Grid-based energy storage methods and systems based on dynamic grid strength sensing

Abstract

This invention discloses a method and system for grid-connected energy storage based on dynamic grid strength sensing, belonging to the field of grid-connected energy storage technology. The method specifically includes: collecting grid information at the grid connection point of the grid-connected energy storage system, generating a grid strength state quantity based on the grid information, adaptively adjusting the grid-connected control parameters according to the grid strength state quantity, constructing stable operation constraints for the grid-connected energy storage system based on the grid strength state quantity, generating active and / or reactive power output commands for the grid-connected energy storage system under the stable operation constraints, adaptively adjusting the output power of the grid-connected energy storage system according to changes in grid strength, and triggering grid control reconfiguration and readjusting the grid control parameters when a sudden change in the grid strength state quantity is detected. This application can maintain the coordinated consistency between the grid control characteristics and grid characteristics under complex, variable, and even abrupt grid conditions, improving the operational reliability of grid-connected energy storage in both weak and strong grid scenarios.

Description

Technical Field

[0001] This invention belongs to the field of grid-based energy storage technology, specifically a grid-based energy storage method and system based on dynamic sensing of grid strength. Background Technology

[0002] With the high proportion of new energy sources being integrated, the large-scale parallel connection of power electronic equipment, and the gradual decrease in the proportion of traditional large-scale synchronous power sources, the overall inertia and damping level of the power system are continuously declining, and the grid strength exhibits significant fluctuations and regional differences. In particular, in weak grids, end-point grids, and islanded grid operation scenarios, problems such as insufficient voltage support capacity, weakened frequency support capacity, and fragile phase angle stability are more prominent.

[0003] In recent years, grid-based energy storage technology has gradually become an important means of supporting the stable operation of power systems. By providing the grid with virtual inertia, virtual damping, and voltage source characteristics similar to synchronous machines, it is considered an important direction to replace the support capabilities of synchronous machines. However, existing grid-based energy storage technologies mostly adopt static parameter design or empirical range setting methods, failing to achieve dynamic identification of grid strength and lacking the ability to evolve grid parameters in real time with grid characteristics. This leads to problems such as parameter mismatch, control stiffness deviation, and insufficient coordination of voltage and power regulation under different strong and weak grid environments, affecting the stable applicability of grid-based energy storage in complex grids. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention proposes a grid-based energy storage method and system based on dynamic perception of grid strength. This method can dynamically perceive, model, and classify grid strength, and accordingly drive adaptive adjustment of grid-based energy storage control parameters and power output boundaries to adapt to complex and ever-changing grid operating environments, thereby improving the universality and reliability of grid-based energy storage in both weak and strong grid scenarios.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] Grid-based energy storage methods based on dynamic grid strength sensing include:

[0007] Collect grid information on grid-connected energy storage points and generate grid strength state parameters based on the grid information, which includes voltage, current and frequency information;

[0008] Based on the power grid strength state quantity, the grid construction control parameters are adaptively adjusted, including virtual inertia, virtual impedance, and power control coefficient.

[0009] Based on the power grid strength state variables, stable operation constraints for grid-connected energy storage are constructed, and active and / or reactive power output commands for grid-connected energy storage are generated under the stable operation constraints. The output power of grid-connected energy storage is adaptively adjusted according to changes in power grid strength.

[0010] When a sudden change in the power grid strength state is detected, the grid construction control reconfiguration is triggered, and the grid construction control parameters are readjusted.

[0011] Specifically, the process of collecting grid information at the grid-connected energy storage points and generating grid intensity state variables based on the grid information includes:

[0012] The synchronous sampling sequence of three-phase voltage and three-phase current at the grid connection point of the grid-connected energy storage is collected, and the sampling sequence is time-aligned to form grid information under the same time reference;

[0013] The power grid information is segmented, and a set of windows including steady-state windows and transition windows is generated according to a preset time window. A corresponding subset of power grid information is established for each window.

[0014] Within the transition window, a preset perturbation sequence is superimposed, and the voltage and current responses during the period of the perturbation sequence are recorded simultaneously to form identification data pairs.

[0015] Based on the identification data, a set of candidate equivalent impedances is generated in multiple frequency bands or multiple phase sequence dimensions, and the set of candidate equivalent impedances is subjected to consistency screening to obtain the target equivalent impedance parameters.

[0016] The target equivalent impedance parameter is fused with the voltage, current and frequency information within the steady-state window to output a power grid strength state quantity that is updated as the window scrolls.

[0017] Specifically, based on the identification data, a set of candidate equivalent impedances is generated across multiple frequency bands or multiple phase sequence dimensions, and the set of candidate equivalent impedances is subjected to consistency screening to obtain the target equivalent impedance parameters, including:

[0018] The identification data is divided into multiple data subsets according to a preset dimension, and a corresponding identification task identifier is established for each data subset.

[0019] Based on the identification task identifier, voltage and current correlation sample pairs are extracted from the corresponding data subset, and multiple candidate equivalent impedance parameters are generated based on the correlation sample pairs, which are then aggregated to form a candidate equivalent impedance set.

[0020] The candidate equivalent impedance set is subjected to consistency screening. The candidate equivalent impedance parameters are compared pairwise according to a preset comparison rule to generate a set of consistent results.

[0021] The remaining candidate equivalent impedance parameters after consistency screening are fused to output the target equivalent impedance parameter. The fusion includes selection by frequency band priority, selection by phase sequence dimension priority, or selection by preset combination rules.

[0022] Specifically, the target equivalent impedance parameter is fused with the voltage, current, and frequency information within the steady-state window to output a grid strength state quantity that updates as the window rolls, including:

[0023] Within the steady-state window, window feature sets corresponding to voltage, current, and frequency information are generated respectively, and the window feature sets are assigned a timestamp identifier corresponding to the steady-state window. The window feature sets include voltage amplitude features, current amplitude features, and frequency offset features.

[0024] The target equivalent impedance parameter is aligned with the window feature set to form a fused input set. The alignment includes mapping the target equivalent impedance parameter to a window index that is consistent with the timestamp identifier, and replacing the target equivalent impedance parameter at the missing index with a preset placeholder parameter or backtracking parameter.

[0025] Based on the fused input set, a set of candidate state quantities for power grid strength is generated. The set of candidate state quantities for power grid strength includes a first candidate state quantity generated primarily by the target equivalent impedance parameter, a second candidate state quantity generated primarily by the window feature set, and a third candidate state quantity generated by combining the target equivalent impedance parameter and the window feature set according to a preset rule.

[0026] During the window scrolling update process, the set of candidate power grid strength states is selected or switched, and the power grid strength states are output as the window scrolls.

[0027] Specifically, based on the aforementioned power grid strength state variables, the grid control parameters are adaptively adjusted, including:

[0028] Obtain the power grid strength status quantity corresponding to the current window, and map the power grid strength status quantity to the target strength level in the preset strength level set, wherein the strength level set includes weak grid level and strong grid level;

[0029] Based on the target intensity level, select the parameter strategy entry corresponding to the target intensity level from the preset parameter strategy library, and generate a parameter update task identifier for the parameter strategy entry. The preset parameter strategy library includes virtual inertia parameter strategy, virtual impedance parameter strategy and power control coefficient strategy.

[0030] The parameter update task identifiers are decoupled and sorted to form a parameter update sequence according to a preset order, which includes updating the virtual impedance first, then updating the virtual inertia, and finally updating the power control coefficient.

[0031] Based on each parameter in the parameter update sequence, a corresponding parameter transition trajectory is generated and written into the parameter register set of the network controller. The parameter transition trajectory includes a segmented incremental trajectory, a rolling hold trajectory, or an event-triggered trajectory.

[0032] When the next window arrives, the correspondence between the power grid strength state quantity and the written parameter register set is verified. If the correspondence does not meet the preset consistency condition, the parameter strategy entry is backtracked and the parameter update sequence is reconstructed to complete the adaptive adjustment of the network control parameters. The network control parameters include virtual inertia, virtual impedance and power control coefficient.

[0033] Specifically, the parameter update task identifiers are decoupled and sorted to form a parameter update sequence according to a preset order, including:

[0034] The update task identifiers for each parameter are categorized and marked, and then assigned to different parameter groups. A group priority identifier is also assigned to each parameter group.

[0035] Based on the mutual influence relationship between parameter update task identifiers within each parameter group, a sequential constraint linked list is established for any two parameter update task identifiers. Parameter update task identifiers that are directly related are recorded as task pairs with sequential constraints, thus forming a parameter update sequence constraint set.

[0036] According to the group priority identifier and the parameter update order constraint set, perform hierarchical sorting on all parameter update task identifiers: first, determine the update order of each parameter group at the group level, and then sort the parameter update task identifiers locally within the group according to the order constraint linked list to generate an intermediate parameter update sequence.

[0037] The intermediate parameter update sequence is subjected to consistency adjustment, and the local sorting segments that do not meet the group priority identifier or order constraint linked list are rearranged to obtain the parameter update sequence that meets the preset order relationship.

[0038] Specifically, based on each parameter in the parameter update sequence, a corresponding parameter transition trajectory is generated, and the parameter transition trajectory is written into the parameter register set of the network controller, including:

[0039] Read the parameter values ​​and target parameter values ​​of each parameter in the parameter update sequence to form parameter state pairs, and group them according to the virtual impedance, virtual inertia or power control coefficient to which the parameters belong, and add trajectory type labels to each parameter state pair;

[0040] Based on the trajectory type marker, a preset trajectory generation rule is selected for each parameter state pair to generate a parameter transition trajectory template;

[0041] According to the order of the parameter update sequence, each parameter transition trajectory template is expanded into a corresponding discrete parameter sample sequence, and each discrete parameter sample sequence is associated with the control cycle number or time step number to form a parameter transition trajectory set.

[0042] The parameter transition trajectory set is written into the parameter register set of the network controller.

[0043] Specifically, based on the grid strength state variables, stable operation constraints for grid-connected energy storage are constructed, and under these constraints, active and / or reactive power output commands for grid-connected energy storage are generated. The output power of grid-connected energy storage is adaptively adjusted according to changes in grid strength, including:

[0044] Obtain the power grid strength status quantity corresponding to the current window, and map the power grid strength status quantity to the target operating mode identifier in the preset operating mode set. The preset operating mode set includes weak grid mode, transition mode and strong grid mode.

[0045] Based on the target operating mode identifier, a stable operating constraint template corresponding to it is selected from the constraint template library, and the stable operating constraint template is scaled or translated based on the current power grid strength state quantity to generate a stable operating constraint set.

[0046] Collect the expected active power and expected reactive power from the outside, and combine the expected active power and expected reactive power with the set of stable operation constraints to form a power expectation determination result that is inside or outside the set of stable operation constraints.

[0047] After power correction, the power expectation determination result is projected onto the boundary or interior of the stable operation constraint set to generate constrained active power output command and / or reactive power output command.

[0048] During window scrolling, when a change in grid strength is detected, the stable operation constraint template corresponding to the new target operation mode identifier is invoked, and the previously generated power output command is smoothly transitioned to form an active power output command and / or reactive power output command that changes with the grid strength.

[0049] Specifically, after power correction, the power expectation determination result is projected onto the boundary or interior of the stable operation constraint set to generate constrained active power output commands and / or reactive power output commands, including:

[0050] Obtain the power expectation determination result after the determination of the stable operation constraint set, mark the active power expectation and / or reactive power expectation outside the stable operation constraint set as the power expectation to be corrected, and assign a projection priority identifier to each power expectation to be corrected according to the preset strategy of active power priority or reactive power priority.

[0051] Based on the projection priority identifier, the corresponding projection rule set is selected, and the projection direction of the expected power to be corrected is determined. The projection direction includes the active axis direction, the reactive axis direction, or the combined active and reactive axis direction.

[0052] Search for power points that satisfy preset constraints along the projection direction at the boundary or inside the stable operation constraint set, record the power points as a candidate projection power point set, and select the target projection power point from the candidate projection power point set according to a preset sorting rule.

[0053] The active and / or reactive components of the target projected power point are output as constrained active power output commands and / or reactive power output commands, respectively.

[0054] The grid-connected energy storage system based on dynamic grid strength sensing includes: an information acquisition module, a control parameter adjustment module, a command output module, and a sudden change adjustment module;

[0055] The information acquisition module is used to collect grid information at the grid-connected energy storage points and generate grid strength state quantities based on the grid information. The grid information includes voltage, current and frequency information.

[0056] The control parameter adjustment module is used to adaptively adjust the grid construction control parameters according to the grid strength state quantity. The grid construction control parameters include virtual inertia, virtual impedance, and power control coefficient.

[0057] The instruction output module is used to construct stable operation constraints for grid-connected energy storage based on the grid strength state quantity, and generate active and / or reactive power output instructions for grid-connected energy storage under the stable operation constraints, and adaptively adjust the output power of grid-connected energy storage according to changes in grid strength.

[0058] The mutation adjustment module is used to trigger network control reconfiguration and readjust network control parameters when a sudden change in the power grid strength state is detected.

[0059] Compared with the prior art, the beneficial effects of the present invention are:

[0060] This invention proposes a grid-based energy storage method and system based on dynamic grid strength perception. By collecting grid information and identifying disturbances online at the grid connection point of the energy storage system, a grid strength state variable that is updated over time is formed. This grid strength state variable is used as the core to drive the adaptive tuning of grid control parameters, the construction of stable operation constraints, and the dynamic adjustment of power output commands. This realizes the transformation of grid-based energy storage from fixed-parameter grid construction to grid construction that evolves with grid strength. Overall, the system achieves active identification and matching control of grid-based energy storage for different strong and weak grid operating environments. It can maintain the synergistic consistency between grid control characteristics and grid characteristics under complex, variable, and even abrupt grid conditions, thereby improving the operational reliability of grid-based energy storage in both weak and strong grid scenarios. Attached Figure Description

[0061] Figure 1 A flowchart of the grid-based energy storage method based on dynamic sensing of grid strength provided by the present invention;

[0062] Figure 2 A schematic diagram of power grid strength state quantity generation provided by the present invention;

[0063] Figure 3 This is a schematic diagram illustrating the adaptive adjustment of network control parameters provided by the present invention.

[0064] Figure 4 The diagram shows the architecture of the grid-connected energy storage system based on dynamic grid strength sensing provided by this invention. Detailed Implementation

[0065] The present application will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present application, but do not limit the present application in any way. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present application. These all fall within the protection scope of the present application.

[0066] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0067] It should be noted that, unless there is a conflict, the various features in the embodiments of this application can be combined with each other, all of which are within the protection scope of this application. Furthermore, although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than the module division in the device or the order in the flowchart. In addition, the terms "first," "second," and "third" used in this application do not limit the data or execution order, but only distinguish identical or similar items with essentially the same function and effect.

[0068] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The term "and / or" as used in this specification includes any and all combinations of one or more of the associated listed items.

[0069] Example 1

[0070] Please see Figures 1-3 The present invention provides an embodiment of a grid-based energy storage method based on dynamic grid strength sensing, comprising the following specific steps:

[0071] Step S1: Collect grid information of the grid-connected energy storage points and generate grid strength state quantities based on the grid information, which includes voltage, current and frequency information.

[0072] like Figure 2 As shown, the specific steps of step S1 are as follows:

[0073] Step S101: Collect the synchronous sampling sequence of three-phase voltage and three-phase current at the grid connection point of the grid-connected energy storage system, and perform time alignment on the sampling sequence to form grid information under the same time reference.

[0074] In this embodiment, the three-phase voltage and three-phase current at the grid-connected point of the grid-storage energy storage are synchronously sampled to obtain the instantaneous amplitude data at a unified time. Specifically, synchronous sampling can rely on the same time base clock source or a unified trigger pulse to collect voltage sampling channels and current sampling channels connected to the same sampling clock in parallel, so that the original discrete sequences of three-phase voltage and three-phase current are obtained simultaneously at each sampling moment. It should be noted that, considering that there may be time offsets in the actual sampling link due to sampling channel delay, trigger jitter or data buffering, after reading the original sampling sequence, a sampling time mark is added to the data of each channel, and the sampling points of different channels and different phases are time aligned based on the sampling time mark. For example, by interpolating or resampling between adjacent sampling points, the slightly offset sampling points are remapped to a unified time grid, thereby constructing a power grid information dataset under the same time base.

[0075] Step S102: The power grid information is segmented, and a set of windows including steady-state windows and transition windows is generated according to a preset time window. A corresponding subset of power grid information is established for each window.

[0076] In this embodiment, based on the aforementioned unified time reference, the power grid information is processed in the time domain to form a windowed data structure that can be used to distinguish different operating state characteristics. Specifically, firstly, the window length and window sliding interval are set according to the grid energy storage control cycle, the dynamic change rate of the power grid, and the identification accuracy requirements, so that the power grid information is continuously divided into several interrelated time segments on the time axis. Subsequently, by analyzing the amplitude fluctuation, frequency change trend, and phasor stability of the voltage and current sequences in each time segment, the power grid operating behavior corresponding to the time segment is inferred. Thus, time segments with approximately stable operating states and slow parameter changes are classified as steady-state windows, while time segments with parameter mutations, disturbance responses, or control switching signs are classified as transition windows.

[0077] Step S103: Within the transition window, a preset perturbation sequence is superimposed, and the voltage and current responses during the operation of the perturbation sequence are recorded synchronously to form identification data pairs.

[0078] In this embodiment, for the time interval determined as a transition window, without changing the overall output trend of grid-connected energy storage, a perturbation sequence with limited amplitude, controlled duration, and known frequency band characteristics is superimposed on its control variables, so that the grid obtains an identifiable controlled excitation input within this window. Specifically, the perturbation sequence can be superimposed on the grid voltage reference, virtual impedance reference, or other grid control variables, so that the voltage and current at the grid connection point generate response components that can be distinguished from natural fluctuations based on the original change trajectory. It should be noted that during the entire process of superimposing the perturbation sequence, the voltage response and current response are synchronously acquired, and time stamps are used to ensure that they correspond to the same perturbation action state under the same time reference. Subsequently, the perturbation input information at each moment is paired with the synchronously sampled voltage and current response data to form an identification data pair with a clear input-response correlation.

[0079] Step S104: Based on the identification data, generate a set of candidate equivalent impedances in multiple frequency bands or multiple phase sequence dimensions, and perform consistency screening on the set of candidate equivalent impedances to obtain the target equivalent impedance parameters.

[0080] The specific steps of step S104 are as follows:

[0081] Step S1041: Divide the identification data into multiple data subsets according to a preset dimension, and establish a corresponding identification task identifier for each data subset.

[0082] In this embodiment, the preset dimension is set based on parameters such as frequency band interval, phase sequence attribute, time segment label or disturbance type identifier.

[0083] Step S1042: Based on the identification task identifier, extract the associated sample pairs of voltage and current in the corresponding data subset, and generate multiple candidate equivalent impedance parameters based on the associated sample pairs, and collect them to form a candidate equivalent impedance set.

[0084] In this embodiment, in each data subset, firstly, based on the analysis dimension indicated by the identification task identifier, a set of sample points in the voltage response sequence and current response sequence that are at the same time reference, the same disturbance stage, or the same statistical interval is screened out from the data subset. Voltage samples and current samples with corresponding relationships are extracted in pairs to form voltage-current associated sample pairs. Subsequently, each associated sample pair is used as an independent identification input, and its corresponding equivalent impedance parameter is derived according to a preset identification path, so that multiple candidate impedance parameter values ​​from different sources but logically independent can be obtained within a single data subset.

[0085] Step S1043: Perform consistency screening on the candidate equivalent impedance set, compare the candidate equivalent impedance parameters pairwise according to the preset comparison rules, and generate a set of consistent results.

[0086] In this embodiment, each candidate equivalent impedance parameter is assigned a comparison label based on its source dimension, the data window type involved in the identification, and the generation path. Then, according to pre-defined comparison rules (including amplitude consistency rules, phase consistency rules, and stability consistency rules), the candidate equivalent impedance parameters are paired to form a set of parameter pairs requiring consistency verification. Subsequently, for each parameter pair, based on the impedance amplitude change relationship, impedance phase change relationship, and stability degree as the identification dimension changes, a difference measurement is performed to infer whether the parameter pair satisfies the condition of being considered an equivalent description under the same power grid operating conditions. It should be noted that by recording the comparison results of all parameter pairs one by one, a unified set of consistency results is formed.

[0087] Step S1044: The remaining candidate equivalent impedance parameters after consistency screening are fused to output the target equivalent impedance parameter, wherein the fusion includes selection by frequency band priority, selection by phase sequence dimension priority, or selection by preset combination rules.

[0088] In this embodiment, frequency band priorities are set based on the differences in sensitivity of grid-connected energy storage control to different frequency bands. This allows candidate equivalent impedance parameters that are within the critical frequency band and have high consistency to be given priority in fusion. Simultaneously, different participation weights are assigned to different phase sequence dimensions based on three-phase imbalance, phase sequence coupling degree, or grid structure characteristics, thus forming a fusion framework that combines frequency band weights and phase sequence weights. Subsequently, under this fusion framework, the remaining candidate parameters are selected hierarchically: when the target result focuses more on describing the grid's response behavior in a specific frequency domain, the output is optimized according to frequency band priority; when it is necessary to reflect three-phase coupling and phase sequence consistency, the output can be optimized according to phase sequence dimension priority; when it is necessary to integrate both frequency domain information and phase sequence balance, multiple candidate parameters are systematically integrated through preset combination rules to generate a single target equivalent impedance parameter that can serve as the basis for subsequent control. The preset combination rules are used to uniformly integrate multiple candidate equivalent impedance parameters in a weighted convergence or primary-secondary superposition manner under the joint constraints of frequency band weights and phase sequence weights.

[0089] Step S105: The target equivalent impedance parameter is fused with the voltage, current and frequency information in the steady-state window, and the power grid strength state quantity is output as the window is updated.

[0090] The specific steps of step S105 are as follows:

[0091] Step S1051: Within the steady-state window, generate window feature sets corresponding to voltage, current, and frequency information respectively, and assign a timestamp identifier corresponding to the steady-state window to the window feature sets. The window feature sets include voltage amplitude features, current amplitude features, and frequency offset features.

[0092] In this embodiment, voltage data is first statistically analyzed within a steady-state window. By analyzing the amplitude distribution, fluctuation boundaries, and stable intervals of voltage sampling points within the window, voltage amplitude features are extracted to characterize the voltage level. Subsequently, the current sequence is processed using the same window boundary. By evaluating the magnitude, relative concentration range, and trend of current amplitude within the window, corresponding current amplitude features are formed. Simultaneously, based on the frequency sampling information within the steady-state window, the degree of frequency deviation relative to the reference frequency during this time period is analyzed, and frequency deviation features are extracted to reflect the operating state of the power grid frequency under this window. It should be noted that after completing the above feature extraction, the formed voltage amplitude features, current amplitude features, and frequency deviation features are uniformly collected into a window feature set and a timestamp identifier corresponding to the steady-state window is attached.

[0093] Step S1052: Align the target equivalent impedance parameter with the window feature set to form a fused input set. The alignment includes mapping the target equivalent impedance parameter to a window index that is consistent with the timestamp identifier, and replacing the target equivalent impedance parameter at the missing index with a preset placeholder parameter or backtracking parameter.

[0094] In this embodiment, firstly, based on the division of steady-state windows on the time axis, a window index system is established with window timestamps as unique identifiers, ensuring that each window feature set has a clear time index position. Subsequently, using the generation time of the target equivalent impedance parameter or the time label of the corresponding transition window as a reference, the target equivalent impedance parameter is mapped to the window index position closest to or directly corresponding to its time attribute, thus anchoring the position of the target equivalent impedance parameter within the same window sequence. It should be noted that in some windows, the target equivalent impedance parameter may be missing due to non-identification or incomplete identification. In this case, the window is not directly discarded, but rather, according to preset rules, a placeholder parameter is introduced in the form of a fixed value, or the target equivalent impedance parameter already confirmed in a temporally adjacent window is used for backtracking and filling, ensuring that each valid window index has bidirectional information on impedance parameters and window features. Finally, through the above mapping, completion, and structural alignment processes, the target equivalent impedance parameter and the voltage amplitude characteristics, current amplitude characteristics, and frequency offset characteristics of each steady-state window are incorporated into a unified data framework, forming a fused input set with temporal homogeneity and structural consistency.

[0095] Step S1053: Generate a set of candidate state quantities for power grid strength based on the fused input set. The set of candidate state quantities for power grid strength includes a first candidate state quantity generated primarily by the target equivalent impedance parameter, a second candidate state quantity generated primarily by the window feature set, and a third candidate state quantity generated by combining the target equivalent impedance parameter and the window feature set according to a preset rule.

[0096] In this embodiment, in the first layer of processing, the target equivalent impedance parameter is used as the main reference quantity. The focus is on the amplitude, phase, and relative stability of the grid's equivalent impedance over time under different windows, thereby generating a first candidate state quantity that reflects the differences in the grid's equivalent support capability. In the second layer of processing, voltage amplitude characteristics, current amplitude characteristics, and frequency deviation characteristics are the primary factors. By analyzing the correlation between voltage levels, current load levels, and frequency deviation characteristics within a window, a second candidate state quantity is constructed to characterize the strength trend of the grid based on operational characteristics. Furthermore, in the third layer of processing, impedance information and window characteristics are introduced... The joint evaluation concept of power grid strength information, through preset combination rules, integrates the target equivalent impedance parameter with the window feature set, so that impedance characteristics and steady-state operation indicators form a synergistic constraint within the same evaluation system, thereby generating a third candidate state quantity that reflects both structural equivalence characteristics and operational performance. It should be noted that through the hierarchical construction of the above three types of candidate state quantities, the representation of power grid strength is not limited to a single data source, but forms a multi-candidate expression system that exists in parallel at the impedance layer, operational feature layer, and combination layer, ultimately forming a set of candidate state quantities for power grid strength with sufficient information dimensions and differentiated reference basis.

[0097] Step S1054: During the window rolling update process, select or switch the set of candidate power grid strength states and output the power grid strength states that are updated with the window rolling.

[0098] In this embodiment, when each window arrives, the candidate state quantity category with priority interpretation capability is first determined from the first, second, and third candidate state quantities based on the operating scenario attributes corresponding to the window, the source type of candidate state quantities, and the correlation between the output state quantities of the previous window. Then, by comparing the changing trends of different candidate values ​​within the category with the historical evolution trajectory of the window sequence, the candidate value that best matches the current window's operating state is selected as the grid strength output reference for this window. It should be noted that during the window rolling process, if a sudden change in the grid operating state is detected, the difference between candidate state quantities increases significantly, or the priority interpretation logic shifts, a candidate category switching operation is performed, replacing the current category with another candidate category that is more suitable for the current operating condition. At the same time, the switching boundary is smoothed to avoid discontinuity in the output transition between windows. Finally, through the joint processing of the above candidate screening, priority judgment, and necessary switching, the grid strength state quantity can be continuously updated as the window rolls and always maintain a descriptive capability consistent with the current grid operating behavior, thereby forming a grid strength state quantity that is rolled out over time.

[0099] Combination Figure 2 This paper describes the process from four stages: data acquisition, identification and screening, steady-state fusion, and output of grid strength state parameters. In the data acquisition stage, sampling devices installed at the grid-connected energy storage points synchronously acquire the three-phase voltage and current signals of the power grid. Based on the time axis, the continuously sampled grid information is divided into two data intervals: a steady-state window and a transition window. The steady-state window is mainly used for steady-state feature extraction, while the transition window is mainly used for identifying the equivalent impedance of the power grid. Within the transition window, a preset disturbance signal is superimposed, and the voltage and current response signals during the disturbance are recorded synchronously, thus forming identification data with input-response correlation.

[0100] During the identification and screening phase, the equivalent impedance of the power grid is calculated based on the identification data, resulting in a set of multiple candidate equivalent impedance parameters, such as... Figure 2 The Z1, Z2, ..., Z marked in the middle n As shown, Z nThe nth equivalent impedance parameter is represented. Consistency screening is then performed on each candidate equivalent impedance parameter. By comparing the performance of different candidate parameters across multiple frequency bands or phase sequence dimensions, abnormal candidate values ​​that do not meet the consistency conditions are eliminated, retaining only parameters with stable consistency characteristics. The target equivalent impedance parameter is then determined accordingly. Subsequently, in the steady-state fusion stage, voltage amplitude characteristics, current amplitude characteristics, and frequency offset characteristics are extracted within the steady-state window. This feature information is then fused with the target equivalent impedance parameter to form three categories of candidate state variables: the first candidate state variable (impedance-dominated), the second candidate state variable (operation measurement-dominated), and the third candidate state variable (combined fusion-dominated). Through comprehensive judgment of the candidate state variable set, a comprehensive characterization of the grid's support capacity, grid coupling strength, and operational robustness is achieved. Finally, in the grid strength state variable output stage, based on the rolling changes of the candidate state variables in different time windows, a suitable candidate state variable is selected or switched as the final output result, forming a grid strength state variable that can be dynamically updated over time and displayed using a weak-to-strong quantization scale.

[0101] Step S2: Based on the power grid strength state quantity, adaptively adjust the grid construction control parameters, which include virtual inertia, virtual impedance, and power control coefficient.

[0102] like Figure 3 As shown, the specific steps of step S2 are as follows:

[0103] Step S201: Obtain the power grid strength status quantity corresponding to the current window, and map the power grid strength status quantity to the target strength level in the preset strength level set, wherein the strength level set includes weak grid level and strong grid level.

[0104] In this embodiment, the grid strength state quantity corresponding to the current window is first read when the current window arrives. This state quantity comprehensively reflects information such as equivalent impedance characteristics, steady-state voltage and current levels, and frequency operation deviation. Its magnitude and trend can characterize the relative strength of the grid support capacity and the grid coupling stiffness. Subsequently, a strength mapping rule is constructed based on a pre-established set of strength levels. The grid strength state quantity is compared with the threshold intervals in the set of levels one by one. When the state quantity is in a low strength interval or exhibits weak stiffness characteristics, it is mapped to a weak grid level. When the state quantity is in a high strength interval or exhibits strong coupling characteristics, it is mapped to a strong grid level. The strong and weak grid levels are set according to the actual equivalent impedance. It should be noted that this mapping process not only relies on the strength value at a single moment but also combines the continuous changes of the window within a short time range to avoid misjudgment of the level due to instantaneous fluctuations. Finally, a target strength level with clear semantic indication is formed in the current window.

[0105] Step S202: Based on the target intensity level, select the parameter strategy entry corresponding to the target intensity level from the preset parameter strategy library, and generate a parameter update task identifier for the parameter strategy entry. The preset parameter strategy library includes virtual inertia parameter strategy, virtual impedance parameter strategy and power control coefficient strategy.

[0106] In this embodiment, firstly, based on the power grid operation characteristics corresponding to the target strength level, the key control focus of the network control under this window is determined. For example, under a weak network level, the focus is on enhancing support capabilities, while under a strong network level, the focus is on improving response flexibility. Accordingly, a set of strategy entries matching the level is selected from the parameter strategy library. The parameter strategy library is divided into a set of virtual inertia parameter strategies, a set of virtual impedance parameter strategies, and a set of power control coefficient strategies according to the control mechanism. Different strategy entries provide parameter value ranges, adjustment directions, and applicable conditions, so that the same target strength level can be mapped to multiple control parameter adjustment paths with clear structures. It should be noted that after the strategy entries are selected, the control parameters are not changed immediately. Instead, a corresponding parameter update task identifier is generated for each selected strategy entry to clarify the independence, execution order, and participation method of each strategy entry in the subsequent update process. This realizes that under the logical chain of level judgment-strategy selection-task generation, the abstract strength level conclusion is transformed into a set of parameter update tasks that can be identified and scheduled by the control execution unit.

[0107] Step S203: Decouple and sort the parameter update task identifiers, and form a parameter update sequence according to a preset order. The preset order includes updating the virtual impedance first, then updating the virtual inertia, and finally updating the power control coefficient.

[0108] The specific steps of step S203 are as follows:

[0109] Step S2031: Classify and mark the update task identifiers of each parameter, assign them to different parameter groups, and assign a group priority identifier to each parameter group.

[0110] In this embodiment, firstly, based on the control object attributes associated with each parameter update task identifier, they are categorized and labeled according to their mechanism of action and control level. For example, task identifiers related to grid impedance matching are categorized into the virtual impedance parameter group, task identifiers related to inertia and damping adjustment are categorized into the virtual inertia parameter group, and task identifiers related to power distribution and response sensitivity adjustment are categorized into the power control parameter group, so that each task identifier has a clear group affiliation. Subsequently, based on the logical relationship of the impact of different group parameters on grid construction behavior and the degree of safety sensitivity, a group priority identifier is set for each parameter group, so that parameter updates have a schedulable execution order at the macro level.

[0111] Step S2032: Based on the mutual influence relationship between parameter update task identifiers within each parameter group, establish a sequence constraint linked list for any two parameter update task identifiers, and record directly related parameter update task identifiers as task pairs with sequential constraints, thus forming a parameter update sequence constraint set.

[0112] In this embodiment, firstly, for the parameter update task identifier within each parameter group, based on the position of the corresponding control parameter in the network control link, the coupling path, and the degree of influence of the execution result on other parameters, it is determined whether there is a dependency relationship between any two tasks that need to be executed sequentially. For example, some parameter update tasks used for stable control foundation support need to be executed before parameter update tasks used for performance optimization. Subsequently, when it is determined that there is a direct relationship, the two task identifiers are recorded in the form of predecessor task-successor task, and a sequential constraint linked list is constructed to describe the relationship between the two, so that each constrained task clearly knows its predecessor task that must be completed. On this basis, by traversing each parameter group, all task pairs with sequential dependencies are uniformly aggregated to form a parameter update sequence constraint set, so that the originally independent multiple parameter update tasks form a constrained execution network at the logical level.

[0113] Step S2033: According to the group priority identifier and the parameter update order constraint set, perform hierarchical sorting on all parameter update task identifiers: first, determine the update order of each parameter group at the group level, and then sort the parameter update task identifiers locally within the group according to the order constraint linked list to generate an intermediate parameter update sequence.

[0114] In this embodiment, the group priority identifier is first used as the top-level constraint to determine the macro-execution order between different groups within the entire parameter group scope. Parameter groups with higher priority and more significant impact on network control are placed first, while parameter groups used for fine-tuning or performance compensation are placed later. This establishes a framework for the order of parameter updates at the group level. Subsequently, within each parameter group with a determined order, instead of simply using a fixed order, the dependency relationship of all parameter update task identifiers within the group is checked one by one based on the aforementioned parameter update order constraint set. According to the predecessor-successor task relationship recorded in the order constraint linked list, task identifiers with direct or indirect dependencies are arranged in sequence, forming a local sorting structure within the group that meets the logical order requirements. It should be noted that after completing the inter-group sorting and intra-group local sorting, the intermediate parameter update sequence in an overall sense can be obtained by unfolding each parameter group according to its priority order and splicing its corresponding local sorting results.

[0115] Step S2034: Perform consistency adjustment on the intermediate parameter update sequence, rearrange the local sorting segments that do not meet the group priority identifier or order constraint list, and obtain a parameter update sequence that meets the preset order relationship.

[0116] In this embodiment, the intermediate parameter update sequence is first scanned using the parameter grouping priority identifier as the top-level constraint to identify whether there are sorted segments where low-priority grouping parameters are inserted before high-priority grouping parameters. If so, the segment is marked as a segment to be reordered. Subsequently, the parameter update order constraint list is used as the second constraint to further check the consistency of the preceding and following relationships within the intermediate sequence, determining whether there are sorted segments that violate the preceding-following task dependency chain, and such segments are also included in the reordering scope. Based on this, all segments to be reordered are reconstructed according to the principle of priority constraint first and order constraint second. By adjusting the internal position of the segment and its connection with the preceding and following sequences, it is made to satisfy the predetermined grouping priority relationship and order dependency relationship again. It should be noted that after the above correction is completed, a global consistency verification is performed on the updated overall sequence to confirm that there are no new conflicts or hidden dependency violations. Finally, a parameter update sequence with continuous structure, logical consistency and complete conformity to the preset order relationship is output.

[0117] Step S204: Based on each parameter in the parameter update sequence, generate corresponding parameter transition trajectories and write the parameter transition trajectories into the parameter register set of the network controller, wherein the parameter transition trajectories include segmented incremental trajectories, rolling hold trajectories or event-triggered trajectories.

[0118] The specific steps of step S204 are as follows:

[0119] Step S2041: Read the parameter values ​​and target parameter values ​​of each parameter in the parameter update sequence to form parameter state pairs, and group them according to the virtual impedance, virtual inertia or power control coefficient to which the parameters belong, and add trajectory type labels to each parameter state pair.

[0120] In this embodiment, after the parameter state pair is formed, it is not regarded as a homogeneous update object. Instead, the parameter state pair is further grouped and associated according to the functional category to which the parameter belongs in the aforementioned parameter grouping stage. Parameter state pairs belonging to the virtual impedance class, parameter state pairs belonging to the virtual inertia class, and parameter state pairs belonging to the power control coefficient class are respectively attached with different trajectory type tags, so that parameters of different categories have differentiated evolution path indications in the subsequent trajectory generation stage.

[0121] Step S2042: Based on the trajectory type marker, select a preset trajectory generation rule for each parameter state pair to generate a parameter transition trajectory template.

[0122] In this embodiment, the current parameter value and target parameter value, as well as the trajectory type marker attached to each parameter state pair, are first read. By querying the preset trajectory rule library, trajectory generation rules that focus more on gradual adjustment and coupling balance are selected for virtual impedance parameters, trajectory generation rules that reflect the gradual reconstruction characteristics of control inertia are selected for virtual inertia parameters, and trajectory generation rules that reflect the gradual release or convergence characteristics of response sensitivity are selected for power control coefficient parameters. Subsequently, under the action of the corresponding rules, the change path form, discrete update step size allocation method, and trajectory smoothing constraint conditions adopted by the parameter in the transition from the current value to the target value are determined. Thus, a parameter transition trajectory template that can completely describe the parameter update behavior is constructed at the logical level.

[0123] Step S2043: According to the order of the parameter update sequence, expand each parameter transition trajectory template into a corresponding discrete parameter sample sequence, and associate each discrete parameter sample sequence with the control cycle number or time step number to form a parameter transition trajectory set.

[0124] In this embodiment, the parameter update sequence is used as a guide. The corresponding parameter transition trajectory templates are read sequentially according to the order of the parameters in the sequence. Based on the definitions of the change path form and update step size allocation method in the trajectory template, the trajectory expression in the sense of continuous description is gradually expanded into a sample sequence composed of multiple discrete parameter samples. This makes the process of parameter evolution over time clearly decomposed into discrete update nodes that can be called on a cycle by cycle. Subsequently, after obtaining the discrete parameter sample sequence of each parameter, the sample value itself is not simply retained. Instead, each discrete sample is further bound to its effective time in the control process by using the control cycle number or time step number as an index. This makes the parameter update not only have the evolution order in the numerical dimension, but also the activation order in the time axis dimension. It should be noted that by collecting the discrete sample sequences corresponding to all parameters in a unified structure, a parameter transition trajectory set for multiple parameters and multiple control cycles is formed.

[0125] Step S2044: Write the parameter transition trajectory set into the parameter register set of the network controller.

[0126] Step S205: When the next window arrives, verify the correspondence between the power grid strength state quantity and the written parameter register set. If the correspondence does not meet the preset consistency condition, backtrack to select parameter strategy entries and rebuild the parameter update sequence to complete the adaptive adjustment of the network control parameters. The network control parameters include virtual inertia, virtual impedance and power control coefficient.

[0127] In this embodiment, when a new window is started, the grid strength state quantity corresponding to that window is first obtained and compared with the target direction of the virtual inertia, virtual impedance, and power control coefficient currently written into the grid controller parameter register set to determine whether the current parameter trajectory still maintains a logical matching relationship with the latest grid strength state. Subsequently, based on the pre-set consistency conditions, if the comparison result indicates that the parameter evolution trend deviates from the direction of grid strength change, or the parameter has been updated to a certain stage but the grid strength state has reversed, the strategy backtracking process is triggered. The strategy entry matching the current grid strength state quantity is retrieved again in the parameter strategy library, and the parameter grouping, update order, and trajectory generation process are reconstructed accordingly to form a new parameter update sequence. It should be noted that through the above-mentioned cyclic verification and backtracking reconstruction, the grid control parameters are not passively maintained after being set once, but are continuously constrained and corrected by the grid strength state quantity during the window scrolling process. This logically establishes a closed-loop adaptive adjustment path of state-driven - parameter evolution - consistency verification - necessary backtracking, and finally realizes the control objective of dynamically matching and adjusting the virtual inertia parameter, virtual impedance parameter, and power control coefficient with the change of grid strength.

[0128] The preset consistency condition is used to determine whether the adjustment direction, update stage and effect intensity of the current grid control parameters are consistent with the trend of strength change and control requirements indicated by the latest grid strength state quantity. If the two are mismatched in trend, stage or target, they are considered not to meet the consistency condition.

[0129] Combination Figure 3 The adaptive adjustment process of network control parameters is described in three stages: parameter update sequence determination, parameter transition trajectory generation, and parameter register writing. First, based on the parameter update sequence formed in steps S203-S204, the network control parameters that need adjustment are determined within the current window and arranged in a predetermined order. The parameter update sequence includes at least three types of parameters: virtual impedance parameters, virtual inertia parameters, and power control coefficients. Second, based on each parameter in the parameter update sequence, a corresponding parameter transition trajectory is generated. Specifically, for each parameter to be updated, based on its current and target parameter values, a path for parameter change on the time axis is generated according to a preset trajectory generation rule. This can form different forms of trajectory expression, such as segmented gradual trajectory, rolling hold trajectory, or event-triggered trajectory, ensuring that the parameter value does not switch abruptly but gradually transitions from the initial value to the target value in a controllable manner. Figure 3 The trajectory process of parameters gradually adjusting over time is represented by a stepped schematic curve.

[0130] Finally, the parameter transition trajectory corresponding to each parameter is expanded into a discrete parameter sample sequence and bound to the control cycle number or time step number to form a complete parameter update sequence data structure. This parameter update sequence is then uniformly written into the parameter register set in the network controller.

[0131] Step S3: Construct stable operation constraints for grid-connected energy storage based on the grid strength state variables, and generate active and / or reactive power output commands for grid-connected energy storage under the stable operation constraints, and adaptively adjust the output power of grid-connected energy storage according to changes in grid strength.

[0132] The specific steps of step S3 are as follows:

[0133] Step S301: Obtain the power grid strength status quantity corresponding to the current window, and map the power grid strength status quantity to the target operating mode identifier in the preset operating mode set. The preset operating mode set includes weak grid mode, transition mode and strong grid mode.

[0134] In this embodiment, when the current window is started, the power grid strength state quantity corresponding to that window is first read. This state quantity comprehensively reflects the equivalent impedance of the power grid, the level of support capability, the steady-state voltage and current characteristics, and the frequency operation deviation. Through its numerical range, variation amplitude, and evolution trend between adjacent windows, it can be inferred that the current power grid is closer to a weak grid environment, is in a dynamic transition stage, or exhibits a strong grid support structure. Subsequently, based on the pre-constructed set of operating modes, mapping rules are established to match the power grid strength state quantity with the weak grid mode interval, the transition mode interval, and the strong grid mode interval in sequence. When the state quantity exhibits low stiffness, weak coupling, or insufficient support capability, it is mapped to a weak grid mode. When the state quantity is in the critical interval and shows a clear evolution trend, it is mapped to a transition mode. When the state quantity is in the high support and high stability interval, it is mapped to a strong grid mode. It should be noted that the mapping process not only considers the static value of a single window but also combines the continuity of changes of multiple consecutive windows to avoid misjudgment caused by instantaneous disturbances. Finally, the corresponding target operating mode identifier is output for the window.

[0135] Step S302: Based on the target operating mode identifier, select the corresponding stable operating constraint template from the constraint template library, and scale or translate the stable operating constraint template based on the current power grid strength state quantity to generate a stable operating constraint set.

[0136] In this embodiment, the constraint template category is first identified based on the target operating mode. For example, the template corresponding to the weak grid mode is more inclined to strengthen voltage support and vibration suppression requirements, while the template corresponding to the strong grid mode is more inclined to improve dynamic response capability and power regulation freedom. The transition mode template focuses on establishing a transition constraint structure between the two. Subsequently, after determining the template category, it is not directly used as the final constraint boundary. Instead, the current grid strength state quantity is used as a measurement reference to scale or shift the active power output boundary, reactive power output boundary, and coupling restriction between active and reactive power in the template. This allows the constraint boundary to be appropriately widened as the grid strength increases and appropriately tightened as the grid strength decreases, thereby forming a constraint amplitude consistent with the current grid strength. It should be noted that through the above template selection-strength-driven adjustment method, the stable operation constraint is transformed from a fixed static boundary into a constraint set that can be dynamically adapted to the operating mode and grid strength.

[0137] Step S303: Collect the expected active power and expected reactive power from the outside, and combine the expected active power and expected reactive power with the set of stable operation constraints to form a power expectation determination result that is inside or outside the set of stable operation constraints.

[0138] In this embodiment, within the current window, the expected active power and reactive power quantities from the system scheduling layer, load-side demand, or upper-level control commands are first collected or received, and both are used as references for the expected output of grid-connected energy storage in this window. Subsequently, using the stable operation constraint set generated in step S302 as the judgment benchmark, the expected active power and reactive power quantities are mapped to the same power point in the constraint set. It is determined whether the power point falls within the allowable area defined by the constraint set. When the power point is inside or near the boundary of the allowable area, it is determined that it belongs to the directly executable power expectation. When the power point is outside the allowable area, it is determined that it exceeds the output range that grid-connected energy storage can stably undertake under the current grid strength and operation conditions. It should be noted that through the above judgment process, the external power command is transformed from a simple demand quantity into an executable judgment result related to stable operation capability, and finally a power expectation judgment result that clearly distinguishes between those inside and outside the constraints is formed.

[0139] Step S304: After power correction of the power expectation determination result, project it onto the boundary or interior of the stable operation constraint set to generate constrained active power output command and / or reactive power output command.

[0140] The specific steps of step S304 are as follows:

[0141] Step S3041: Obtain the power expectation determination result after the determination of the stable operation constraint set, mark the active power expectation and / or reactive power expectation outside the stable operation constraint set as the power expectation to be corrected, and assign a projection priority identifier to each power expectation to be corrected according to the preset strategy of active power priority or reactive power priority.

[0142] In this embodiment, the power expectation determination result obtained by joint determination is first read in the current window. The power expectations located inside the stable operation constraint set and the power expectations located outside the stable operation constraint set are identified. The active power expectation and / or reactive power expectation corresponding to the latter are extracted separately and marked as power expectations to be corrected, which are the mandatory processing objects for subsequent correction calculations. Subsequently, considering that the importance of active output and reactive power support differs under different operating scenarios, a power adjustment strategy with active power priority or reactive power priority is pre-set according to the system operation strategy or network control design principle. This strategy is applied to each power expectation to be corrected, and different projection priority labels are assigned to different power components, so that the power components with high priority are given priority in obtaining the degree of satisfaction or approaching the target boundary first in the subsequent projection and correction process.

[0143] Step S3042: Based on the projection priority identifier, select the corresponding projection rule set and determine the projection direction of the expected power to be corrected. The projection direction includes the active power axis direction, the reactive power axis direction, or the combined active and reactive power direction.

[0144] In this embodiment, firstly, based on the projection priority identifier set for each power expectation to be corrected in step S3041, the projection rule set corresponding to the priority in the preset projection rule base is called. This makes the projection rule corresponding to the active power priority strategy more inclined to converge along the active power axis, and the projection rule corresponding to the reactive power priority strategy more inclined to converge along the reactive power axis. In the scenario where both active and reactive power need to be adjusted in a coordinated manner, a joint projection rule set that reflects the joint correction characteristics of the two is selected. Subsequently, after selecting the projection rule set, the spatial direction of the current power expectation to be corrected is judged according to its structured constraint definition to determine its most matching projection direction, thus forming a projection path along the active power axis, along the reactive power axis, or along the joint direction of active and reactive power. It should be noted that this method realizes the transition from the power over-limit fact to the directional correction based on priority and rule constraints, so that the power correction process is no longer a simple truncation in a single fixed direction, but forms a projection direction selection result with purpose and interpretability under rule control.

[0145] Step S3043: Search for power points that satisfy preset constraints along the projection direction at the boundary or inside of the stable operation constraint set, record the power points as a candidate projection power point set, and select the target projection power point from the candidate projection power point set according to the preset sorting rules.

[0146] In this embodiment, after determining the projection direction, the original power expectation point to be corrected is used as the starting point, and the projection proceeds gradually towards the boundary or interior of the stable operation constraint set. During the process, each candidate power point is checked to see if it simultaneously meets the preset constraints in the constraint set, such as active power output amplitude limits, reactive power support capacity limits, and active-reactive power coupling limits. Power points that meet the constraints are recorded one by one to form a structured candidate projection power point set. Subsequently, after the candidate projection power point set is constructed, no point that meets the constraints is directly used as the output. Instead, the set is optimized according to a preset sorting rule. The sorting rule can be based on factors such as the distance relationship with the original power expectation point, the priority of satisfying different constraints, or the degree of matching with the operating mode to comprehensively evaluate and select the more representative target projection power point that is more in line with the current operating strategy requirements from the candidate set. It should be noted that the preset sorting rule is used to comprehensively sort the candidate projection power points in the set according to the minimum deviation from the original power expectation point, the priority of satisfying key constraints, and the degree of matching with the operating strategy, so as to prioritize the power point with the smallest correction amplitude of the original expectation and the highest operating rationality under the premise of ensuring compliance with stable constraints.

[0147] Step S3044: Output the active power component and / or reactive power component of the target projected power point as constrained active power output command or reactive power output command, respectively.

[0148] In this embodiment, after the target projected power point is determined, the active and reactive components corresponding to the power point are first analyzed and separated. The obtained active and reactive power values ​​represent the acceptable power output level under the current grid strength conditions and stable operation constraint boundaries. Subsequently, the active component of the target projected power point is used as the constrained active power output command, and its reactive component is used as the constrained reactive power output command. These commands are written into the grid-based energy storage control command chain, so that the subsequent control module can directly use these commands as reference values ​​when performing power allocation and output adjustment. It should be noted that by explicitly mapping the target power point after power expectation correction to independent active and reactive power output commands, not only is the correspondence between the power correction result and the stable operation constraint set maintained, but also a continuous transition from expected power to corrected power to control command is achieved in terms of control execution semantics. This allows the power output of grid-based energy storage within the current window to be determined in a structured and executable form.

[0149] Step S305: During the window scrolling process, when a change in grid strength is detected, the stable operation constraint template corresponding to the new target operation mode identifier is invoked, and the previously generated power output command is smoothly transitioned to form an active power output command and / or reactive power output command that changes with the grid strength.

[0150] In this embodiment, during the window scrolling update, the grid strength state quantity corresponding to the new window is continuously monitored, and the operating mode is determined in real time according to the mapping rules of step S301 to determine whether the operating mode has transitioned from a weak grid mode to a transitional mode or a strong grid mode, or has retreated from a strong grid mode to the weak grid side. When a difference is detected between the target operating mode identifier and the previous window, the constraint template switching process is triggered. The stable operating constraint template corresponding to the new target operating mode identifier is re-called from the constraint template library, and a new constraint scaling or translation is completed in combination with the current grid strength state quantity to form an updated stable operating constraint set. Subsequently, the newly formed stable operating constraint set is used as the power correction... The new benchmark performs continuity determination on active power output commands and / or reactive power output commands that have been generated in the previous window and are still in execution. By introducing a transition buffer or a phased correction method, the original commands are gradually guided to the target power region that meets the new constraint set, avoiding abrupt jumps in commands at the moment of form switching. It should be noted that through the above-mentioned processing chain of form switching-constraint replacement-command smooth transition, the power output exhibits control characteristics that are synchronously reconstructed with changes in grid strength and have temporal continuity during the evolution of grid strength, ultimately forming active power output commands and / or reactive power output commands that can be dynamically adjusted with changes in grid strength.

[0151] Step S4: When a sudden change in the power grid strength state is detected, the grid construction control reconfiguration is triggered, and the grid construction control parameters are readjusted.

[0152] In this embodiment, the time-series changes of the grid strength state are continuously monitored during the window rolling update process. When a jump amplitude exceeding the preset abrupt change threshold is detected between adjacent windows, or when multiple strength levels or operating mode intervals are continuously crossed within a very short period of time, it is determined that the current grid has entered an abrupt change state from the normal evolution state. At this time, the grid construction control reconfiguration process is triggered. In this process, the current grid construction control parameters are first frozen or protected by gradual change to prevent the parameters from being further disturbed while they are still executing the transition trajectory. Then, the strategy is re-evaluated based on the latest grid strength state. The virtual inertia parameter strategy, virtual impedance parameter strategy, and power control coefficient strategy adapted to the new strength environment are reselected from the parameter strategy library, and the parameter grouping, sequence constraints, and parameter trajectories are reconstructed so that the grid construction control parameters are logically returned to the starting point for replanning. It should be noted that through this abrupt change triggering-reconfiguration adjustment design, when facing extreme operating conditions or rapidly deteriorating grid environments, grid construction energy storage no longer relies on the original gradual adaptive process, but can directly enter a rapid re-adaptation mode centered on the new state, thereby forming a self-healing grid construction regulation capability to cope with abrupt changes at the control mechanism level.

[0153] For example, the following numerical example is provided to illustrate the calculation process of how to readjust the network control parameters. This example is only used to explain the relationship between the calculation link and the dimensions. The selected parameters and values ​​are illustrative and do not represent the actual calibration results or engineering recommended values.

[0154] Parameter settings: Window and sampling: Window length T = 1s; Calculate the grid strength state quantity at the end of each window (or when the next window arrives); Grid strength state quantity: Grid strength state quantity of the previous window. =3.0; Current window grid strength state quantity =2.2; Change = - =-0.8 indicates a decrease in grid strength; the grid control parameters that need to be adaptively adjusted are: virtual inertia H (unit: s), virtual impedance Z (unit: Ω), and power control coefficient K (unit: kW / Hz); the parameter safety boundaries are: virtual inertia [2.0, 6.0]s, virtual impedance [0.10, 0.35]Ω, and power control coefficient [80, 160]kW / Hz.

[0155] The previous window has written register sets and policy entries: The parameter register sets written in the previous window , This represents the virtual inertia of the parameter register set already written in the previous window. This represents the virtual impedance of the parameter register set already written in the previous window. This represents the power control coefficients of the parameter register set written in the previous window. The parameter strategy library will then use the current window's grid strength state variables. Divided into three segments, with target parameters given for each segment. (This can be understood as the recommended target point obtained through calibration / experience / offline tuning):

[0156] Strategy A (Strong Network Segment): ≥3.0, , Indicates strategy A and strategy B (medium strength): 2.0 ≤ <3.0, , Indicates strategy B and strategy C (weak network segment): <2.0, , Let C represent strategy C, where H, Z, and K with superscripts A, B, and C respectively represent the virtual inertia, virtual impedance, and power control coefficient corresponding to each strategy.

[0157] If the correspondence does not meet the preset consistency condition, the backtracking link is set with the following consistency condition: Condition C1: If If the value is less than 0 (power grid strength decreases), then the parameter adjustment direction should satisfy the following condition: ≥0, ≥0, ≤0, where (Change in virtual inertia) (Virtual impedance change) and (Change in power control coefficient) is obtained from historical register records; Condition C2: when | When |≥0.5, it is considered a significant change in intensity, requiring rematching of the strategy segment. In this example: = -0.8, violating the C1 directionality requirement while satisfying the C2 threshold, therefore consistency is not satisfied, triggering policy backtracking and reconstruction of the parameter update sequence; based on the current window measurement =2.2, falling into strategy B (2.0≤ <3.0), therefore the strategy entries for backtracking selection are: The sequence is reconstructed using an update method that combines target points and a smoothing coefficient. The smoothing coefficient λ is set to 0.6 (dimensionless, representing a 60% convergence to the target point). For any parameter x∈{H,Z,K}, the update is... ,in This represents any parameter among H, Z, and K after the update. This indicates that any parameter in register set H, Z, K has been written to the previous window. This indicates that the target point unit is consistent with x, and the updated virtual inertia is obtained by substituting specific values. =3.8 + 0.6 × (4.2 - 3.8) = 4.04 s, updated virtual impedance =0.18 + 0.6 × (0.22 - 0.18) = 0.204Ω, the updated power control coefficient =130+0.6×(110−130)=118kW / Hz. When the grid strength decreases (from 3.0 to 2.2), the reconstructed parameters show an increasing H, increasing Z, and decreasing K direction, satisfying the aforementioned consistency direction requirement; the three results are limited. =4.04s∈[2.0,6.0]s, through, virtual impedance =0.204Ω∈[0.10,0.35]Ω, through. =118kW / Hz∈[80,160]kW / Hz, passed, therefore the result is written to the parameter register set. .

[0158] Example 2

[0159] Please see Figure 4 Another embodiment of the present invention provides a grid-based energy storage system based on dynamic sensing of grid strength, comprising: an information acquisition module, a control parameter adjustment module, an instruction output module, and a sudden change adjustment module;

[0160] The information acquisition module is used to collect grid information at the grid-connected energy storage points and generate grid strength state quantities based on the grid information. The grid information includes voltage, current and frequency information.

[0161] The control parameter adjustment module is used to adaptively adjust the grid construction control parameters according to the grid strength state quantity. The grid construction control parameters include virtual inertia, virtual impedance, and power control coefficient.

[0162] The instruction output module is used to construct stable operation constraints for grid-connected energy storage based on the grid strength state quantity, and generate active and / or reactive power output instructions for grid-connected energy storage under the stable operation constraints, and adaptively adjust the output power of grid-connected energy storage according to changes in grid strength.

[0163] The mutation adjustment module is used to trigger network control reconfiguration and readjust network control parameters when a sudden change in the power grid strength state is detected.

[0164] In addition, the parts of the technical solutions provided in the embodiments of this application that are consistent with the implementation principles of the corresponding technical solutions in the prior art have not been described in detail, so as to avoid excessive elaboration.

[0165] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A grid-based energy storage method based on dynamic grid strength sensing, characterized in that, include: Collect grid information on grid-connected energy storage points and generate grid strength state parameters based on the grid information, which includes voltage, current and frequency information; Based on the power grid strength state quantity, the grid construction control parameters are adaptively adjusted, including virtual inertia, virtual impedance, and power control coefficient. Based on the power grid strength state variables, stable operation constraints for grid-connected energy storage are constructed, and active and / or reactive power output commands for grid-connected energy storage are generated under the stable operation constraints. The output power of grid-connected energy storage is adaptively adjusted according to changes in power grid strength. When a sudden change in the power grid strength state is detected, the grid construction control reconfiguration is triggered, and the grid construction control parameters are readjusted. The process of collecting grid information at grid-connected energy storage points and generating grid intensity state variables based on this information includes: The synchronous sampling sequence of three-phase voltage and three-phase current at the grid connection point of the grid-connected energy storage is collected, and the sampling sequence is time-aligned to form grid information under the same time reference; The power grid information is segmented, and a set of windows including steady-state windows and transition windows is generated according to a preset time window. A corresponding subset of power grid information is established for each window. Within the transition window, a preset perturbation sequence is superimposed, and the voltage and current responses during the period of the perturbation sequence are recorded simultaneously to form identification data pairs. Based on the identification data, a set of candidate equivalent impedances is generated in multiple frequency bands or multiple phase sequence dimensions, and the set of candidate equivalent impedances is subjected to consistency screening to obtain the target equivalent impedance parameters. The target equivalent impedance parameter is fused with the voltage, current and frequency information within the steady-state window to output a power grid strength state quantity that is updated as the window scrolls.

2. The grid-based energy storage method based on dynamic grid strength sensing as described in claim 1, characterized in that, Based on the identified data, candidate equivalent impedance sets are generated across multiple frequency bands or multiple phase sequence dimensions. Consistency screening is then performed on these candidate equivalent impedance sets to obtain the target equivalent impedance parameters, including: The identification data is divided into multiple data subsets according to a preset dimension, and a corresponding identification task identifier is established for each data subset. Based on the identification task identifier, voltage and current correlation sample pairs are extracted from the corresponding data subset, and multiple candidate equivalent impedance parameters are generated based on the correlation sample pairs, which are then aggregated to form a candidate equivalent impedance set. The candidate equivalent impedance set is subjected to consistency screening. The candidate equivalent impedance parameters are compared pairwise according to a preset comparison rule to generate a set of consistent results. The remaining candidate equivalent impedance parameters after consistency screening are fused to output the target equivalent impedance parameter. The fusion includes selection by frequency band priority, selection by phase sequence dimension priority, or selection by preset combination rules.

3. The grid-based energy storage method based on dynamic grid strength sensing as described in claim 2, characterized in that, The target equivalent impedance parameter is fused with the voltage, current, and frequency information within the steady-state window to output a grid strength state quantity that is updated as the window scrolls, including: Within the steady-state window, window feature sets corresponding to voltage, current, and frequency information are generated respectively, and the window feature sets are assigned a timestamp identifier corresponding to the steady-state window. The window feature sets include voltage amplitude features, current amplitude features, and frequency offset features. The target equivalent impedance parameter is aligned with the window feature set to form a fused input set. The alignment includes mapping the target equivalent impedance parameter to a window index that is consistent with the timestamp identifier, and replacing the target equivalent impedance parameter at the missing index with a preset placeholder parameter or backtracking parameter. A set of candidate state quantities for power grid strength is generated based on the fused input set. The set of candidate state quantities for power grid strength includes a first candidate state quantity, a second candidate state quantity, and a third candidate state quantity. During the window scrolling update process, the set of candidate power grid strength states is selected or switched, and the power grid strength states are output as the window scrolls.

4. The grid-based energy storage method based on dynamic grid strength sensing as described in claim 3, characterized in that, Based on the aforementioned power grid strength state parameters, adaptive adjustments are made to the grid construction control parameters, including: Obtain the power grid strength status quantity corresponding to the current window, and map the power grid strength status quantity to the target strength level in the preset strength level set, wherein the strength level set includes weak grid level and strong grid level; Based on the target intensity level, select the parameter strategy entry corresponding to the target intensity level from the preset parameter strategy library, and generate a parameter update task identifier for the parameter strategy entry. The preset parameter strategy library includes virtual inertia parameter strategy, virtual impedance parameter strategy and power control coefficient strategy. The parameter update task identifiers are decoupled and sorted to form a parameter update sequence according to a preset order, which includes updating the virtual impedance first, then updating the virtual inertia, and finally updating the power control coefficient. Based on each parameter in the parameter update sequence, a corresponding parameter transition trajectory is generated and written into the parameter register set of the network controller. The parameter transition trajectory includes a segmented incremental trajectory, a rolling hold trajectory, or an event-triggered trajectory. When the next window arrives, the correspondence between the power grid strength state quantity and the written parameter register set is verified. If the correspondence does not meet the preset consistency condition, the parameter strategy entry is backtracked and the parameter update sequence is reconstructed to complete the adaptive adjustment of the network control parameters. The network control parameters include virtual inertia, virtual impedance and power control coefficient.

5. The grid-based energy storage method based on dynamic grid strength sensing as described in claim 4, characterized in that, The parameter update task identifiers are decoupled and sorted to form a parameter update sequence according to a preset order, including: The update task identifiers for each parameter are categorized and marked, and then assigned to different parameter groups. A group priority identifier is also assigned to each parameter group. Based on the mutual influence relationship between parameter update task identifiers within each parameter group, a sequential constraint linked list is established for any two parameter update task identifiers. Parameter update task identifiers that are directly related are recorded as task pairs with sequential constraints, thus forming a parameter update sequence constraint set. According to the group priority identifier and the parameter update order constraint set, perform hierarchical sorting on all parameter update task identifiers: first, determine the update order of each parameter group at the group level, and then sort the parameter update task identifiers locally within the group according to the order constraint linked list to generate an intermediate parameter update sequence. The intermediate parameter update sequence is subjected to consistency adjustment, and the local sorting segments that do not meet the group priority identifier or order constraint linked list are rearranged to obtain the parameter update sequence that meets the preset order relationship.

6. The grid-based energy storage method based on dynamic grid strength sensing as described in claim 5, characterized in that, Based on each parameter in the parameter update sequence, a corresponding parameter transition trajectory is generated, and the parameter transition trajectory is written into the parameter register set of the network controller, including: Read the parameter values ​​and target parameter values ​​of each parameter in the parameter update sequence to form parameter state pairs, and group them according to the virtual impedance, virtual inertia or power control coefficient to which the parameters belong, and add trajectory type labels to each parameter state pair; Based on the trajectory type marker, a preset trajectory generation rule is selected for each parameter state pair to generate a parameter transition trajectory template; According to the order of the parameter update sequence, each parameter transition trajectory template is expanded into a corresponding discrete parameter sample sequence, and each discrete parameter sample sequence is associated with the control cycle number or time step number to form a parameter transition trajectory set. The parameter transition trajectory set is written into the parameter register set of the network controller.

7. The grid-based energy storage method based on dynamic grid strength sensing as described in claim 6, characterized in that, Based on the grid strength state variables, stable operation constraints for grid-connected energy storage are constructed, and active and / or reactive power output commands for grid-connected energy storage are generated under these constraints. The output power of grid-connected energy storage is adaptively adjusted according to changes in grid strength, including: Obtain the power grid strength status quantity corresponding to the current window, and map the power grid strength status quantity to the target operating mode identifier in the preset operating mode set. The preset operating mode set includes weak grid mode, transition mode and strong grid mode. Based on the target operating mode identifier, a stable operating constraint template corresponding to it is selected from the constraint template library, and the stable operating constraint template is scaled or translated based on the current power grid strength state quantity to generate a stable operating constraint set. Collect the expected active power and expected reactive power from the outside, and combine the expected active power and expected reactive power with the set of stable operation constraints to form a power expectation determination result that is inside or outside the set of stable operation constraints. After power correction, the power expectation determination result is projected onto the boundary or interior of the stable operation constraint set to generate constrained active power output command and / or reactive power output command. During window scrolling, when a change in grid strength is detected, the stable operation constraint template corresponding to the new target operation mode identifier is invoked, and the previously generated power output command is smoothly transitioned to form an active power output command and / or reactive power output command that changes with the grid strength.

8. The grid-based energy storage method based on dynamic grid strength sensing as described in claim 7, characterized in that, After power correction, the power expectation determination result is projected onto the boundary or interior of the stable operation constraint set to generate constrained active power output commands and / or reactive power output commands, including: Obtain the power expectation determination result after the stable operation constraint set is determined, mark the active power expectation and / or reactive power expectation outside the stable operation constraint set as power expectation to be corrected, and assign a projection priority identifier to each power expectation to be corrected according to the preset strategy of active power priority or reactive power priority. Based on the projection priority identifier, the corresponding projection rule set is selected, and the projection direction of the expected power to be corrected is determined. The projection direction includes the active axis direction, the reactive axis direction, or the combined active and reactive axis direction. Search for power points that satisfy preset constraints along the projection direction at the boundary or inside the stable operation constraint set, record the power points as a candidate projection power point set, and select the target projection power point from the candidate projection power point set according to a preset sorting rule. The active and / or reactive components of the target projected power point are output as constrained active power output commands and / or reactive power output commands, respectively.

9. A grid-connected energy storage system based on dynamic grid strength sensing, used to implement the grid-connected energy storage method based on dynamic grid strength sensing as described in any one of claims 1-8, characterized in that, include: Information acquisition module, control parameter adjustment module, command output module, and sudden change adjustment module; The information acquisition module is used to collect grid information at the grid-connected energy storage points and generate grid strength state quantities based on the grid information. The grid information includes voltage, current and frequency information. The control parameter adjustment module is used to adaptively adjust the grid construction control parameters according to the grid strength state quantity. The grid construction control parameters include virtual inertia, virtual impedance, and power control coefficient. The instruction output module is used to construct stable operation constraints for grid-connected energy storage based on the grid strength state quantity, and generate active and / or reactive power output instructions for grid-connected energy storage under the stable operation constraints, and adaptively adjust the output power of grid-connected energy storage according to changes in grid strength. The mutation adjustment module is used to trigger network control reconfiguration and readjust network control parameters when a sudden change in the power grid strength state is detected.

Images