequency bands, achieving accurate perception of wideband oscillation information across the entire domain. By combining a pre-established wideband stability domain model, the invention calculates the parameter safety ranges of virtual inertia, damping coefficient, and virtual impedance transfer function in virtual synchronous machine control in real time online. This adapts to the grid topology, operating state, and nonlinear time-varying characteristics of wideband impedance, defining stability boundaries for adaptive parameter adjustment. By constructing a hierarchical collaborative adaptive controller containing adaptive adjustment units for different frequency bands, the invention activates corresponding adjustment units based on the actual oscillation frequency band and constrains the output parameters of each unit within the safety range. This enables precise control of oscillations in each frequency band while avoiding instability caused by parameter over-limit adjustments. Finally, by summarizing the control parameters output by each unit through preset priorities and inputting them into a voltage and current dual closed-loop controller, the invention achieves coordinated matching and unified output of multi-band control parameters. On the one hand, it breaks through the limitations of existing technologies that only focus on low-frequency parameter optimization and cannot actively sense and suppress mid-to-high frequency resonance peaks. It abandons the traditional fixed parameter control method and can adapt the control parameters in real time with the dynamic changes of the power grid, effectively improving the stability margin of the mid-to-high frequency band and eliminating the wideband oscillation problem caused by mid-to-high frequency series and parallel resonance. At the same time, it avoids the drawbacks of poor adaptability, control lag, and sacrifice of dynamic response speed of fixed parameter control. On the other hand, it solves the defects of existing technologies where the control modules of each frequency band are isolated and there is no coordinated adjustment mechanism. It eliminates the problem of conflict between high and low frequency band control strategies and the risk of inducing new oscillations. It can track the changes in the stability region of the entire frequency band of the power grid in real time and realize the integrated coordinated suppression of multi-mode oscillations. Ultimately, it effectively improves the grid full-band adaptability of grid-based energy storage converters and solves the technical problems of poor wideband oscillation suppression effect and inability to adapt to complex wideband power grid scenarios due to insufficient mid-to-high frequency stability margin and lack of wideband coordinated control capability in existing technologies. Attached Figure Description

[0022] Figure 1 This is a flowchart of the adaptive adjustment-based grid-type energy storage control method in the first embodiment of the present invention; Figure 2 This is a structural block diagram of a grid-type energy storage control system based on adaptive adjustment in the third embodiment of the present invention.

[0023] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0024] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0025] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0026] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0027] Example 1 Please see Figure 1 The figure shows a grid-type energy storage control method based on adaptive adjustment in the first embodiment of the present invention, the method including steps S10 to S13.

[0028] Step S10: Obtain the broadband electrical quantity signal at the grid connection point of the grid-connected energy storage converter. The broadband electrical quantity signal includes three-phase voltage and three-phase current. Perform broadband oscillation feature extraction on the broadband electrical quantity signal to obtain the broadband oscillation feature vector.

[0029] In specific implementation, a high-precision broadband sampling sensor is installed at the grid-connected port of the grid-connected energy storage converter. The sensor sampling frequency is set to no less than 10kHz, covering a wide range of grid signals from DC to several kilohertz, continuously acquiring the three-phase voltage and current signals at the grid connection point in real time. The acquired raw electrical signals are first processed by pre-processing anti-aliasing filtering to remove high-frequency noise interference and ensure the authenticity and validity of the raw signals. Subsequently, full-band oscillation feature extraction is performed on the pre-processed broadband electrical signals to mine the oscillation mode information of each frequency band contained in the signal. Finally, a broadband oscillation feature vector that can comprehensively characterize the current grid oscillation state is generated. This feature vector contains core characteristic parameters such as amplitude, frequency, damping, and phase of each frequency band oscillation, providing data support for subsequent parameter calculation and adjustment.

[0030] For example, the step of extracting wideband oscillation features from wideband electrical quantity signals to obtain wideband oscillation feature vectors includes: A set of adaptive bandpass filter banks with dynamically shiftable center frequency is used to perform parallel filtering on wideband electrical signals, and the signals are divided into five analysis subbands according to low frequency band, subsynchronous frequency band, fundamental frequency neighborhood, mid frequency band and high frequency band. For each sub-band, the oscillation amplitude, oscillation frequency, damping ratio and phase difference of that frequency band are calculated in real time using a recursive least squares identification algorithm, and a wideband oscillation feature vector is output. The broadband oscillation eigenvector is represented as:

[0031] in, Representing the first The oscillation amplitude, oscillation frequency, damping ratio, and phase difference of each sub-band.

[0032] In this embodiment, an adaptive bandpass filter bank with dynamically shiftable center frequency is selected. This filter bank contains five independent bandpass filter channels, each of which supports real-time fine-tuning of the center frequency to adapt to the dynamic offset characteristics of the power grid oscillation frequency. The pre-processed three-phase voltage and three-phase current broadband electrical quantities are simultaneously input into the five filter channels for parallel synchronous filtering. According to the industry-standard division of power grid oscillation frequency bands, the filter channels are divided into five non-overlapping analysis sub-bands: 0.1Hz ≤ low frequency band < 5Hz, 5Hz ≤ sub-synchronous frequency band < 50Hz, 50Hz ≤ fundamental frequency neighborhood < 60Hz, 60Hz ≤ mid-frequency band < 300Hz, and 300Hz ≤ high frequency band. The fundamental frequency neighborhood is a specific frequency range or band surrounding the system's rated fundamental frequency (e.g., 50Hz). The parallel filtering mode can simultaneously complete the separation and extraction of signals across the entire frequency band, avoiding signal delay caused by serial filtering, ensuring the real-time performance of oscillation feature extraction, and ultimately outputting independent frequency band signals corresponding to the five sub-bands.

[0033] For the frequency band signals output by the five analysis sub-bands, a recursive least squares identification algorithm is used to identify parameters in real time. This algorithm has the advantages of fast convergence speed, strong anti-interference ability, and suitability for online real-time identification, and can adapt to the real-time solution of oscillation characteristics under dynamic power grid conditions. Within each control cycle, the algorithm iteratively calculates the corresponding sub-band signal one by one, accurately solving for four core oscillation characteristic parameters of the current frequency band: oscillation amplitude, oscillation frequency, system damping ratio, and signal phase difference, thus fully characterizing the oscillation characteristics of a single sub-band. After completing the parameter identification of all five sub-bands, the characteristic parameters of all sub-bands are summarized and integrated to generate the final broadband oscillation characteristic vector.

[0034] Step S11: Based on the wideband oscillation characteristic vector and combined with the pre-established wideband stability domain model, calculate online the parameter safety range required to ensure stability of virtual inertia, damping coefficient and virtual impedance transfer function in the virtual synchronous machine control at the current operating point within a wideband range.

[0035] During real-time system operation, the extracted wideband oscillation feature vector is input into a pre-built wideband stability domain model. Combined with real-time operating conditions such as the active power output, reactive power output, grid load level, and grid impedance of the current energy storage system, the model iteratively calculates the three core control parameters—virtual inertia, damping coefficient, and virtual impedance transfer function—in the virtual synchronous machine control strategy under the current operating conditions. This ensures a safe parameter range for stable system operation across the entire wideband. All subsequent parameter adjustments must be limited to this range, thus preventing system instability caused by excessive parameter adjustments.

[0036] For example, the expression for the parameter safety interval is:

[0037] in, For virtual inertia, The damping coefficient is... This is the virtual impedance transfer function. For the first Damping ratio of the belt, This is the preset minimum allowable damping ratio; The expression for the virtual impedance transfer function is:

[0038] in, For virtual resistance, For virtual inductance, For the differential operator, the third term represents multiple dynamically configurable notch / peak compensation units. For the center frequency, To compensate for the gain, The quality factor is used to optimize the configuration in real time based on the stability boundary of the mid-frequency band. This represents the total number of compensation units currently in use by the system. For the first Branch number of each independent notch / peak compensation unit.

[0039] In this embodiment of the invention, the parameter safety interval is limited by a multi-dimensional damping constraint formula. The core constraint logic is that the system damping ratio of all frequency bands is not lower than the preset minimum allowable damping ratio, thereby ensuring stability across the entire frequency band. The constraint relationship of the parameter safety interval uses virtual inertia, damping coefficient, and virtual impedance transfer function as core variables, and the real-time damping ratio of each sub-band and the minimum allowable damping ratio as constraint conditions. The minimum allowable damping ratio is preset to a fixed value according to the grid connection standard, taking a value of 0.05 under normal operating conditions, and can be adaptively fine-tuned according to different voltage level grid connection requirements. During the parameter solution process, the system traverses the feasible value range of virtual inertia, damping coefficient, and virtual impedance transfer function, and selects parameter combinations that ensure that the damping ratio of all sub-bands is greater than or equal to the preset minimum allowable damping ratio. This set of combinations constitutes the parameter safety interval at the current operating point. All adaptively adjusted parameter values ​​must fall within this interval, mathematically ensuring the wideband stability of the system.

[0040] In practical implementation, the dynamic configuration strategy of virtual impedance is based on the mid-frequency stability boundary. The system solves the stability critical boundary of mid-frequency oscillation under the current operating conditions in real time, and dynamically optimizes the center frequency, compensation gain, and quality factor parameters of each compensation unit according to the offset state of the stability boundary. When a resonance peak occurs in the mid-frequency band, the notch compensation unit corresponding to the center frequency is automatically matched to increase the compensation gain to offset the resonance peak. When the mid-frequency damping margin is insufficient, the peak compensation unit is activated to improve the system damping capability, realizing the adaptive optimization of the amplitude-frequency characteristics of virtual impedance and accurately adapting to the mid-frequency oscillation suppression requirements.

[0041] Step S12: Construct a hierarchical and collaborative adaptive controller. The adaptive controller includes adaptive adjustment units for different frequency bands. Based on the frequency band where the current oscillation occurs, the corresponding adaptive adjustment units are activated to perform corresponding adjustments, and the output parameters of each adjustment unit are limited to the parameter safety range.

[0042] The hierarchical adaptive controller constructed in this embodiment of the invention adopts a frequency band hierarchical architecture, integrating five independent adaptive adjustment units corresponding to low-frequency band, subsynchronous frequency band, fundamental frequency neighborhood, mid-frequency band, and high-frequency band. Each adjustment unit operates independently and corresponds to its own dedicated oscillation frequency band, enabling targeted control of oscillations in different frequency bands. The system identifies the dominant frequency band of the current power grid oscillation signal in real time. Based on the matching results of the oscillation frequency band, the adaptive adjustment unit corresponding to the oscillation frequency band is activated individually to perform parameter adjustment operations, while the other adjustment units that do not correspond to the frequency band remain in standby mode, avoiding parameter conflicts caused by disordered adjustment of multiple units. At the same time, a parameter limiting mechanism is set in the program logic to strictly constrain the virtual inertia, damping coefficient, and virtual impedance-related parameters output by each adjustment unit, ensuring that all output parameters are always within the safe range of parameters calculated in the second step, and preventing system stability damage caused by out-of-range parameter output.

[0043] For example, the step of activating the corresponding adaptive adjustment unit to perform corresponding adjustments based on the frequency band where the current oscillation occurs includes: If the oscillation is in the low-frequency band or subsynchronous frequency band, adjust the virtual inertia and damping coefficient within the safe range of parameters; If it is a mid-frequency oscillation, the amplitude-frequency characteristics of the virtual impedance are adjusted within the parameter safety range. Specifically, notch compensation units are dynamically configured in the virtual impedance to compensate for the resonant peak value in reverse. If it is a high-frequency oscillation, then increase the virtual resistance and reduce the current loop control bandwidth within the parameter range; Among them, the adjustment of virtual inertia and damping coefficient adopts incremental fuzzy PID control law, so that the damping ratio in the low frequency band or subsynchronous frequency band rises back to the upper limit of the parameter safety range. The virtual resistance automatically increases to a preset upper limit value in the high-frequency band, and the current loop control bandwidth decreases from the first preset value to the second preset value.

[0044] When the dominant oscillation is detected to be a low-frequency or subsynchronous oscillation, the system only activates the adaptive adjustment unit corresponding to the low-frequency and subsynchronous oscillations. Within the pre-calculated parameter safety range, it prioritizes adjusting two core parameters: virtual inertia and damping coefficient, while keeping the virtual impedance parameter constant. This embodiment uses an incremental fuzzy PID control law to achieve adaptive parameter adjustment. Compared to traditional PID algorithms, incremental fuzzy PID can dynamically correct the incremental PID parameters based on the oscillation amplitude and damping ratio deviation, possessing nonlinear adaptation capabilities and accurately matching the slow dynamic characteristics of low-frequency and subsynchronous oscillations. In implementation, the deviation value and rate of change of the damping ratio of the current frequency band from the upper limit of the safety range are used as inputs to the fuzzy controller. The adjustment increments of virtual inertia and damping coefficient are obtained through reasoning from a preset fuzzy rule base, gradually optimizing the parameter values, ultimately raising the system damping ratio of the low-frequency and subsynchronous frequency bands back to the upper limit of the parameter safety range, maximizing the low-frequency oscillation suppression margin.

[0045] When the dominant oscillation is detected to be a mid-frequency oscillation, the mid-frequency adaptive adjustment unit is activated, focusing on adjusting the amplitude-frequency characteristics of the virtual impedance while keeping the virtual inertia and damping coefficient constant in a steady state. Specifically, based on the resonant frequency and peak value of the mid-frequency oscillation, a corresponding notch compensation unit is dynamically configured in the virtual impedance transfer function. Through the inverse gain characteristic of the notch compensation, the resonant peak value of the mid-frequency band is precisely canceled, suppressing the mid-frequency resonant oscillation. The system identifies the mid-frequency oscillation mode in real time and automatically matches the optimal center frequency and compensation gain of the compensation unit to ensure the accuracy of the notch compensation. Simultaneously, all adjustment parameters are strictly limited within the safe range to avoid overcompensation that could trigger new oscillation problems.

[0046] When a high-frequency oscillation is detected as the dominant oscillation, the high-frequency adaptive adjustment unit is activated, executing a combined adjustment strategy of resistance increase and bandwidth reduction. Within the parameter safety range, the virtual resistance value is automatically increased until a preset upper limit is reached. This increases the high-frequency oscillation energy and attenuates the amplitude of the high-frequency oscillation by increasing the high-frequency virtual resistance. Simultaneously, the converter current loop control bandwidth is reduced. The current loop control bandwidth is the highest frequency at which the current control loop in the energy storage system can accurately track the AC current command signal. The current loop control bandwidth is reduced from a first preset high-frequency value to a second preset low-frequency value. The first preset value is typically set to 1000Hz, and the second preset value is typically set to 500Hz. By reducing the control bandwidth, high-frequency interference signals are filtered out, preventing the control loop from amplifying the response to high-frequency noise and oscillations, thus achieving rapid suppression of high-frequency oscillations.

[0047] It should be noted that in the embodiments of the present invention, only non-power frequency oscillations are adjusted, while the fundamental frequency neighborhood is essentially a small offset of the rated 50Hz power frequency itself, fundamental amplitude / phase imbalance, and interharmonics, which do not belong to the category of "oscillation".

[0048] Furthermore, the step of rolling back the execution parameters to the previous set of valid parameters includes: Construct a circular parameter sequence buffer of length L to store the set of controller parameters and their corresponding damping ratio satisfaction function values ​​that are within the parameter safety range in the past L effective control cycles. When a rollback is triggered, the following strategy is executed: Level 1 rollback: Roll back to the original saved set of valid parameters from the previous cycle; if the number of rollbacks exceeds a preset value within a preset number of consecutive control cycles, it is determined that an unrecoverable continuous conflict state has been entered. Secondary rollback: If a rollback is triggered again after the primary rollback, the circular parameter sequence buffer is traversed, and the set of parameters that form a weak response relationship with the current oscillation sub-band, i.e., the set with the smallest absolute value of the damping ratio sensitivity factor, is selected as the new target rollback point. The feedforward compensation channel is forcibly activated, and the current loop control bandwidth is simultaneously reduced. Level 3 safety rollback: If the rollback is triggered multiple times after the Level 2 rollback strategy is enabled, the rollback will be reverted to the pre-specified absolute safety parameters, and a device-level shutdown protection request will be issued and reported to the power grid dispatch. After the rollback control takes effect, set the satisfaction enhancement boundary function for restoring the adjustment weights of each conflicting sub-band. Only when satisfy The first set weight is restored only when the value is greater than or equal to the preset threshold; otherwise, the second set weight is maintained.

[0049] First, a parameter storage buffer is constructed. The system pre-constructs a circular parameter sequence buffer of length L, preset to 8, to store controller parameter data for the most recent 8 effective control cycles. The buffer records in real time the set of virtual inertia, damping coefficient, and virtual impedance transfer function parameters within the parameter safety range for each effective cycle. Simultaneously, it calculates and stores the damping ratio satisfaction function value for the corresponding parameters. The satisfaction function value is calculated from the deviation between the average damping ratio across the entire frequency band and the preset optimal damping ratio, used to quantitatively evaluate the control effect of the parameters. The circular parameter sequence buffer adopts a cyclic overwrite storage mode, where new effective parameters automatically overwrite the oldest historical parameters, ensuring that the stored data is the latest effective operating condition data.

[0050] When the system triggers the parameter rollback condition, a three-level differentiated rollback strategy is executed sequentially. The first level is a level one rollback strategy, which prioritizes rolling back the controller parameters directly to the set of valid parameters from the previous cycle stored in the circular parameter sequence buffer, quickly canceling erroneous adjustments and restoring the system to a stable state. Simultaneously, the system continuously counts the number of rollback triggers within consecutive control cycles. If the cumulative number of rollbacks exceeds five within a preset ten consecutive control cycles, the system is determined to have entered an unrecoverable state of continuous parameter conflict, indicating a risk of persistent instability.

[0051] The second level is a two-stage rollback strategy. If the rollback condition is triggered again during subsequent operation after the first-stage rollback is completed, it indicates that the effective parameters of the previous cycle are still not suitable for the current oscillation condition. At this time, the second-stage rollback is initiated. The system traverses all historical effective parameter sets stored in the ring parameter sequence buffer, calculates the damping ratio sensitivity factor of each parameter set with the current oscillation subband, and selects the parameter set with the smallest absolute value of the sensitivity factor. This parameter set has the weakest correlation with the response of the current oscillation mode and will not cause parameter conflicts. It is determined as the target rollback point and the parameter update is completed. At the same time, the system's feedforward compensation channel is forcibly activated. The feedforward compensation amount is used to offset the oscillation interference in advance. The feedforward compensation channel directly collects / predicts the interference amount in advance and outputs a reverse compensation signal before the disturbance affects the system, thus weakening the oscillation disturbance from the source. For example, the current broadband oscillation amplitude, frequency and other disturbance information can be extracted in real time by relying on the five-band subband filter to calculate the corresponding reverse compensation amount, and simultaneously further reduce the current loop control bandwidth to enhance the high-frequency interference suppression effect and stabilize the system operation.

[0052] The third level is the three-level safety rollback strategy. If the rollback action continues to be triggered within three consecutive control cycles after the second-level rollback strategy is enabled, it means that all currently valid historical parameters are no longer suitable for extreme abnormal operating conditions, and the system faces a serious risk of instability. At this time, the third-level safety rollback is immediately executed, directly reverting the controller parameters to the factory-preset absolute safety parameter combination. This parameter combination is a universally stable parameter combination for all operating conditions, which can maximize the basic operational stability of the system. At the same time, the system generates an equipment-level shutdown protection request command, sends it to the energy storage converter protection unit, executes power reduction or shutdown protection actions, and reports abnormal operating condition information and parameter anomaly information to the power grid dispatching platform in real time, completing fault warning and operating condition reporting.

[0053] After any level of rollback control takes effect, the system initiates weight recovery control logic, constructing a satisfaction improvement boundary function Ψ(Δξ,t) for the conflict sub-band. The function inputs are the damping ratio change and the continuous running time after rollback, used to quantitatively evaluate the recovery status of the operating condition. The system calculates the boundary function value in real time. Only when the function value is greater than or equal to a preset threshold is the current operating condition determined to be stable and the oscillation risk eliminated, and the second set weight of the corresponding regulating unit is gradually restored, i.e., the maximum control weight. If the function value is lower than the preset threshold, the operating condition is determined to still have potential instability risks, and the second set weight of the regulating unit is maintained, i.e., a low-weight participation state, to avoid triggering parameter conflicts and oscillation problems again.

[0054] Furthermore, in some optional embodiments of the present invention, after the step of activating the corresponding adaptive adjustment unit to perform corresponding adjustments according to the frequency band where the current oscillation occurs, the method further includes: After each adjustment cycle, calculate the change in damping ratio of all sub-bands before and after the adjustment; If the damping ratio of all sub-bands does not decrease and no new oscillation mode appears, then lock the current parameters; If the damping ratio of any sub-band decreases or a new oscillation characteristic appears, the execution parameters are rolled back to the previous effective parameter set, and the participation weight of the corresponding adjustment unit is reduced.

[0055] After the corresponding adaptive adjustment unit is activated based on the oscillation frequency band to complete parameter adjustment, parameter validity verification and dynamic weight correction logic are added to achieve closed-loop verification of the adjustment effect, avoiding parameter degradation and new oscillations during the adaptive adjustment process. Specifically: After each adjustment cycle, the change in damping ratio of each sub-band is statistically analyzed. This invention sets a fixed control adjustment cycle, with a single cycle duration of 10ms, to adapt to the control response speed of the energy storage converter. After each adaptive parameter adjustment action is completed and the current adjustment cycle ends, the updated real-time damping ratios of the five sub-bands are recalculated using a recursive least squares identification algorithm. The difference between the updated damping ratios of each sub-band and the original damping ratios before this adjustment is calculated to obtain the change in damping ratio for all sub-bands. Simultaneously, it detects whether any existing oscillation modes disappear or new oscillation modes are generated in each sub-band, thus fully recording the system oscillation state changes after this adjustment.

[0056] The parameter locking logic is executed based on the changes in damping ratio and oscillation mode. A comprehensive assessment of the damping ratio changes and oscillation mode states for all sub-bands is performed. If the damping ratio changes for all five sub-bands are greater than or equal to zero (meaning the damping ratio has not decreased across all frequency bands, and no new oscillation modes are generated throughout the process), it indicates that the adaptive parameter adjustment is effective, and the system's broadband stability performance has been optimized or maintained at its optimal state. At this point, the system executes a parameter locking command, fixing the current virtual inertia, damping coefficient, and virtual impedance transfer function parameters. In the next adjustment cycle, no further parameter adjustment will be triggered, maintaining the steady-state of the system control parameters and preventing invalid parameter fluctuations from affecting grid stability.

[0057] Based on abnormal conditions, the system executes parameter rollback and weight downgrade logic. If the damping ratio change of any sub-band is detected to be less than zero, indicating a decrease in the damping ratio of the corresponding frequency band, the system's ability to suppress oscillations in that frequency band is weakened. Alternatively, if any sub-band exhibits entirely new oscillation characteristics or a new oscillation mode, it indicates that the adaptive adjustment action has a negative effect, and the parameter adjustment is not suitable for the current operating conditions. In this case, the system immediately performs a parameter rollback operation, restoring all controller parameters to the effective parameter set of the previous control cycle, eliminating the adverse effects of erroneous parameters. Simultaneously, the system locates the corresponding frequency band adjustment unit where the damping has decreased or a new oscillation has appeared, and reduces the participation weight of that unit. The weight reduction is preset to 30% to 50%, limiting the subsequent parameter adjustment intensity of that unit and preventing it from continuously outputting erroneous adjustment values ​​that could worsen system oscillations. The weights of other normal adjustment units remain unchanged, ensuring that the system's basic adjustment capability is not affected.

[0058] Step S13: The final virtual inertia, damping coefficient and virtual impedance transfer function output by each adjustment unit are summarized according to the preset priority and output to the voltage and current dual closed-loop controller of the grid-type energy storage converter to complete the adaptive suppression of wideband oscillation.

[0059] In this process, after each adaptive adjustment unit completes its corresponding parameter adjustment and outputs valid parameter values, the various parameters output from multiple frequency bands are merged and summarized according to preset priorities. High-priority parameters directly cover the conflicting adjustment amounts of low-priority parameters, while non-conflicting parameters are directly superimposed and retained to obtain the final optimal combination of control parameters. The summarized virtual inertia, damping coefficient, and virtual impedance transfer function parameters are then sent in real time to the voltage and current dual-loop controller of the grid-connected energy storage converter to update the core control parameters of the dual-loop controller. Through the dynamic adjustment of the controller, the output voltage and current of the converter are corrected, offsetting the oscillation components of various frequency bands in the power grid, ultimately achieving adaptive suppression of wideband oscillations and maintaining the stable operation of the energy storage grid-connected system.

[0060] In summary, the adaptive adjustment-based grid-connected energy storage control method in the above embodiments of the present invention acquires the three-phase voltage and three-phase current broadband electrical quantity signals at the grid connection point of the grid-connected energy storage converter, and extracts broadband oscillation features from the broadband electrical quantity signals to obtain broadband oscillation feature vectors. This enables comprehensive capture of the oscillation mode characteristics of multiple frequency bands in the power grid, including low-frequency, subsynchronous, and medium-to-high-frequency bands, achieving accurate perception of broadband oscillation information across the entire domain. By combining a pre-established broadband stability domain model, the method calculates the parameter safety ranges of virtual inertia, damping coefficient, and virtual impedance transfer function in the virtual synchronous machine control in real time online. It can adapt to power grid topology, operating status, and wideband impedance nonlinear time-varying characteristics, defining a stable boundary for adaptive parameter adjustment. By constructing a hierarchical collaborative adaptive controller containing adaptive adjustment units of different frequency bands, the corresponding adjustment unit is activated according to the actual oscillation frequency band, and the output parameters of each unit are constrained within a safe range. This enables targeted and precise control of oscillations in each frequency band, while avoiding instability caused by parameter adjustment exceeding limits. Finally, by pre-setting priorities, the control parameters output by each unit are summarized and input into a voltage and current dual closed-loop controller, achieving coordinated matching and unified output of multi-frequency band control parameters. On the one hand, it breaks through the limitations of existing technologies that only focus on low-frequency parameter optimization and cannot actively sense and suppress mid-to-high frequency resonance peaks. It abandons the traditional fixed parameter control method and can adapt the control parameters in real time with the dynamic changes of the power grid, effectively improving the stability margin of the mid-to-high frequency band and eliminating the wideband oscillation problem caused by mid-to-high frequency series and parallel resonance. At the same time, it avoids the drawbacks of poor adaptability, control lag, and sacrifice of dynamic response speed of fixed parameter control. On the other hand, it solves the defects of existing technologies where the control modules of each frequency band are isolated and there is no coordinated adjustment mechanism. It eliminates the problem of conflict between high and low frequency band control strategies and the risk of inducing new oscillations. It can track the changes in the stability region of the entire frequency band of the power grid in real time and realize the integrated coordinated suppression of multi-mode oscillations. Ultimately, it effectively improves the grid full-band adaptability of grid-based energy storage converters and solves the technical problems of poor wideband oscillation suppression effect and inability to adapt to complex wideband power grid scenarios due to insufficient mid-to-high frequency stability margin and lack of wideband coordinated control capability in existing technologies.

[0061] Example 2 This embodiment also proposes an adaptive adjustment-based grid-type energy storage control method. The difference between the adaptive adjustment-based grid-type energy storage control method in this embodiment and the adaptive adjustment-based grid-type energy storage control method in Embodiment 1 is as follows: The step of calculating the safe range of parameters required to ensure stability of virtual inertia, damping coefficient, and virtual impedance transfer function in virtual synchronous machine control over a wide frequency range at the current operating point also includes: A broadband discrete closed-loop control state-space model of a grid-type energy storage system is established and constructed as an augmented state-space matrix. Two consecutive first-order eigenvalue perturbation analyses are performed on the augmented state-space matrix to solve the perturbation sensitivity factor of the damping ratio of the current controller parameters along the coordinate axis and along a specific Jacobian descent direction. After normalizing each sensitivity factor, construct the parameter-damping ratio Jacobian matrix at the current operating point, and calculate the projection of the current parameter adjustment direction onto this matrix to generate the feedforward control quantity in a directional manner. The feedforward control quantity is directly superimposed on the output of each hierarchical control unit, so that before the feedback loop in each control unit responds, the feedforward component actively adjusts the parameters to counteract the foreseeable trend of decreasing broadband damping ratio.

[0062] To further enhance the foresight of parameter tuning and eliminate the lag in feedback tuning, after calculating the parameter safety range, perturbation sensitivity analysis, Jacobian matrix construction, and feedforward control compensation logic are added to achieve predictive parameter tuning. The specific implementation method is as follows: First, a wideband discrete closed-loop state-space model is constructed and the perturbation sensitivity factor is solved. Firstly, combining the discrete control logic, main circuit topology, and virtual synchronous machine control closed-loop architecture of the grid-connected energy storage converter, a discrete closed-loop control state-space model adapted to wideband operating conditions is established. Virtual inertia, damping coefficient, and virtual impedance parameters are used as state variables, grid oscillation disturbances are used as input variables, and the damping ratio of each frequency band is used as the output variable. Further, the basic state-space model is augmented, introducing parameter perturbation increments to construct an augmented state-space matrix, fully covering the impact of minute parameter changes on the system state. Subsequently, two consecutive eigenvalue first-order perturbation analyses are performed on the augmented state-space matrix. The first perturbation analysis solves for the parameter perturbation sensitivity factor of the damping ratio of each frequency band when the controller parameters change along a single coordinate axis, characterizing the degree of influence of independent adjustment of a single parameter on damping. The second perturbation analysis solves for the damping ratio perturbation sensitivity factor when the parameters change along the Jacobian decreasing direction, characterizing the degree of coupling influence of multi-parameter coordinated adjustment on damping, obtaining comprehensive parameter-damping sensitivity characteristics.

[0063] Secondly, a parameter-damping ratio Jacobian matrix is ​​constructed and the optimal adjustment direction is determined. All calculated sensitivity factors are normalized to eliminate calculation errors caused by differences in the dimensions of different parameters, ensuring the comparability of each sensitivity factor. Based on the normalized sensitivity factors, a parameter-damping ratio Jacobian matrix is ​​constructed at the current operating point. This matrix accurately represents the nonlinear mapping relationship and coupling characteristics between all control parameters and the full-band damping ratio. The system substitutes the parameter adjustment direction of the current feedback regulation output into the Jacobian matrix, calculates the optimal projection vector of the adjustment direction in the matrix space, and accurately locates the parameter adjustment direction that maximizes the improvement of system damping and suppresses damping decrease. Based on this, the corresponding feedforward regulation control quantity is generated. Finally, predictive stability regulation is achieved by superimposing feedforward control quantities. The calculated feedforward regulation control quantities are directly superimposed on the parameter output terminals of each hierarchical adaptive regulation unit, and then input into the dual closed-loop controller after being merged with the feedback regulation output quantities. The core advantage of this control logic is its proactive regulation. In the period before the feedback loops of each regulation unit respond to changes in grid oscillations and before the damping ratio decreases, the feedforward component can actively correct the controller parameters based on the prediction results, offsetting the foreseeable broadband damping ratio decline trend in advance. This compensates for the lag in traditional pure feedback regulation, significantly improves the dynamic response speed and suppression accuracy of the grid-type energy storage system to broadband oscillations, and further enhances the system's broadband operational stability.

[0064] In summary, the adaptive adjustment-based grid-connected energy storage control method in the above embodiments of the present invention acquires the three-phase voltage and three-phase current broadband electrical quantity signals at the grid connection point of the grid-connected energy storage converter, and extracts broadband oscillation features from the broadband electrical quantity signals to obtain broadband oscillation feature vectors. This enables comprehensive capture of the oscillation mode characteristics of multiple frequency bands in the power grid, including low-frequency, subsynchronous, and medium-to-high-frequency bands, achieving accurate perception of broadband oscillation information across the entire domain. By combining a pre-established broadband stability domain model, the method calculates the parameter safety ranges of virtual inertia, damping coefficient, and virtual impedance transfer function in the virtual synchronous machine control in real time online. It can adapt to power grid topology, operating status, and wideband impedance nonlinear time-varying characteristics, defining a stable boundary for adaptive parameter adjustment. By constructing a hierarchical collaborative adaptive controller containing adaptive adjustment units of different frequency bands, the corresponding adjustment unit is activated according to the actual oscillation frequency band, and the output parameters of each unit are constrained within a safe range. This enables targeted and precise control of oscillations in each frequency band, while avoiding instability caused by parameter adjustment exceeding limits. Finally, by pre-setting priorities, the control parameters output by each unit are summarized and input into a voltage and current dual closed-loop controller, achieving coordinated matching and unified output of multi-frequency band control parameters. On the one hand, it breaks through the limitations of existing technologies that only focus on low-frequency parameter optimization and cannot actively sense and suppress mid-to-high frequency resonance peaks. It abandons the traditional fixed parameter control method and can adapt the control parameters in real time with the dynamic changes of the power grid, effectively improving the stability margin of the mid-to-high frequency band and eliminating the wideband oscillation problem caused by mid-to-high frequency series and parallel resonance. At the same time, it avoids the drawbacks of poor adaptability, control lag, and sacrifice of dynamic response speed of fixed parameter control. On the other hand, it solves the defects of existing technologies where the control modules of each frequency band are isolated and there is no coordinated adjustment mechanism. It eliminates the problem of conflict between high and low frequency band control strategies and the risk of inducing new oscillations. It can track the changes in the stability region of the entire frequency band of the power grid in real time and realize the integrated coordinated suppression of multi-mode oscillations. Ultimately, it effectively improves the grid full-band adaptability of grid-based energy storage converters and solves the technical problems of poor wideband oscillation suppression effect and inability to adapt to complex wideband power grid scenarios due to insufficient mid-to-high frequency stability margin and lack of wideband coordinated control capability in existing technologies.

[0065] Example 3 Please see Figure 2 The figure shows a grid-type energy storage control system based on adaptive adjustment proposed in the third embodiment of the present invention. The system includes: The acquisition module 100 is used to acquire the broadband electrical quantity signal at the grid connection point of the grid-connected energy storage converter. The broadband electrical quantity signal includes three-phase voltage and three-phase current. The broadband oscillation feature is extracted from the broadband electrical quantity signal to obtain the broadband oscillation feature vector. Module 200 is established to calculate online the parameter safety range required to ensure stability of virtual inertia, damping coefficient and virtual impedance transfer function in the virtual synchronous machine control at the current operating point within a wide frequency range, based on the wide frequency oscillation characteristic vector and the pre-established wide frequency stability domain model. The startup module 300 is used to construct a hierarchical and collaborative adaptive controller. The adaptive controller includes adaptive adjustment units for different frequency bands. Based on the frequency band where the current oscillation occurs, the corresponding adaptive adjustment unit is started to perform corresponding adjustments, and the output parameters of each adjustment unit are limited to the parameter safety range. The control module 400 is used to summarize the final virtual inertia, damping coefficient and virtual impedance transfer function output by each regulation unit according to the preset priority and output them to the voltage and current dual closed-loop controller of the grid-type energy storage converter to complete the adaptive suppression of wideband oscillation.

[0066] The functions or operation steps implemented by the above modules are largely the same as those in the above method embodiments, and will not be repeated here.

[0067] Example 4 In another aspect, the present invention provides a readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the steps of the method described in any one of Embodiments 1 to 2 above.

[0068] Example 5 In another aspect, the present invention provides an electronic device, the electronic device including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the program to implement the steps of any one of the methods described in Embodiments 1 to 2 above.

[0069] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0070] Those skilled in the art will understand that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequential list of executable instructions for implementing logical functions, and can be embodied in any computer-readable storage medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable storage medium" can mean any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0071] More specific examples (a non-exhaustive list) of computer-readable storage media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable storage media can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0072] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0073] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0074] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A grid-type energy storage control method based on adaptive adjustment, characterized in that, The method includes: Obtain the broadband electrical quantity signal at the grid connection point of the grid-connected energy storage converter. The broadband electrical quantity signal includes three-phase voltage and three-phase current. Extract broadband oscillation features from the broadband electrical quantity signal to obtain a broadband oscillation feature vector. Based on the broadband oscillation characteristic vector and combined with the pre-established broadband stability domain model, the safe range of parameters required to ensure stability of virtual inertia, damping coefficient and virtual impedance transfer function in the virtual synchronous machine control at the current operating point is calculated online. A hierarchical and collaborative adaptive controller is constructed, which includes adaptive adjustment units for different frequency bands. Based on the frequency band where the current oscillation occurs, the corresponding adaptive adjustment unit is activated to perform corresponding adjustments, and the output parameters of each adjustment unit are limited to the parameter safety range. The final virtual inertia, damping coefficient, and virtual impedance transfer function output by each regulation unit are summarized according to a preset priority and then output to the voltage and current dual closed-loop controller of the grid-type energy storage converter to complete the adaptive suppression of wideband oscillation. The wideband frequency range includes 10Hz to 1000Hz; The expression for the parameter safety interval is: in, For virtual inertia, The damping coefficient is... This is the virtual impedance transfer function. For the first Damping ratio of the belt, This is the preset minimum allowable damping ratio; The expression for the virtual impedance transfer function is: in, For virtual resistance, For virtual inductance, For the differential operator, the third term represents multiple dynamically configurable notch / peak compensation units. For the center frequency, To compensate for the gain, The quality factor is used to optimize the configuration in real time based on the stability boundary of the mid-frequency band. This represents the total number of compensation units currently in use by the system. For the first Branch number of each independent notch / peak compensation unit.

2. The adaptive adjustment-based grid-type energy storage control method according to claim 1, characterized in that, The step of extracting broadband oscillation features from broadband electrical quantity signals to obtain broadband oscillation feature vectors includes: A set of adaptive bandpass filter banks with dynamically shiftable center frequency is used to perform parallel filtering on wideband electrical signals, and the signals are divided into five analysis subbands according to low frequency band, subsynchronous frequency band, fundamental frequency neighborhood, mid frequency band and high frequency band. For each sub-band, the oscillation amplitude, oscillation frequency, damping ratio and phase difference of that frequency band are calculated in real time using a recursive least squares identification algorithm, and a wideband oscillation feature vector is output. The broadband oscillation eigenvector is represented as: in, Representing the first The oscillation amplitude, oscillation frequency, damping ratio, and phase difference of each sub-band.

3. The adaptive adjustment-based grid-type energy storage control method according to claim 2, characterized in that, Following the step of activating the corresponding adaptive adjustment unit to perform corresponding adjustments based on the frequency band where the current oscillation occurs, the method further includes: After each adjustment cycle, calculate the change in damping ratio of all sub-bands before and after the adjustment; If the damping ratio of all sub-bands does not decrease and no new oscillation mode appears, then lock the current parameters; If the damping ratio of any sub-band decreases or a new oscillation characteristic appears, the execution parameters are rolled back to the previous effective parameter set, and the participation weight of the corresponding adjustment unit is reduced.

4. The adaptive adjustment-based grid-type energy storage control method according to claim 3, characterized in that, The steps of activating the corresponding adaptive adjustment unit to perform corresponding adjustments based on the frequency band where the current oscillation occurs include: If the oscillation is in the low-frequency band or subsynchronous frequency band, adjust the virtual inertia and damping coefficient within the safe range of parameters; If it is a mid-frequency oscillation, the amplitude-frequency characteristics of the virtual impedance are adjusted within the parameter safety range. Specifically, notch compensation units are dynamically configured in the virtual impedance to compensate for the resonant peak value in reverse. If it is a high-frequency oscillation, then increase the virtual resistance and reduce the current loop control bandwidth within the parameter range; Among them, the adjustment of virtual inertia and damping coefficient adopts incremental fuzzy PID control law, so that the damping ratio in the low frequency band or subsynchronous frequency band rises back to the upper limit of the parameter safety range. The virtual resistance automatically increases to a preset upper limit value in the high-frequency band, and the current loop control bandwidth decreases from the first preset value to the second preset value.

5. The adaptive adjustment-based grid-type energy storage control method according to claim 4, characterized in that, The step of rolling back the execution parameters to the previous set of valid parameters includes: Construct a circular parameter sequence buffer of length L to store the set of controller parameters and their corresponding damping ratio satisfaction function values ​​that are within the parameter safety range in the past L effective control cycles. When a rollback is triggered, the following strategy is executed: Level 1 rollback: Roll back to the original saved set of valid parameters from the previous cycle; if the number of rollbacks exceeds a preset value within a preset number of consecutive control cycles, it is determined that an unrecoverable continuous conflict state has been entered. Secondary rollback: If a rollback is triggered again after the primary rollback, the circular parameter sequence buffer is traversed, and the set of parameters that form a weak response relationship with the current oscillation sub-band, i.e., the set with the smallest absolute value of the damping ratio sensitivity factor, is selected as the new target rollback point. The feedforward compensation channel is forcibly activated, and the current loop control bandwidth is simultaneously reduced. Level 3 safety rollback: If the rollback is triggered multiple times after the Level 2 rollback strategy is enabled, the rollback will be reverted to the pre-specified absolute safety parameters, and a device-level shutdown protection request will be issued and reported to the power grid dispatch. After the rollback control takes effect, set the satisfaction enhancement boundary function for restoring the adjustment weights of each conflicting sub-band. Only when satisfy The first set weight is restored only when the value is greater than or equal to the preset threshold; otherwise, the second set weight is maintained.

6. The adaptive adjustment-based grid-type energy storage control method according to claim 1, characterized in that, The step of calculating the safe range of parameters required to ensure stability of virtual inertia, damping coefficient, and virtual impedance transfer function in virtual synchronous machine control over a wide frequency range at the current operating point also includes: A broadband discrete closed-loop control state-space model of a grid-type energy storage system is established and constructed as an augmented state-space matrix. Two consecutive first-order eigenvalue perturbation analyses are performed on the augmented state-space matrix to solve the perturbation sensitivity factor of the damping ratio of the current controller parameters along the coordinate axis and along a specific Jacobian descent direction. After normalizing each sensitivity factor, construct the parameter-damping ratio Jacobian matrix at the current operating point, and calculate the projection of the current parameter adjustment direction onto this matrix to generate the feedforward control quantity in a directional manner. The feedforward control quantity is directly superimposed on the output of each hierarchical control unit, so that before the feedback loop in each control unit responds, the feedforward component actively adjusts the parameters to counteract the foreseeable trend of decreasing broadband damping ratio.

7. A grid-type energy storage control system based on adaptive adjustment, characterized in that, The system includes: The acquisition module is used to acquire the broadband electrical quantity signal at the grid connection point of the grid-connected energy storage converter. The broadband electrical quantity signal includes three-phase voltage and three-phase current. The broadband oscillation feature is extracted from the broadband electrical quantity signal to obtain the broadband oscillation feature vector. A module is established to calculate online the parameter safety range required to ensure stability of virtual inertia, damping coefficient and virtual impedance transfer function in virtual synchronous machine control at the current operating point within a wide frequency range, based on the wide frequency oscillation characteristic vector and a pre-established wide frequency stability domain model. The startup module is used to build a hierarchical and collaborative adaptive controller. The adaptive controller includes adaptive adjustment units for different frequency bands. Based on the frequency band where the current oscillation occurs, the corresponding adaptive adjustment unit is started to perform corresponding adjustments, and the output parameters of each adjustment unit are limited to the parameter safety range. The control module is used to summarize the final virtual inertia, damping coefficient and virtual impedance transfer function output by each regulation unit according to the preset priority and output them to the voltage and current dual closed-loop controller of the grid-type energy storage converter to complete the adaptive suppression of wideband oscillation. The wideband frequency range includes 10Hz to 1000Hz; The expression for the parameter safety interval is: in, For virtual inertia, The damping coefficient is... This is the virtual impedance transfer function. For the first Damping ratio of the belt, This is the preset minimum allowable damping ratio; The expression for the virtual impedance transfer function is: in, For virtual resistance, For virtual inductance, For the differential operator, the third term represents multiple dynamically configurable notch / peak compensation units. For the center frequency, To compensate for the gain, The quality factor is used to optimize the configuration in real time based on the stability boundary of the mid-frequency band. This represents the total number of compensation units currently in use by the system. For the first Branch number of each independent notch / peak compensation unit.

8. A readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method as described in any one of claims 1 to 5.

9. An electronic device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor, when executing the program, implements the steps of the method as described in any one of claims 1 to 5.

Images