Port green power virtual energy storage aggregation regulation and control method

Abstract

The invention relates to the technical field of power grids, and discloses a port green power virtual energy storage aggregation regulation and control method, which comprises the following steps: acquiring an aggregation scene and resource data; generating a power hard constraint mapping and pile division execution reference by calculating a multi-band energy index and identifying a system power transfer characteristic; a local rapid control configuration set is generated by solving a virtual impedance feasible region and combining power hard constraint allocation compensation budget; and fusing a sub-pile execution reference (slow channel) generated based on the power hard constraint with a fast channel execution track obtained after local online execution compensation, performing priority arbitration under the constraint of the power hard constraint, and outputting a final execution power track. According to the invention, a double-layer regulation and control framework of slow channel planning and fast channel compensation is constructed, and through a core link of power hard constraint mapping, coordinated operation of fast and slow channels and safety and stability of the whole system are ensured, and fast and reliable dynamic support capability can be provided for a power grid.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power grid, and particularly relates to a port green electricity virtual energy storage aggregation regulation method. BACKGROUND

[0002] With the increasing emphasis on environmental protection and carbon neutralization targets worldwide, as a key hub of international trade, the green and intelligent transformation of ports has become an inevitable trend of industry development. Among them, the full electrification of port operation equipment, such as rubber-tired cranes, rail-mounted cranes, automated guided vehicles (AGV) and container trucks, is a core measure to achieve a zero-carbon port. This transformation not only significantly reduces port carbon emissions and pollutants, but also generates a large amount of distributed mobile energy storage units with charging and discharging capabilities. If these scattered, random charging loads in space and time can be effectively aggregated and cooperatively regulated to form a large-scale port green electricity virtual energy storage system (VESS), not only can the internal energy be optimized and managed, but also can provide high-value auxiliary services such as frequency regulation, backup capacity, and demand-side response as a whole to the power system. This has extremely important research significance and engineering application value for enhancing the stability of regional power grids, promoting the local consumption of renewable energy such as wind and solar power, and opening up new value-added business models for port operators.

[0003] Currently, research on charging facilities aggregation participating in power grid regulation has made some progress. Existing technical solutions usually adopt centralized or decentralized control architecture. In the centralized control scheme, an upper aggregation regulation center calculates the power reference curve of each charging terminal in the future period of time according to the dispatching instructions issued by the power grid or the internal optimization target, and directly issues it to each terminal for execution through the communication network. To achieve basic power grid support, some schemes introduce droop control strategies, that is, each charging pile adjusts its active or reactive power output according to the local measured power grid frequency or voltage deviation according to a fixed droop coefficient, thereby realizing primary frequency regulation or voltage support function. At the same time, to ensure the basic charging needs of vehicles, these schemes usually integrate battery state of charge (SOC) balancing management strategies to avoid affecting normal production operations of the port due to deep discharge of some vehicles. In addition, to suppress the impact on the power grid caused by the simultaneous start-stop or power adjustment of a large number of equipment, existing technologies usually set a fixed and conservative limit value for the power change rate of the charging pile, that is, the power is controlled in a linear ramp manner.

[0004] However, the existing technology still faces deep technical challenges in dealing with complex scenarios of large-scale, high-density power electronic converter (charging pile) aggregation regulation. These challenges mainly come from the inherent contradiction between the rapidity of system dynamic response and the safety of operating boundaries, as well as the inherent stability risk of a large number of converters interacting with the grid. First, the existing control scheme generally decouples the control target from the safety boundary. Second, to achieve high-value grid services, the aggregation system needs to have high-bandwidth response capability, but this will induce instability risk of high-bandwidth control interacting with the grid. SUMMARY

[0005] To solve the above technical problems of the contradiction between fast power response and internal electrical safety in the grid, as well as the instability risk of a large number of converters interacting with unknown grid, embodiments of the present application propose a port green electricity virtual energy storage aggregation regulation method.

[0006] The technical solutions adopted by the present application are as follows: Embodiments of the present application propose a port green electricity virtual energy storage aggregation regulation method, comprising the following steps: obtaining aggregation scenario and resource data, the aggregation scenario and resource data including: a synchronous data set of virtual energy storage, a baseline power trajectory, a device constraint set, and a vehicle and charging facility available capacity set; based on the aggregation scenario and resource data, generating power hard constraint mapping and sub-pile execution reference by calculating multi-band energy index and identifying system power transmission characteristics, the energy index including: power quality index; based on the aggregation scenario and resource data and the power hard constraint mapping, generating a local fast control configuration set by solving a virtual impedance feasible region and apportioning a compensation budget, and obtaining an aligned fast channel execution trajectory by online executing compensation; fusing the sub-pile execution reference, the local fast control configuration set, and the aligned fast channel execution trajectory, and performing priority arbitration under the constraint of the power hard constraint mapping, and outputting a virtual energy storage final execution power trajectory.

[0007] Optionally, generating the power hard constraint mapping and the sub-pile execution reference comprises: according to the device direct current ripple limit value contained in the device constraint set in the aggregation scenario and resource data, reversely deducing the upper limit of the power peak value and the change rate of the power domain to form the power hard constraint mapping; using the power hard constraint mapping to constrain and allocate the baseline power trajectory in the aggregation scenario and resource data to generate the sub-pile execution reference.

[0008] Optionally, the reverse derivation of the upper limit of the power peak value and the change rate of the power domain includes: based on the aggregated scene and resource data, calculating the power change energy in the low frequency band, twice the power grid frequency band and the switching sideband to obtain three frequency band energy indicators; based on the aggregated scene and resource data, identifying the amplitude-frequency transfer characteristic of the power change to the direct current voltage ripple to obtain the power to direct current ripple transfer characteristic; combining the three frequency band energy indicators and the power to direct current ripple transfer characteristic, and solving with the direct current ripple limit value as the boundary condition to determine the upper limit of the power peak value and the change rate.

[0009] Optionally, the generation of the power hard constraint mapping and the segmented execution reference further includes: monitoring the zero crossing times of the aggregated power to obtain a zero crossing count; aggregating the low frequency band energy and insulation monitoring alarms reflected by the aggregated scene and resource data to generate a zero crossing risk indicator; dynamically adjusting the polarity maintaining window length and the buffer bandwidth used to constrain the segmented execution reference according to the zero crossing count and the zero crossing risk indicator.

[0010] Optionally, the generation of the power hard constraint mapping and the segmented execution reference further includes: at the boundary of the polarity maintaining window, performing two-stage smooth switching, wherein the power change rate is slowed down to near zero using trend feedforward; enabling second-order ramp control to complete the zero-crossing switching of the power polarity, and during the switching, if the energy indicator of any frequency band is monitored to rise, the execution time of the second-order ramp is extended.

[0011] Optionally, the generation of the local fast control configuration set includes: based on the aggregated scene and resource data, identifying the equivalent impedance frequency point set of the power grid by injecting a power disturbance sequence and collecting the response; solving the feasible region of the virtual impedance parameters that satisfy the preset frequency domain passivity condition in combination with the equivalent impedance frequency point set of the power grid; selecting a group of parameters from the feasible region of the virtual impedance parameters to form a virtual impedance configuration, and taking it as a component of the local fast control configuration set.

[0012] Optionally, the online execution of compensation to obtain the aligned fast channel execution trajectory includes: in each local control period, superimposing the virtual impedance configuration to the power compensation control loop; taking the equivalent impedance frequency point set of the power grid as the benchmark, online evaluating the passivity margin of the current period and generating a passivity compliance marker; performing safety arbitration on the result of power compensation according to the passivity compliance marker to form the final compensation instruction.

[0013] Optionally, the generation of the local fast control configuration set further includes: determining the total compensation bandwidth budget of the whole station according to the power hard constraint mapping; based on the available capacity of each device in the aggregated scene and resource data, calculating the compensation participation factor of each device; allocating the total compensation bandwidth budget to each device in accordance with the compensation participation factor to form a device-level gain upper limit budget, and incorporating it into the local fast control configuration set.

[0014] Optionally, the online execution compensation to obtain the aligned fast channel execution trajectory further comprises: collecting communication indicators such as communication delay, jitter and packet loss rate, and comprehensively scoring as a communication reliability score; determining a state machine state label according to the communication reliability score and a preset threshold, the state machine state label comprising any one of normal, step loss, shape preservation and derating; and adaptively adjusting the gain of power compensation according to the state machine state label before executing the final compensation instruction.

[0015] Optionally, the priority arbitration of the fusion and segmentation execution reference, the local fast control configuration set and the aligned fast channel execution trajectory under the constraint of the power hard constraint mapping comprises: when fusing and segmenting the execution reference and the aligned fast channel execution trajectory, detecting in real time whether the synthesized power instruction exceeds the boundary of the power hard constraint mapping; if it exceeds the boundary, according to the priority rule of frequency support first, voltage support second and trajectory tracking third, the segmentation execution reference is trimmed to ensure that the fusion result is subject to the power hard constraint mapping.

[0016] The beneficial effects of the present application are: The present application solves the technical problems of the contradiction between fast power response and system internal electrical safety in the prior art, and the instability risk of a large number of converters interacting with unknown power grids, and constructs a double-layer regulation and control architecture of a slow channel plan + fast channel compensation, and realizes the provision of fast and reliable dynamic support capability for the power grid under the premise of actively ensuring the stability of the whole system through the core link of the power hard constraint mapping. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 is a flowchart of a port green electricity virtual energy storage aggregation regulation and control method according to an embodiment of the present application.

[0018] Figure 2 is a flowchart of generating a power hard constraint mapping and a segmented execution reference according to an embodiment of the present application.

[0019] Figure 3 is a flowchart of generating a power hard constraint mapping and a segmented execution reference according to another embodiment of the present application.

[0020] Figure 4 is a flowchart of generating a local fast control configuration set according to an embodiment of the present application.

[0021] Figure 5 is a flowchart of online execution compensation to obtain an aligned fast channel execution trajectory according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to solve the problems in the prior art, the applicant has conducted in-depth research and found that: The decoupling of control targets and safety boundaries is reflected in the fact that existing power instruction generation is mostly aimed at tracking upper-layer scheduling signals or meeting droop responses, while internal electrical safety constraints of the system, such as voltage ripple of a key DC bus, are not directly and dynamically mapped to the power control domain. The controller can usually only rely on a static, conservative power change rate limit value set for the worst-case scenario. This one-size-fits-all static constraint results in the rapid response potential of virtual energy storage being severely suppressed when the system state is good, and even if the static limit value is followed, the rapid power instruction issued by the controller can still excite resonance within the port microgrid, causing the DC ripple to exceed the limit and triggering protection, which unexpectedly interrupts the grid support service that should have been provided. The root cause of this problem lies in the lack of a mechanism that can convert internal, frequency-domain-related power quality constraints (such as ripple) into power-domain time-varying power boundaries (i.e., peak and change rate upper limits) in real time and quantitatively. The instability problem of high-bandwidth control interacting with the grid is even more difficult. When hundreds of charging pile control loops (which need to have a bandwidth of several hertz or even higher to provide high-quality frequency support) interact with an unknown, time-varying grid equivalent impedance, it is easy to form negative damping in a certain frequency band (usually the low and medium frequency bands), thereby causing subsynchronous or supersynchronous oscillation. Existing technologies either cannot effectively deal with this problem due to insufficient control bandwidth, or although advanced control strategies are used, they usually assume that the grid impedance is known or ideal, lack a mechanism that can actively identify the characteristics of the grid and accordingly configure the controller parameters in advance to actively establish the system stability margin, resulting in the system always operating under the risk of potential instability.

[0023] Embodiment one describes the aggregation regulation architecture and data basis of the port green electricity virtual energy storage system, which is used to provide an exemplary system environment and data basis, and details the preparation stage of data collection and preprocessing.

[0024] In this embodiment, a port green electricity virtual energy storage aggregation regulation system is deployed in an automated container terminal. The system physically includes an aggregation regulation master station and communicates with multiple distributed energy units in the terminal through industrial Ethernet or a 5G private network. The distributed energy units are mainly charging piles connected to work equipment such as container trucks (AGV), rubber-tired gantries (RTG), and rail-mounted gantries (RMG), which have bidirectional charging and discharging capabilities and can be aggregated as virtual energy storage resources. The aggregation regulation master station is used to execute the regulation method of the present application.

[0025] Specifically, the first stage related to this embodiment is data aggregation and basis alignment, the process of which includes: Step S11, data acquisition. The aggregated regulation master station reads and aggregates data from various locations in the port in real time through the communication network. For example, the acquired data channels include at least: the operating power of each charging pile, the active power of the port at the point of common coupling (PCC) of the power grid, the voltage effective value of the key bus, the bus frequency, the battery status (such as state of charge SOC, temperature) of each vehicle, the working status (such as standby, charging, discharging) of the charging facility, and the standard time pulse signal from the GPS or Beidou system.

[0026] In order to ensure the synchronization and accuracy of the data, the acquired raw data will be subjected to a series of preprocessing. Among them, time calibration is a key step, the master station takes the received time pulse as the reference, adds a unified time stamp to all uploaded data, and uniformly resamples with a sampling step of 5ms to obtain aligned raw measurements. This is aimed at eliminating the time deviation introduced by the difference between the sampling clocks of each measurement channel, and is the basis for subsequent accurate frequency domain analysis and cooperative control. On this basis, the aligned raw measurements are subjected to missing data repair and outlier suppression, for example, linear interpolation or hold method based on adjacent data points can be used for missing data repair, and 3σ criterion (three sigma criterion) or local outlier factor (LOF) algorithm can be used to identify and eliminate outliers in the data sequence. After the above processing, a high-quality synchronized data set is obtained for subsequent steps.

[0027] Step S12, constraint loading. The aggregated regulation master station loads and analyzes the preset or received from the grid dispatching mechanism operation constraints. These constraints constitute the safety boundary of the regulation decision. Specifically, the constraints can include: the grid-connected power limit value signed by the port and the main grid (for example, the maximum grid-connected power shall not exceed 10MW), the insulation monitoring threshold of the charging facility itself (used to judge whether there is a risk of electric leakage), the trigger voltage condition of the battery pre-charging process, the limit value of the DC bus ripple voltage (for example, the peak-to-peak value shall not exceed 5% of the rated voltage), the maximum power change rate allowed by a single charging and discharging device (for example, not more than 500kW / s), and the compensation gain range allowed when performing fast frequency or voltage support. These constraints are combined to form a device constraint set.

[0028] Step S14, resource capability aggregation. The master station periodically (for example, every 1 minute) statistics the real-time status of all vehicles and charging facilities in the synchronized data set to evaluate the available capacity of the entire virtual energy storage system. Specifically, this process will exclude terminals that are in fault, maintenance or have full charge and do not allow discharging, then accumulate the available charging and discharging power, the remaining available energy (in kWh) of all available terminals, and record their corresponding charging gun connector status. Finally, the vehicle and charging facility available capacity set is aggregated, which dynamically updates to reflect the real-time energy-power reserve of the virtual energy storage pool.

[0029] Embodiment two, this embodiment illustrates the overall process of the polymerization regulation method defined therein. The method process of this embodiment is based on the system architecture and data foundation constructed in embodiment one.

[0030] obtain a synchronization data set, a device constraint set, a vehicle and charging facility available capability set, and a baseline power trajectory; based on the synchronization data set and the device constraint set, generate a power hard constraint mapping and a sub-pool execution reference by calculating a multi-band energy index and identifying a power-to-dc ripple transmission characteristic; based on the synchronization data set, the device constraint set, the vehicle and charging facility available capability set, and the power hard constraint mapping, generate a local fast control configuration set by solving a virtual impedance feasible region and apportioning a compensation budget, and perform compensation online according to the local fast control configuration set to obtain an aligned fast channel execution trajectory; and perform priority arbitration under the constraint of the power hard constraint mapping by fusing the sub-pool execution reference, the local fast control configuration set, and the aligned fast channel execution trajectory, and output a final execution power trajectory.

[0031] The power hard constraint mapping and the sub-pool execution reference generated based on the aggregated scenario and resource data include: parsing the synchronization data set, the device constraint set, and the vehicle and charging facility available capability set from the aggregated scenario and resource data; based on the synchronization data set, calculating the power change energy of the low frequency band, the twice grid frequency band, and the switching sideband to obtain the three-band energy index; injecting a short-time power sequence into the synchronization data set to identify the amplitude-frequency characteristic of the power-to-dc voltage ripple to obtain the power-to-dc ripple transmission characteristic; based on the dc ripple limit value in the device constraint set, combining the power-to-dc ripple transmission characteristic and the three-band energy index, and reversely deducing the allowed power peak value and the upper limit of the change rate of each frequency band to form the power hard constraint mapping; and fusing the vehicle and charging facility available capability set and using the power hard constraint mapping to clip to allocate a target for the charging facility to generate the sub-pool execution reference.

[0032] In other words, as shown in Figure 1 A port green electricity virtual energy storage aggregation regulation method includes: Step one, obtain the aggregated scenario and resource data. This step can specifically refer to the data collection, cleaning, constraint loading, and resource aggregation process of embodiment one. The purpose is to provide comprehensive, accurate, and synchronized basic data input for subsequent regulation decisions, specifically including the synchronization data set, the device constraint set, and the vehicle and charging facility available capability set generated by the aforementioned embodiment one.

[0033] Step two, based on the aggregated scenario and resource data, the power hard constraint mapping and the segmented execution reference are generated by calculating the multi-band energy index and identifying the system transmission characteristics. This step is the key mapping link in the invention from power quality to power control. In this embodiment, in order to avoid the impact of severe power fluctuations on the DC bus, causing overvoltage or overcurrent protection, this step innovatively maps the electric energy quality indicators that are not easy to control directly (such as DC ripple) into the power domain constraints that can be directly executed by the controller (i.e. power hard constraint mapping), and based on this constraint, a safe and smooth power instruction is planned for each charging pile (i.e. segmented execution reference). The detailed implementation of this step will be further described in subsequent embodiments (such as embodiments three and four).

[0034] Step three, based on the aggregated scenario and resource data and the power hard constraint mapping, the local fast control configuration set is generated by solving the virtual impedance feasible region and allocating the compensation budget, and the aligned fast channel execution trajectory is obtained by online compensation. This step aims to establish a local compensation channel that can quickly respond to grid disturbances (such as frequency and voltage fluctuations), i.e. the fast channel. In order to ensure the stability of a large number of charging piles connected to the grid, this step introduces a virtual impedance design based on the passivity theory, which solves the range of virtual impedance parameters that can ensure stable operation by actively identifying the characteristics of the grid. At the same time, the total compensation capacity allowed by the entire station (limited by the power hard constraint mapping generated in step two) is reasonably distributed to each charging pile. These parameters together form the local fast control configuration set, which is sent to each terminal for execution. The detailed implementation of this step will be further described in subsequent embodiments (such as embodiments five and six).

[0035] Step four, the final execution power trajectory is output by fusing the segmented execution reference, the local fast control configuration set, and the aligned fast channel execution trajectory, and performing priority arbitration under the constraint of the power hard constraint mapping. This step is the final instruction synthesis and execution link. As can be seen, the segmented execution reference generated in step two belongs to a low-speed planned power trajectory (which can be called a slow channel), while the fast channel execution trajectory generated in step three is a high-speed compensation power that responds to disturbances. This step fuses the two and checks them against the power hard constraint mapping generated in step two as the highest priority constraint. When the fused instruction exceeds the constraint, it will be arbitrated and cropped according to the pre-set priority rules (for example, the frequency support to ensure the safety of the grid takes priority over planned tracking), and the final power instruction is output for execution by the charging pile. The detailed implementation of this step will be further described in subsequent embodiments (such as embodiment seven).

[0036] In summary, the flow of this embodiment builds a slow channel planning + fast channel compensation double-layer regulation architecture, and through the core link of the power hard constraint mapping, it ensures the coordinated operation of the fast and slow channels and the safety and stability of the entire system.

[0037] Embodiment three describes a power hard constraint mapping generation process based on three-band energy, i.e. a process of inversely mapping power quality constraints (DC ripple) into power domain hard constraints.

[0038] In this embodiment, the aggregation regulation method comprises the following steps: based on the aggregation scenario and resource data, a power hard constraint mapping and a segmented execution reference are generated by calculating a multi-band energy index and identifying system transfer characteristics. In this embodiment, in order to fundamentally avoid the voltage impact on the DC bus of the port microgrid caused by the power fluctuation of a large number of distributed power sources (charging piles), and thus trigger equipment protection actions or even shutdown, the present application proposes a method of converting the DC ripple, which is not easy to directly observe and control, into power instruction upper and lower limits (i.e. power hard constraints) that can be directly understood and executed by the controller.

[0039] Specifically, as shown in Figure 2 According to the equipment DC ripple limit value contained in the aggregation scenario and resource data, the upper limit of the power peak value and the change rate in the power domain is inversely deduced to form a power hard constraint mapping. This inverse deduction process is not achieved by using a fixed margin or limiting amplitude, but by a dynamic and adaptive calculation method, as shown in Figure 3 which comprises: based on the aggregation scenario and resource data, calculating the power change energy in the low frequency band, the twice grid frequency band and the switching sideband to obtain a three-band energy index; based on the aggregation scenario and resource data, identifying the amplitude-frequency transfer characteristics of power change to DC voltage ripple to obtain the power to DC ripple transfer characteristics; combining the three-band energy index and the power to DC ripple transfer characteristics, and solving with the DC ripple limit value as the boundary condition to determine the upper limit of the power peak value and the change rate.

[0040] In order to achieve the above process, the aggregation regulation master station first performs the following preparations: Online identification of power-to-dc ripple transfer characteristic: the system injects a short-time, low-amplitude power perturbation sequence to some charging piles via the aggregated regulation master station in the gap (e.g., the short interval between the operation of two quayside cranes) without affecting the normal operation of the port. Optionally, the sequence can be a multi-sine signal or a pseudo-random binary sequence (PRBS) whose frequency components cover the range of interest but deliberately avoid known sensitive frequency bands. For example, the amplitude of the perturbation sequence can be set to within 2% of the rated power of the charging pile, and the duration is about 1-3 seconds. During the perturbation injection, the master station synchronously collects the injected total power perturbation signal sequence and the voltage ripple response signal sequence on the key DC bus. Based on these two sets of input-output data, a frequency domain model is calculated using the cross-spectrum method or the frequency domain estimation method in the system identification toolbox. The model is the power-to-dc ripple transfer characteristic, whose gain and phase at different frequency points quantitatively describe how much amplitude of the same frequency voltage ripple will be generated on the DC bus when a unit power perturbation is injected under the current topology and load of the port system. For example, if the gain at 100Hz frequency point is identified as 0.05p.u., it means that under the current working condition, an injection of a 1kW 100Hz sinusoidal power fluctuation will generate a 100Hz voltage ripple of 1kW 0.05 / P_base on the DC bus (where P_base is the power reference value).

[0041] Calculation of three-band energy index: the master station continuously analyzes the time series data of the total active power of the PCC point in the synchronous data set obtained in embodiment one. By designing three parallel digital bandpass filters, the power signal is decomposed into three key frequency bands. In this embodiment, the three frequency bands are preferably set as: Low frequency band: for example, 1Hz to 20Hz, power fluctuations in this range are usually related to slow dynamic processes such as start-stop of large equipment, mechanical load changes, etc., and excessive energy may cause system oscillation.

[0042] Two times grid frequency band: for example, 95Hz to 105Hz (for 50Hz power grid), power fluctuations in this range are mainly generated by rectification / inversion links and are one of the main sources of DC ripple.

[0043] Switching sideband: for example, 18kHz to 22kHz (for 20kHz switching frequency), power fluctuations in this range are directly related to the switching action of power electronic converters themselves, and excessive energy may cause high-frequency noise and electromagnetic interference problems.

[0044] For the filtered power signals of each frequency band, the square integral is performed within a sliding time window (e.g., a 20-second window for the low-frequency band and a 500-ms window for the high-frequency band) to obtain the power change energy of each frequency band. These three energy values ​​together constitute the three-band energy index vector, which dynamically reflects the distribution of recent power fluctuation energy in different frequency domains.

[0045] Based on this, the main station executes the core joint solution process to determine the power hard constraints. Specifically, for any frequency band among the low-frequency band, twice the grid frequency band, and the switching sideband, the basic power margin is calculated based on the power-to-DC ripple transfer characteristics and DC ripple limits, and the energy suppression factor is calculated based on the energy indices of the three frequency bands. The energy suppression factor is inversely proportional to the square root of the energy index of the corresponding frequency band. The basic power margin and the energy suppression factor are combined to determine the upper limit of the power peak value and rate of change of the frequency band.

[0046] This joint process can be achieved through the following formula: P allow,b =α b Vlim(||G) P→V || b (1 / sqrt(E)) b +∈)); where: P allow,b The maximum peak power (in kW) allowed to contribute to frequency band b is a component of the upper bound of the power to be solved. b is the frequency band index, taking values ​​for the low-frequency band, twice the grid frequency band, and the switching sideband, respectively. α b The safety factor for frequency band b is a dimensionless parameter used to maintain a certain safety margin. Preferably, its value ranges from [0.6, 0.9], and it can be set differently according to the importance of different frequency bands. lim The DC bus ripple voltage limit (unit: V or pu) defined for the equipment constraint set is the boundary condition for this calculation. ||G P→V || b E represents the power-voltage coupling amplification factor of the system in that frequency band, which is the norm of the power-to-DC ripple transfer characteristic identified above (e.g., the maximum gain or L2 norm in that band). b The calculated energy index for frequency band b is given by the three-band frequency band (unit: (kW)^2·s). ∈ is a small positive number, such as 10, used to prevent the denominator from being zero. -6 .

[0047] It can be seen that the calculation logic of this formula is divided into two parts: the first part V lim / ||G P→V || bThe first part, 1 / sqrt(E_b+ε), can be regarded as a base power margin, which translates the voltage constraint into a power constraint directly based on the physical characteristics of the system; the second part, 1 / sqrt(E_b+ε), can be regarded as an energy suppression factor, which is dynamically adjusted according to the current operating state of the system. When the energy E_b of a certain frequency band is already very high, the suppression factor will become smaller, thereby significantly tightening the power margin of the frequency band and playing a role in actively suppressing potential instability.

[0048] The P allow,b After that, the minimum value of the three is taken, and then processed through a smoothing filter (such as a first-order low-pass filter), to obtain the final power peak hard constraint at the current time. The hard constraint of the power change rate can be derived by a similar method using the derivative or difference relationship of the transfer characteristic. This set of time-varying power peak and change rate curves collectively constitutes the power hard constraint mapping.

[0049] As Figure 2 Finally, the power hard constraint mapping is used to constrain and allocate the aggregated scenario and resource data to generate a split-pile execution reference. is an ideal and smooth total power curve predicted according to the port production plan. The master station compares this curve with the power hard constraint mapping, and performs peak clipping, valley filling or slope limiting processing on the part that exceeds the constraint, to obtain a revised total power target. Then, according to the available capacity set of vehicles and charging facilities, the total power target is decomposed to each available charging pile in a proportional or optimized manner, to form the final split-pile execution reference issued to each terminal.

[0050] As an optional implementation, the identification of the transfer characteristic can not be performed in real time, but can be performed offline in several typical port operation scenarios (such as daytime operation peak and nighttime concentrated charging) and stored in a database. During operation, the system automatically identifies the scenario according to the current time or working condition, and calls the corresponding transfer characteristic model, thereby reducing the complexity of online calculation.

[0051] Embodiment four describes a two-stage smoothing switching process of the polarity maintaining window self-adaptation and the boundary, which provides a refined improvement for the generation and execution process of the split-pile execution reference, and is specially used to solve the stability and safety problem of the aggregated power crossing near zero point (i.e. switching from charging state to discharging state, or vice versa).

[0052] In this embodiment, the method further includes: monitoring the number of zero point crossings to obtain a zero point crossing count; aggregating the low frequency band energy reflected by the aggregated scenario and resource data and the insulation monitoring alarm to generate a zero crossing risk indicator; and dynamically adjusting the polarity maintaining window length and the buffer bandwidth used to constrain the split-pile execution reference according to the zero point crossing count and the zero crossing risk indicator.

[0053] Specifically, when the aggregated power needs to switch from one polarity (e.g. charging, positive power) to the other (e.g. discharging, negative power), if the switching speed is too fast, it is very easy to cause system oscillation and DC voltage instability near zero. To this end, the embodiment introduces the concept of polarity maintenance window and buffer band.

[0054] The polarity maintenance window defines a minimum time period during which the aggregated power should maintain its polarity, avoiding frequent crossing near zero.

[0055] The buffer band defines a small power interval (e.g. ±5% of the rated power) centered at zero power, when the power enters this interval, a special slow control strategy must be followed.

[0056] To make these two parameters adapt to the changes in system state, the embodiment adopts an adaptive adjustment mechanism. The master station continuously performs the following calculations: Zero crossing count: in a sliding time window (e.g. the past 5 minutes), count the number of times the zero is crossed in embodiment three.

[0057] Over-zero risk indicator: this is a comprehensive indicator that aggregates multiple risk factors through weighted sum. For example, over-zero risk indicator = w1 x (low frequency band energy) + w2 x (insulation monitoring alarm times). Where w1 and w2 are weight coefficients. High low frequency band energy means that the system has a low frequency oscillation risk; insulation monitoring alarm may be related to the grounding of the DC system, and the risk is higher when the power reverses.

[0058] Dynamic adjustment: according to the above indicators, adjust the window length and buffer band width through pre-set rules. For example: window length = T_base + k1 x (zero crossing count) + k2 x (over-zero risk indicator); buffer band width = B_base + h1 x (over-zero risk indicator). Where T_base and B_base are base values, k1, k2, h1 are positive coefficients. This means that when the system detects frequent crossing or high over-zero risk, it will automatically extend the polarity maintenance window and widen the buffer band, forcing the system to perform power reversal in a more conservative and slower manner.

[0059] Further, to ensure the process is extremely smooth when crossing the buffer band, the method of the embodiment further comprises: at the boundary of the polarity maintenance window, performing a two-stage smooth switching, wherein first, the trend feedforward is used to slow down the power change rate to near zero; then, a second-order ramp control is enabled to complete the zero-crossing of the power polarity, and during the switching, if the energy indicator of any frequency band is monitored to rise, the execution time of the second-order ramp is extended.

[0060] This two-stage switching process is implemented as follows: Stage 1: Trend feedforward deceleration. When the aggregated power trajectory is approaching the buffer band, the controller does not simply linearly decelerate. Instead, it activates a trend feedforward module that predicts the inertia of the power as it rushes into the buffer band based on its short-term rate of change before entering the boundary. Based on this prediction, the controller applies a reverse and smooth control force to actively reduce the power rate of change to a very small value (close to zero), allowing it to gently reach the boundary of the buffer band.

[0061] Stage 2: Second-order ramp across zero. Once the power enters the buffer band, the control strategy switches to a second-order slope. This means that not only is the rate of change of power limited, but the rate of change of the rate of change (i.e., acceleration) is also limited, achieving the smoothest zero crossing. During the execution of this ramp, the master station continuously monitors the three-band energy indicators in Example Three at high frequency. If it finds that the energy of any band, especially the low-frequency band and the two times the grid frequency band, rises abnormally and exceeds the warning threshold, it indicates that this zero crossing may be triggering unstable modes in the system. At this time, the controller will immediately extend the execution time of the second-stage second-order ramp, for example, from the original planned 200ms to 350ms. By further slowing down the zero-crossing speed, it actively suppresses the development of oscillations, ensuring the absolute safety of the switching process.

[0062] As an optional implementation, the parameters of the junction of the two-stage switching and the extension logic can be optimized online by an expert system or reinforcement learning agent to adapt to more complex operating conditions. In addition, before entering the first stage, the master station can issue early warning information to the local controller about the upcoming power reversal, allowing it to pre-adjust the compensation gain and form a more coordinated response.

[0063] Example Five describes the configuration of a passive-constrained local fast control channel, explaining how to configure the parameters of the local fast control channel for each distributed power source (charging pile) to achieve fast and stable grid support.

[0064] In large port areas, a large number of charging piles operate as power electronic inverters (Inverter) connected to the grid. Their high-frequency switching characteristics and control loop behavior may interact with the equivalent impedance of the grid, causing resonance or instability in certain frequency bands. To fundamentally solve this problem, the invention proposes a method to actively ensure system stability, which generates a set of configuration parameters for each local controller to ensure its passivity before issuing specific compensation instructions. The process is as follows: based on the mapping of aggregated scenario and resource data and power hard constraints, a virtual impedance feasible region is solved and the compensation budget is allocated to generate a set of local fast control configuration parameters.

[0065] Specifically, as shown in Figure 4 This step includes the following core links: Online identification of grid equivalent impedance: By injecting power disturbance sequence and collecting the response, the equivalent impedance frequency set of the grid is identified. The purpose of this process is to obtain the dynamic characteristic portrait of the local grid at different frequencies.

[0066] The equivalent impedance frequency set of the grid refers to a data set that contains the complex impedance values (including amplitude and phase angle) presented from the public grid connection point (PCC) of the charging pile cluster into the grid at a series of discrete frequency points.

[0067] The aggregation control master station plans and executes a low-amplitude short-window, multi-frequency separation disturbance injection. For example, a 3-second time window with a relatively stable work load is selected, and a small, additional power disturbance reference is synchronously issued to all or part of the charging piles in the system. The disturbance signal is preferably a multi-sine superposition signal, with frequency points distributed in the target frequency band of 5Hz to 2kHz, but avoiding known sensitive areas such as power frequency (50Hz) and double frequency (100Hz); the total amplitude is strictly limited to within 2% of the rated power of the entire station to ensure no disturbance to normal production. During the injection, the high-speed monitoring unit (such as PMU or high-frequency recording device) deployed at the PCC point synchronously records the instantaneous values of the injection point voltage and current at a sampling rate not lower than 10kHz. After the collection is completed, the voltage and current response sequences are subjected to Fourier transform (FFT), and at each injected frequency point ω, the complex estimate value of the grid admittance Y_grid(jω)=ΔI(jω) / ΔV(jω) is calculated; where ΔI(jω) is the current response phasor, and ΔV(jω) is the voltage response phasor. Finally, by taking the reciprocal Z_grid(jω)=1 / Y_grid(jω), the equivalent impedance of the grid at this frequency point is obtained. By repeating this process for all frequency points, the required impedance frequency point set is formed.

[0068] Through active and safe online identification, the impedance characteristics reflecting the real working conditions of the current grid can be obtained, replacing the traditional control design method that relies on inaccurate system models or typical parameters, providing an accurate data basis for subsequent stability design, and being a prerequisite for realizing high-reliability adaptive control.

[0069] Solution of virtual impedance feasible region: According to the pre-set frequency domain passivity condition and combined with the equivalent impedance frequency set of the grid, the virtual impedance parameter feasible region that satisfies the condition is solved.

[0070] Virtual impedance is a control algorithm that modifies the voltage or current control loop of the converter, making it present additional, programmable resistance, inductance or capacitance characteristics, thereby improving the dynamic performance of the system. The virtual impedance parameter feasible region is a set of parameter points in the multi-dimensional parameter space (such as virtual resistance R_v, virtual inductance L_v) that can make the system satisfy the passivity constraint.

[0071] In this embodiment, the preset frequency domain passivity condition is: at any frequency point in the target frequency band, the real part of the series impedance composed of the equivalent impedance of the device and the equivalent impedance frequency point set of the grid is not less than the preset passivity margin δ. This condition can be expressed by the formula: Re{Z device (jω)+Z grid (jω)}≥δ; Where: Re{...} is the real part of the complex number operation; Z_grid(jω) is the grid impedance identified in the previous step; Z_device(jω) is the equivalent output impedance of the charging pile converter, which can be expressed as Z_device(jω)=Z_phy(jω)+Z_v(jω), Z_phy(jω) is the inherent physical impedance of the converter, and Z_v(jω) is the virtual impedance to be designed, whose expression is related to the parameters to be solved, such as virtual resistance R_v and virtual inductance L_v; δ is the preset passivity margin, which is a small positive real number (for example, 0.02 p.u.), used to ensure that the system remains stable under parameter perturbation or model uncertainty. Expanding the above inequality at a series of discrete frequency points in the target frequency band (such as 5Hz-300Hz) forms a linear inequality system about R_v and L_v. By using grid search method or convex optimization toolkit to solve the inequality system, a two-dimensional or multi-dimensional feasible region can be obtained.

[0072] Based on the strict passivity theory, the system stability problem is transformed into a mathematical constraint that can be explicitly solved. By solving the feasible region, the stability of the fuzzy target is guaranteed to be explicitly converted into the specific value range of the controller parameters (virtual impedance), so that the stability design is changed from post-validation to pre-constraint, greatly improving the design efficiency and reliability of the control system.

[0073] Allocation of compensation budget: after determining the stability region, the output upper limit of each device when performing fast compensation needs to be determined.

[0074] Determine the total compensation bandwidth budget: according to the power hard constraint mapping, determine the total compensation bandwidth budget of the whole station. Specifically, the aggregation control master station obtains the power rate hard constraint curve generated by embodiment three, which changes over time, and takes the minimum value of the future control cycle as Rate_limit_hard. At the same time, aggregate the sum of the physical maximum power rate of all available charging piles Rate_limit_phy. The total compensation bandwidth budget Budget_total of the whole station is carefully set to min(Rate_limit_hard, Rate_limit_phy)×k_safety, where k_safety is a safety factor (such as 0.8). This closely couples the dynamic capability of the fast channel with the global safety constraint of the slow channel.

[0075] Calculate compensation participation factor: Based on the aggregated scenario and resource data, the available capability of each device is calculated to obtain the compensation participation factor of each device. For the i-th device, its participation factor Factor_i can be weightedly synthesized by its available power P_avail_i, available energy E_avail_i, etc., such as Factor_i = w_p x (P_avail_i / ∑P_avail) + w_e x (E_avail_i / ∑E_avail).

[0076] Allocate device-level budget: According to the compensation participation factor, the total compensation bandwidth budget is allocated to each device. The initial budget of the i-th device is Budget_i = Budget_total x Factor_i. Optionally, to cope with the situation that the budget allocated to some devices exceeds its own physical limit, an iterative algorithm of peak clipping and backfilling can be used: if the Budget_i of a device i exceeds its Rate_limit_phy_i, the excess part ΔB = Budget_i - Rate_limit_phy_i is proportionally backfilled to other unsaturated devices until the budget of all devices is within their physical constraints.

[0077] Finally, from the feasible region of the virtual impedance parameters, a set of parameters is selected (for example, R_v with the smallest additional loss or L_v that can provide the best damping is selected under the premise of meeting the constraints), together with the above-mentioned allocated device-level gain upper limit budget, and the preset second-order slope parameters and priority arbitration table, to encapsulate a data packet, i.e., a local fast control configuration set, which is issued to each local controller.

[0078] Embodiment six describes the online execution and state adaptation of the local fast control channel, and details the online application and dynamic execution process of the local fast control configuration set. It describes how the local controller of the charging pile autonomously and intelligently executes fast compensation at each local control cycle (e.g., every 10 ms).

[0079] The process is: online execution of compensation to obtain the aligned fast channel execution trajectory, as shown in detail in Figure 5 .

[0080] Communication health assessment and state machine decision: At the beginning of each control cycle, the local controller first assesses the communication quality with the master station. Communication indicators such as communication delay, jitter, and packet loss rate are collected and integrated into a communication reliability score; according to the communication reliability score and the preset threshold, the state machine state label is determined, which includes any one of normal, step loss, shape preservation, and derating.

[0081] The controller maintains a communication quality monitor, which computes the mean delay and jitter by analyzing the time stamps of received messages, and the packet loss rate by examining the message sequence numbers. An exemplary scoring function is Score = w_d x Delay_norm + w_j x Jitter_norm + w_l x Loss_rate. According to the value of Score, the controller switches among four states: Normal: Score < θ_normal, communication is good, all functions are enabled.

[0082] Out-of-sync: θ_normal < Score < θ_lost, communication is slightly deteriorated, at which point the controller will actively adapt the gain of the power compensation according to the state tag of the state machine, e.g., halving the compensation gain, and operate in a conservative mode.

[0083] Preserving: Score > θ_lost, or consecutive packet loss, communication is severely deteriorated. The controller enters the preserving state, freezes the last received partition execution reference, and continues to execute local compensation based on it, maintaining the power waveform, and waits for communication recovery.

[0084] Degraded: long-term communication interruption or internal controller failure, enters a safe degraded or shutdown state.

[0085] Budgeted compensation and online passive preserving: in the normal or out-of-sync state, the controller measures the deviation of the local bus frequency and voltage, and calculates an original power compensation instruction according to the compensation gain (which has been adaptively adjusted in the out-of-sync state).

[0086] Virtual impedance superposition: the controller superimposes a virtual impedance configuration to the power compensation control loop. This means that before generating the final PWM drive signal, the control algorithm simulates the effect of the virtual resistance R_v and virtual inductance L_v configured in embodiment five.

[0087] Online passive margin evaluation: at the same time, based on the equivalent impedance frequency point set of the grid, the passive margin of the current period is evaluated online, and a passive compliance flag is generated. The controller uses the Z_grid(jω) data received from the master station last time, combined with its own current Z_device(jω), to calculate Re{Z_device+Z_grid} in real time. If the value at any frequency point is lower than the margin δ, an uncompliant flag is generated.

[0088] Safety arbitration: based on the passive compliance flag, the result of power compensation is subjected to safety arbitration to form the final compensation instruction. If the flag is uncompliant, the arbitration logic will take immediate measures, such as temporarily increasing the value of the virtual resistance R_v or further reducing the compensation gain, until the passive margin is restored, thereby ensuring that the system is always running within the stability domain.

[0089] Error-constrained ramp reset: When communication recovers from a severe degradation (coasting state), the local controller receives a brand new dispatch execution reference, which can be significantly different from the current power value being executed. To avoid a jump in the command, the deviation between the slow power command and the fast channel execution trajectory is detected when the state machine state tag enters the coasting state; and based on the deviation, an error-constrained ramp reset mechanism is initiated to realign the fast channel execution trajectory to the latest slow power command with a controlled slope. That is, the controller does not immediately follow the new command, but rather smoothly transitions from the current power value to the new reference value with a ramp that is constrained by an upper limit on the power change rate (determined by the budget of embodiment five).

[0090] Coupled execution of the perturbation and online identification: To enable continuous system characteristic monitoring, the method injects an orthogonal perturbation sequence into the final compensation command, where the orthogonal perturbation sequence has the characteristics of zero mean, bounded amplitude, and low cross-correlation with the perturbation sequences of other devices below a preset threshold. For example, each controller is assigned a unique pseudo-random binary sequence as its perturbation signal, with an amplitude limited to within 5% of the compensation command. Furthermore, the injection of the orthogonal perturbation sequence, the error-constrained ramp reset mechanism, the superposition of the virtual impedance configuration, and the application of the device-level gain upper limit budget are all coupled to be executed under the unified scheduling of the state machine state tag. This is a highly integrated control logic: for example, the perturbation injection is only activated in the normal state; the virtual impedance remains in the normal and step-out states, but can be adjusted to a more conservative value in the coasting state; and the ramp reset mechanism is triggered when exiting the coasting state. This coupled design ensures that various advanced functions work safely and coordinately in complex communication and power grid environments.

[0091] Finally, the power command formed after all the above processes is the aligned fast channel execution trajectory, which is sent to the underlying control loop of the converter for execution and uploaded to the master station for subsequent fusion arbitration.

[0092] Embodiment seven describes the fusion arbitration of the fast and slow channels and the cross-cycle adaptive correction process, explaining how to safely and intelligently fuse the low-speed planned power trajectory (slow channel) with the high-speed dynamic compensation trajectory (fast channel) in the final execution phase of the aggregation control method, and how the system achieves cross-cycle self-learning and optimization through evaluation of the execution results.

[0093] In this embodiment, the core step of the method is to fuse the dispatch execution reference, the local fast control configuration set, and the aligned fast channel execution trajectory, and to perform priority arbitration under the constraint of the power hard constraint mapping to output the final execution power trajectory.

[0094] Fusion and priority arbitration: At each control cycle, the aggregation controller combines the slow reference (P_ref(t)) obtained from embodiment three with the aligned fast trajectory (P_fast(t)) obtained from embodiment six.

[0095] The resulting synthetic power command P_syn(t) is the algebraic sum of both: P_syn(t) = P_ref(t) + P_fast(t). At the same time, the rate of change dP_syn(t) / dt is computed. Subsequently, the controller checks in real time whether the synthetic power command exceeds the boundaries of the power hard constraints map. The boundaries here refer to the time-varying upper limits on power peak and power rate of change generated in embodiment three.

[0096] When a boundary is detected to be exceeded by P_syn(t) or dP_syn(t) / dt, the arbitration mechanism is activated. If a boundary is exceeded, the slow reference is clipped according to the priority rule that frequency support has the highest priority, voltage support the second priority, and trajectory tracking the third priority, to ensure that the fusion result respects the power hard constraints map.

[0097] The fast trajectory P_fast(t) can itself be decomposed into P_fast(t) = P_freq(t) + P_volt(t), where P_freq(t) is the compensation component for frequency deviation and P_volt(t) is the compensation component for voltage deviation.

[0098] Arbitration logic: When a conflict occurs, the controller performs the reduction or clipping of the commands in the following order: First, the trajectory tracking component with the lowest priority, i.e. the slow reference P_ref(t), is attempted to be reduced. For example, if dP_syn(t) / dt is out of limits, the rate of change of P_ref(t) is limited, or even temporarily reversed, to make room for the higher priority compensations.

[0099] If reducing P_ref(t) alone is not sufficient to bring the synthetic command back within the boundaries, the controller starts reducing the voltage support component P_volt(t) with the second lowest priority.

[0100] The frequency support component P_freq(t) with the highest priority is preserved to the maximum extent, and is only reduced in extreme cases.

[0101] This priority rule embodies the highest principle of guaranteeing the overall security and stability of the power grid. The stability of the power grid frequency is the first priority of system security, followed by the stability of the local voltage, and the priority of the economic operation plan of the port area itself (reflected in P_ref(t)) is the lowest. Through this clear arbitration mechanism, it is ensured that in any emergency, the virtual energy storage system can first and maximally provide support to the power grid that is most beneficial, while never breaking the power hard constraint boundary determined by its own physical characteristics, achieving the goal of both strong external support and absolute internal safety.

[0102] In order to make the system have the ability of self-learning and self-optimization, the application further comprises: comparing the final execution power trajectory with the segmented execution reference, and combining the three-frequency band energy index of the execution period, an evaluation index set is calculated and obtained; according to the evaluation index set, a budget adjustment amount, a virtual impedance fine-tuning amount, and a window and buffer band correction amount for the next period are generated; the local fast control configuration set is updated using the budget adjustment amount and the virtual impedance fine-tuning amount, and the polarity maintaining window and buffer band parameters are updated using the window and buffer band correction amount.

[0103] On a longer time scale (e.g. every 15 minutes), the evaluation module of the master station is started once.

[0104] Calculate the evaluation index set: this index set is a set of numerical values that quantify the performance of the system, such as: the integral value of the frequency / voltage deviation, the number of times the arbitration mechanism is triggered, the peak and average values of the three-frequency band energy index, the root mean square value of the DC ripple, and the tracking error between the final execution power trajectory and the segmented execution reference, etc.

[0105] Generate correction amount: a decision engine based on expert rules or fuzzy logic generates correction amounts for key control parameters according to the evaluation index set. For example: If the evaluation finds that the arbitration mechanism is frequently triggered due to exceeding the power change rate constraint, the decision engine will generate a negative budget adjustment amount to appropriately lower the total compensation bandwidth budget allocated in Embodiment Five in the next period, so that the behavior of the fast channel is more convergent.

[0106] If the evaluation finds that even within the hard constraint, the energy index of a certain frequency band is still consistently high, a positive virtual impedance fine-tuning amount may be generated to slightly increase the virtual impedance value corresponding to that frequency band in the next period, to provide stronger damping.

[0107] If the evaluation finds that the actual process of power zero-crossing still accompanies energy spikes, a positive window and buffer band correction amount will be generated to further widen the adaptive window parameters in Embodiment Four in the next period.

[0108] This closed-loop correction mechanism builds a top-level learning loop across multiple core modules. It enables the system to learn from its own historical performance, automatically discover and correct potential parameter mismatches or suboptimal configurations, and continuously converge to a safer, more efficient, and more stable operating point without human intervention, embodying a high degree of intelligence.

[0109] Embodiment eight describes a timing offset-based group coordination control process, which can replace the passive constraint-based online execution compensation method in embodiment six, and is particularly suitable for scenarios with limited communication bandwidth or lower requirements for controller computing power.

[0110] In this embodiment, the step of online execution compensation obtaining the aligned fast lane execution trajectory is replaced by: based on the rated power and maximum power change rate of each device in the aggregated scene and resource data, calculating the response time constant, and dividing the devices into at least two response groups according to the response time constant; determining the power instruction issuance timing offset interval between the at least two response groups, and issuing instructions to each response group in turn according to the timing offset interval.

[0111] Dynamic grouping of devices based on response capability: First, calculate the response time constant τ_i for each available charging pile i: τ_i = P_rated_i / (dP / dt)_max_i; where P_rated_i is the rated power and (dP / dt)_max_i is the maximum power change rate. The smaller the τ_i value, the faster the device responds. Then, according to the value of τ_i, use clustering algorithms (such as K-Means) or simple threshold methods to divide all devices into several response groups. For example, devices with τ < 50ms are divided into a fast response group, devices with 50ms ≤ τ < 200ms are divided into a medium response group, and devices with τ ≥ 200ms are divided into a slow response group.

[0112] This grouping method classifies a large number of heterogeneous devices into several homogeneous groups with similar response characteristics. This zero-to-one approach greatly simplifies the complexity of control, allowing the controller to manage several groups of consistent behavior rather than independently scheduling hundreds or thousands of devices.

[0113] Determination of timing offset interval and sequential issuance: When compensation of total power P_total is needed, the power is allocated to each response group. The controller plans the starting sequence for each group, usually fast response group first. To avoid huge impact on the grid caused by all devices acting at the same time, the starting time of each group is artificially staggered by a time interval Δt_group. The interval can be calculated based on the allowed voltage change rate of the bus, for example Δt_group=P_step_group / (dV / dt)_allow; where P_step_group is the power step generated by the starting of a response group, and (dV / dt)_allow is the allowed voltage change rate of the bus defined in the device constraints. The controller issues the instructions in the sequence fast response group -> Δt_group -> medium response group -> Δt_group -> slow response group.

[0114] The core idea of the method is to change the simultaneous into sequential, by staggering in time, to decompose a large power impact into multiple small, bearable power steps, so as to realize the smoothing of the system dynamic process in a programmed, open-loop scheduling mode without relying on complex stability control (such as virtual impedance).

[0115] Further, in order to introduce feedback regulation to this open-loop scheduling, the embodiment further includes: monitoring the bus voltage change caused by the starting of any response group, and dynamically adjusting the time interval of the subsequent response group according to the bus voltage change; and when it is monitored that the actual response power of any response group is lower than the planned value, calculating the compensation time and accordingly advancing the instruction issuing time of the next response group.

[0116] Interval adaptive adjustment: for example, after the starting of the fast response group, the controller monitors the actual bus voltage sag ΔV_measured. If ΔV_measured is much smaller than the allowed value, the controller will dynamically shorten Δt_group before the starting of the next group to speed up the overall response speed. On the contrary, if ΔV_measured is close to the limit value, Δt_group will be extended.

[0117] Compensation advance mechanism: if the controller monitors that the actual power P_actual emitted by the fast response group within the specified time is lower than the planned value P_plan, it will calculate the power shortage ΔP=P_plan-P_actual, and estimate the time Δt_comp needed for the next response group (medium response group) to make up for this shortage. Then, the originally scheduled starting time of the medium response group is advanced by Δt_comp to ensure the timeliness of the total power response.

[0118] Embodiment eleven describes the core algorithm in embodiment three, that is, how to generate a specific power hard constraint value from the system state and constraints.

[0119] Set input parameters and boundary conditions: Assume at a certain time t, the aggregation control master station obtains the following input data: The DC bus rated voltage is 750V, and the DC ripple limit V_lim defined in the equipment constraints is peak-to-peak value not more than 5% of the rated voltage, i.e. V_lim=750V x 5%=37.5V.

[0120] Through online identification (such as Example Three), the gain norm ||G_P→V||_b of the power-to-DC ripple transfer characteristic in three key frequency bands under the current working condition is obtained as follows: Low frequency band (b=1): ||G_P→V||_1=0.1V / kW; Twice grid frequency band (b=2): ||G_P→V||_2=0.5V / kW; Switching sideband (b=3): ||G_P→V||_3=0.2V / kW; Through real-time calculation (such as Example Three), the current three-frequency band energy indicators E_b are obtained as follows: Low frequency band (b=1): E_1=4.0(kW)²·s.

[0121] Twice grid frequency band (b=2): E_2=1.0(kW)²·s.

[0122] Switching sideband (b=3): E_3=0.25(kW)²·s.

[0123] The safety factor α_b of all frequency bands is uniformly set to 0.8.

[0124] The prevention of zero small positive number ε is set to 10^-6.

[0125] Step-by-step calculation process: According to the formula P_allow,b=α_b (V_lim / ||G_P→V||_b) (1 / sqrt(E_b+ε)) in Example Three.

[0126] Step one: Calculate the basic power margin V_lim / ||G_P→V||_b of each frequency band; Low frequency band margin = 37.5V / 0.1V / kW = 375kW; Twice grid frequency band margin = 37.5V / 0.5V / kW = 75kW; Switching sideband margin = 37.5V / 0.2V / kW = 187.5kW.

[0127] Step two: Calculate the energy suppression factor 1 / sqrt(E_b+ε) of each frequency band

[0128] Low band factor = 1 / sqrt(4.0+10^-6)≈0.500.

[0129] Two times grid band factor = 1 / sqrt(1.0+10^-6)≈1.000.

[0130] Switching band factor = 1 / sqrt(0.25+10^-6)≈2.000.

[0131] Step three: Calculate the independent allowed power peak value P_allow,b of each band combined with the safety factor.

[0132] Low band allowed value P_allow,1 = 0.8x375kWx0.500 = 150kW.

[0133] Two times grid band allowed value P_allow,2 = 0.8x75kWx1.000 = 60kW.

[0134] Switching band allowed value P_allow,3 = 0.8x187.5kWx2.000 = 300kW.

[0135] Step four: Determine the final power hard constraint

[0136] The aggregation control master station takes the minimum value of all band allowed values as the final power peak hard constraint P_hard_limit at the current time.

[0137] P_hard_limit = min(P_allow,1, P_allow,2, P_allow,3) = min(150, 60, 300) = 60kW.

[0138] Through the above calculation, the power peak hard constraint generated by the system at this time t is 60kW. As can be seen, although the system can still withstand power fluctuations of more than 150kW in the low band and switching band, since the transfer gain of the current two times grid band is the highest (i.e. the system is most sensitive to power disturbance in this frequency band), and there is already a certain energy accumulation in this band, therefore the final power hard constraint is determined by the calculation result of this band. This calculation result will serve as an insurmountable red line for the fusion arbitration link in embodiment seven, to ensure that any power instruction will not cause the DC bus ripple to exceed the limit, thereby guaranteeing the power quality and safety and stability of the system.

[0139] Embodiment nine describes an adaptive grouping and customized control process based on charging behavior characteristics, rather than treating all charging piles as homogeneous resources. Instead, by deeply mining their operation data, their unique behavior portraits are identified, and based on this, dynamic grouping and differentiated and refined collaborative control are carried out.

[0140] In the embodiment, the step of obtaining the aggregated scenario and resource data and generating the split execution reference further comprises: extracting, for each device, a behavior feature vector containing a power change rate standard deviation and a power ripple ratio based on the power curve in the synchronization dataset; and dynamically grouping the devices according to the similarity of the behavior feature vectors to form a hierarchical group structure, wherein the dynamic grouping includes merging, splitting or dissolving of the groups.

[0141] Online behavior feature extraction and update: The aggregation control master station maintains a dynamically updated behavior feature vector F_i for each online charging pile i. For example, in a 5-minute rolling time window, the master station extracts the power curve of the charging pile from the synchronization dataset and calculates the following features: Power change rate standard deviation σ_P: std(dP / dt), reflecting the smoothness of the device power output. The value is higher for devices that frequently start and stop and have large power fluctuations.

[0142] Power ripple ratio RPP: (P_max-P_min) / P_avg, reflecting the size of the power ripple of the device in steady state operation.

[0143] State transition number N_trans: The number of times the device switches between charging, discharging, standby and other states, reflecting the intermittency of its tasks.

[0144] Charging efficiency η: P_dc·dt / P_ac·dt, reflecting the energy conversion efficiency of the device.

[0145] Battery state SOC and rated power P_rated: included as static or slowly changing features.

[0146] These features together constitute a multi-dimensional behavior feature vector F_i=[σ_P, RPP, N_trans, η, SOC, P_rated]. To ensure the real-time and stability of the features, an exponential weighted average method can be used to update the vector smoothly: F_i(t)=β×F_i(t-1)+(1-β)×F_i_new; where β is the forgetting factor, for example, take 0.8. This allows the feature vector to smoothly reflect the recent behavior pattern of the device.

[0147] Hierarchical dynamic clustering: Based on the above feature vectors, the system adopts a hierarchical clustering strategy. First, a coarse-grained hierarchical clustering is performed according to the rated power P_rated, for example, the devices are divided into a high-power fast-charging group (such as > 50kW), a medium-power slow-charging group (such as 7-22kW), and a V2G bidirectional group. Subsequently, within each large level, a density-based clustering algorithm (such as DBSCAN) is used to perform fine-grained dynamic clustering according to the similarity (for example, Euclidean distance) of other dynamic behavior features (such as σ_P, RPP).

[0148] This clustering is dynamic, and the system periodically monitors the compactness and separation of each group. When the center points of two groups are less than the merging threshold θ_merge, they will be merged; when the internal variance of a group is greater than the splitting threshold θ_split, it will be split into two or more more compact subgroups; when the number of members of a group is too small (such as less than 2), the group will be dissolved, its members become discrete points or be reassigned. This dynamic management mechanism ensures that the clustering result always accurately reflects the current state of the system.

[0149] On this basis, for each group in the hierarchical group structure, a set of customized control strategies is matched according to its behavior characteristics, forming a group customized control strategy set; wherein the customized control strategy at least includes configuring different power ripple limits or power margins for different groups, and generating a segmented execution reference according to the group customized control strategy set.

[0150] The system tailors the regulation and control strategy for groups with different behavior characteristics. For example: For dynamic task type fast charging groups with high σ_P and high N_trans (usually serving high frequency operation vehicles such as AGVs), the system matches a fast response strategy for them. This strategy allows it to have a larger power ripple (for example, allowing a ±5% ripple) when participating in frequency regulation, and gives it a higher compensation participation factor to fully utilize its fast response capability.

[0151] For static energy storage type V2G groups with low σ_P and low N_trans (usually serving long-parked employee vehicles or energy storage dedicated piles), the system matches a smooth support strategy for them. This strategy imposes strict ripple limits (for example, no more than ±2%) on its power output, and reserves a large power margin (for example, always maintains a 20% power reserve) for it, making it the main force to provide stable, reliable voltage support or peak shaving services.

[0152] For suspected abnormal groups with high RPP, the system matches a conservative observation strategy for them, temporarily reduces their participation in all grid services, and pushes a diagnostic alarm to the operation and maintenance system.

[0153] In generating the final disaggregated execution reference, the aggregation regulation master station no longer simply allocates the total power target among all devices, but adopts a hierarchical allocation mode: first, the total power target (e.g. total frequency modulation capacity, total peak clipping power) is allocated among groups according to their customized strategies. Then, within each group, the task amount allocated to the group is further decomposed to each device in the group according to the group's internal coordination rules (e.g. considering SOC balance).

[0154] Through such adaptive grouping and customized regulation, the system can achieve fine and differentiated management of massive heterogeneous resources. It goes beyond the rough aggregation based on device nameplate parameters, and instead utilizes data-driven behavioral insights to match tasks to appropriate resource groups, thereby achieving higher control efficiency, stronger system stability, and longer device service life in the macroscopic sense.

[0155] Embodiment Ten describes the generation of a safety boundary based on dynamic power envelope, i.e. provides an optional solution to generating the power hard constraint mapping. Instead of relying on complex system identification and frequency domain analysis, a data-driven method based on historical data statistics and future periodic prediction is adopted to construct a dynamic and adaptive safe operation envelope.

[0156] In this embodiment, the step of generating the power hard constraint mapping is replaced by: based on historical execution data within a rolling time window, constructing a statistical power envelope containing power standard deviation; using the daily periodicity and weekly similarity of historical execution data, constructing a predicted power envelope; and weighting and fusing the statistical power envelope and the predicted power envelope to form the power hard constraint mapping.

[0157] Construction of statistical and predicted envelope: Statistical power envelope: the aggregation regulation master station maintains a rolling power history database, e.g. the past 24 hours. Within a shorter sliding time window T (e.g. 15 minutes), the maximum max(P) and minimum min(P) of historical power P(t) within the window are calculated, as well as the standard deviation σ_P. Based on this, the upper envelope P_env_up_stat(t) = max(P(t-T:t)) + k_σ × σ_P(t) and the lower envelope P_env_low_stat(t) = min(P(t-T:t)) - k_σ × σ_P(t) of statistical power are constructed; where k_σ is the standard deviation coefficient (e.g. take 1 or 2) for controlling the width of the envelope. This envelope reflects the recent actual operation range and fluctuation intensity of the system.

[0158] Considering the strong daily and weekly periodicity of port operations, the system utilizes this prior knowledge to construct the prediction model. A simple yet effective model is: P_pred(t) = P_hist(t-24h) x k_day x k_week. Where P_hist(t-24h) is the historical power value at the same time 24 hours ago; k_day and k_week are the daily and weekly correction factors (e.g., to account for the load difference between weekdays and weekends). Based on this prediction curve, a prediction power envelope can also be constructed.

[0159] The final power envelope is a weighted fusion of the above two: P_env_final = w_stat x P_env_stat + w_pred x P_env_pred. For example, the weights can be set as w_stat = 0.6 and w_pred = 0.4. The purpose of the fusion is to take into account both the actual short-term fluctuations of the system (reflected by the statistical envelope) and the long-term operating rules (reflected by the prediction envelope). This fused P_env_final is then used as the basic power hard constraint mapping.

[0160] Further, to make the envelope more intelligent and adaptive, the embodiment also introduces dynamic margins and scenario management.

[0161] First, a dynamic safety margin is calculated and applied to adjust the power hard constraint mapping; where the amplitude of the dynamic safety margin is proportional to the distance between the current real-time power value and the envelope boundary defined by the power hard constraint mapping.

[0162] At any time t, the safety margin Margin(t) is dynamically calculated, for example: Margin(t) = M_base x (1 + k_m x dist(P_actual(t), P_env_final(t)) / Bandwidth(t)); where M_base is the base margin; k_m is the gain coefficient; P_actual(t) is the current real-time power; dist(...) is the distance of the current power to the nearest envelope boundary; Bandwidth(t) is the total width of the current envelope. This dynamic margin can be used to adjust the gain of the controller or the allowed compensation range. When the real-time power P_actual(t) is near the center of the envelope, dist(...) is maximum, Margin(t) is also maximum, and the system is allowed to have a larger amplitude of fast response. When P_actual(t) approaches the envelope boundary, dist(...) decreases, and Margin(t) also tightens, thereby automatically suppressing aggressive control behavior that may lead to a breach.

[0163] Secondly, based on the current time and real-time power value information, the current operating scenario is identified; and according to the operating scenario, the corresponding scenario-related parameters such as envelope width and safety factor are extracted from the scenario parameter library to configure the power hard constraint mapping; and the envelope utilization rate of the power hard constraint mapping is periodically calculated, and the parameters in the scenario parameter library are adjusted across cycles based on the envelope utilization rate to achieve self-learning.

[0164] The system has a built-in scene recognition module that matches the system status to a preset scene library based on the current time (e.g., 7:00-9:00), total power level, number of online devices, etc., such as morning peak operation, midday stable charging, evening peak V2G support, and nighttime off-peak maintenance.

[0165] Each scenario is associated with a set of optimal envelope generation parameters in the parameter library (such as statistical window length T, fusion weight w_stat, base margin M_base, etc.). Once the current scenario is identified, the system loads the corresponding parameter set to generate a customized power envelope for that scenario.

[0166] After each scenario ends (e.g., after the daily morning rush hour scenario ends), the system calculates the envelope utilization rate for that period, i.e., Utilization = |P_actual|dt / P_env_finaldt. The system uses an ideal utilization target (e.g., 85%) as a benchmark. If the actual utilization is too low (e.g., <70%), it indicates that the envelope is too loose, and the system will automatically tighten the parameters for the same scenario the next day (e.g., reduce k_σ). If the utilization is too high (e.g., >95%), it indicates that the envelope is too tight, limiting the normal operation of the system, and the parameters will be automatically relaxed. Through this daily fine-tuning, the system can autonomously learn and optimize the power envelope for each scenario, making it both safe and efficient.

[0167] Example 12 provides a refined and enhanced implementation method for the data aggregation and subsequent instruction issuance process in Example 1.

[0168] Firstly, in the data preprocessing stage, in order to improve the accuracy of the subsequent algorithm, the embodiment further introduces a weighted processing mechanism based on noise spectrum on the basis of the cleaning step of embodiment one. Specifically, the aggregation regulation master station will conduct offline or online statistical analysis on the historical collected synchronous data set of each channel to estimate the noise power spectral density of each measurement channel (for example, the bus voltage of a certain area, the power feedback of a certain type of charging pile). As can be seen, some channels may exhibit greater noise in a certain frequency band due to factors such as sensor quality, electromagnetic environment, etc. When performing data-dependent calculations in the subsequent, for example, in embodiment three, the identification of power to DC ripple transmission characteristics, or in embodiment four, the extracted trend, the algorithm will use this noise spectrum information as a weighting factor. For data channels with low noise levels in key frequency bands, the weight in the calculation is higher, and vice versa. In this way, noise interference can be effectively suppressed, and the identified model and extracted trend are closer to the real physical process.

[0169] Secondly, in terms of instruction issuance and execution, to solve the problem of synchronization loss of instruction execution caused by inconsistent network delays in distributed systems, the embodiment adopts a collaborative control mechanism based on effective time stamp. Specifically, the time synchronization strategy module in the aggregation regulation master station arranges a unified effective time for all slow instructions and parameter packages that need to be synchronized. For example, when the master station completes a calculation and generates a new local fast control configuration set (from embodiment five) and a pile execution reference (from embodiment three), it does not immediately issue and execute it. Instead, the master station will encapsulate these two data packages and attach a unified, future effective time stamp to them, for example, September 2, 2025, 14:30:00.500. The local controller of each charging pile will store the time-stamped instruction package in the buffer and continue to execute the old instructions when it receives the time-stamped instruction package. When the local controller's own GPS / Beidou synchronized clock reaches the exact time of the effective time stamp, it will instantly and atomically enable the new configuration set and execution reference. It can be understood that through this mechanism, even if the instruction packages arrive at different controllers in different orders due to different network paths, all devices can eventually switch to the new working state at the same time, ensuring the precise synchronization of the entire port area virtual energy storage resources in control logic and power target, and avoiding system transient oscillation that may be caused by inconsistent states.

[0170] Embodiment thirteen, this embodiment aims to further describe several key details in the core control algorithm of the invention, and to explain the robustness strategy designed to ensure that the system can still operate stably under non-ideal or abnormal working conditions.

[0171] In the online compensation execution of embodiment six, a ramp controller is commonly used to limit the power change rate. In a preferred implementation of this embodiment, a second-order power ramp is used. Compared with the traditional first-order ramp (power changes linearly with a constant rate), the power output curve of the second-order ramp is S-shaped. This means that the power change rate first increases smoothly from zero and then decreases smoothly to zero, and the change rate of the change rate (i.e., the power acceleration) is limited. In a discrete implementation, the first-order ramp only limits the difference between the target and the current output, while the second-order ramp limits the output again on the basis of the first-order ramp. This can significantly reduce the high-frequency components in the command change, avoid unnecessary switching stress on power electronic devices, and reduce high-frequency noise injected into the power grid, which is beneficial for improving power quality and prolonging equipment life.

[0172] In the compensation budget allocation link of embodiment five, to ensure that the budget allocated to each device does not exceed its physical limit, this embodiment uses an iterative allocation algorithm with peak clipping and backfilling. The specific process is as follows: first, the total compensation budget is allocated initially according to the compensation participation factors of each device. Then, check whether the allocated budget of any device exceeds its maximum power change rate limit. If so, add all the excess parts together to form a backfilling pool. In the next iteration, the budget in the backfilling pool is redistributed according to the relative proportion of the participation factors of all unsaturated (i.e., not reaching its physical upper limit) devices. Repeat this process until the backfilling pool is empty or all devices are saturated. For example, the total budget is 100 kW / s, the participation factors of devices A, B, and C are 0.5:0.3:0.2, and the physical upper limit of B is 25 kW / s. The initial allocation is A=50, B=30, and C=20. B exceeds the upper limit by 5 kW / s. This 5 kW / s is placed in the backfilling pool and redistributed according to the participation factor proportion of A and C, which is 5:2. A increases by 5x(0.5 / (0.5+0.2))≈3.57, and C increases by 5x(0.2 / (0.5+0.2))≈1.43. The final allocation result is approximately A=53.57, B=25, and C=21.43. This refined allocation ensures that the total budget is fully utilized while strictly respecting the physical constraints of each device.

[0173] In the process of grouping in Embodiment Three, in order to improve efficiency and control pertinence, this embodiment introduces the intermediate steps of only charging / only discharging group solving and intra-group power corridor shaping. When the total power target is net discharging, the system does not require all vehicles to reduce charging, but through an optimization solver, a part of vehicles with high SOC and allowed to discharge are selected to form a discharging-only group, and another part of vehicles are selected to form a charging-only group, so that the algebraic sum of the power of the two groups is exactly equal to the total power target. This can reduce unnecessary bidirectional energy flow and improve the overall efficiency of the system. After forming the groups, the system designs a dynamic power corridor (i.e. power upper and lower limits and change rate upper and lower limits) for each group, which takes into account the average SOC, maximum battery temperature, and estimated departure time of all members in the group. The final power allocation is carried out under the constraints of the corridor.

[0174] Fallback strategy under abnormal conditions: to improve the robustness of the system, an abnormality handling and fallback path is designed.

[0175] Optionally, in the power grid impedance identification link of Embodiment Five, if the signal-to-noise ratio of the response signal collected after the disturbance injection is too low, resulting in a confidence level of the identification result lower than a preset threshold (e.g. coherence less than 0.7), the system will discard this identification result and automatically load a set of default virtual impedance parameters stored locally and suitable for most working conditions, while sending an impedance identification failure alarm to the upper system.

[0176] Optionally, in the power hard constraint mapping calculation link of Embodiment Three, if the calculated power change rate upper limit is lower than the minimum controllable step size physically achievable by all devices, it indicates that the current system state is very fragile. At this time, the system will trigger a linkage protection, on the one hand, it will forcibly extend the polarity retention window and widen the buffer band through the mechanism of Embodiment Four, on the other hand, it will actively impose a larger slope limit on the slow channel, reducing the demand for power fluctuation from the source until the hard constraint calculation result returns to a reasonable range.

[0177] In a preferred implementation of this embodiment, the system recognizes that the power to DC ripple transmission characteristics (Embodiment Three) or the grid equivalent impedance (Embodiment Five) are not fixed, but closely related to the overall operating scenario of the port area. For example, there are multiple quaysides operating simultaneously (providing shore power) during the day and only AGVs charging in concentration at night, the system dynamic characteristics under these two scenarios are very different.

[0178] The system performs system identification for multiple typical and repeatable operating scenarios (e.g. daytime peak, night-time concentrated charging, weekend low load) through offline analysis or online learning, and establishes a version library of transfer characteristics or a version library of grid impedance. Each version is associated with a scenario label. In actual operation, a scenario identification module determines which predefined scenario the system is in according to the current time, total load level, key device operating state and other information. Once the scenario is determined, the controller loads the optimal model parameters matching the scenario from the version library for subsequent power hard constraint calculation or virtual impedance configuration.

[0179] This scenario-based adaptive mechanism enables the core algorithm of the application to always make decisions based on the most accurate system model, greatly improving the accuracy and stability of control compared to using a single, averaged model.

[0180] To facilitate system maintenance, fault diagnosis and continuous optimization, the embodiment constructs a complete closed-loop audit and data archiving system.

[0181] After each major control event (e.g. a complete grid frequency support process) or at fixed time intervals (e.g. every hour), the system automatically generates and stores an audit and reproduction instance. This instance is a highly structured data snapshot that not only contains the final execution result, but also includes the full-link decision basis leading to the result. Specifically, an instance contains at least: Input data: snapshot of synchronized data set at the time of the event.

[0182] Intermediate calculation results: set of grid equivalent impedance frequency points identified at the time, calculated power hard constraint mapping curve, generated local fast control configuration set, etc.

[0183] Instructions and outputs: issued segmented execution reference, uploaded fast channel execution trajectory of each device, and finally fused final execution power trajectory.

[0184] State and alarms: state machine label change record and any alarm information generated by all devices during the event.

[0185] By establishing such an audit mechanism, when the system exhibits any unexpected behavior, the operation and maintenance personnel can easily retrieve the corresponding reproduction instance and replay the entire event process in an offline environment, thereby quickly locating the root cause of the problem. In addition, these accumulated massive instances also constitute a valuable database that can be used for regression testing, iterative optimization of algorithms and training of more advanced AI decision models, thereby forming a complete data-decision-execution-audit-optimization closed loop, providing a solid foundation for the long-term evolution of the system.

[0186] The present application solves the problem of decoupling control target and safety boundary in the background technology by constructing a power-to-dc ripple transfer characteristic model, and calculating the energy indicators of the three key frequency bands of low frequency band, twice grid frequency band and switching sideband in real time, and then solving the problem in combination with the dc ripple limit value. Thus, the power quality safety boundary (dc ripple) in the frequency domain which is not easy to control directly is creatively transformed into a power hard constraint (instantaneous power peak and upper limit of change rate) in the time domain which can be directly executed by the controller. This completely solves the fundamental problem of decoupling control target and safety boundary in the background technology. In the actual scene of port virtual energy storage, this means that when responding to any grid dispatching instruction (such as fast frequency modulation), the power instruction issued by the aggregation regulation system naturally embeds the prediction and protection of the internal dc bus stability. Instead of relying on a static and conservative power change rate limit, the system can dynamically calculate the maximum power change capability that can be safely borne at the moment according to the current system state (reflected in the frequency band energy indicator) and system characteristics (reflected in the transfer function). Therefore, when the system state is good, it can release all the potential, provide faster and larger response, and obtain higher income; when the system is close to the instability boundary, it can instantly and accurately self-restrict to avoid triggering internal protection due to over-response, ensuring the power supply safety of key production loads in the port area, and realizing the unification of external high-performance response and internal high-reliability operation.

[0187] The present application solves the problem of decoupling control target and safety boundary in the background technology by constructing a power-to-dc ripple transfer characteristic model, and calculating the energy indicators of the three key frequency bands of low frequency band, twice grid frequency band and switching sideband in real time, and then solving the problem in combination with the dc ripple limit value. Thus, the power quality safety boundary (dc ripple) in the frequency domain which is not easy to control directly is creatively transformed into a power hard constraint (instantaneous power peak and upper limit of change rate) in the time domain which can be directly executed by the controller. This completely solves the fundamental problem of decoupling control target and safety boundary in the background technology. In the actual scene of port virtual energy storage, this means that when responding to any grid dispatching instruction (such as fast frequency modulation), the power instruction issued by the aggregation regulation system naturally embeds the prediction and protection of the internal dc bus stability. Instead of relying on a static and conservative power change rate limit, the system can dynamically calculate the maximum power change capability that can be safely borne at the moment according to the current system state (reflected in the frequency band energy indicator) and system characteristics (reflected in the transfer function). Therefore, when the system state is good, it can release all the potential, provide faster and larger response, and obtain higher income; when the system is close to the instability boundary, it can instantly and accurately self-restrict to avoid triggering internal protection due to over-response, ensuring the power supply safety of key production loads in the port area, and realizing the unification of external high-performance response and internal high-reliability operation.

[0188] The application solves the problems of how to balance multiple control targets and how to realize long-term performance evolution of the system under complex constraints by constructing fast-slow channel fusion and priority arbitration mechanism and designing a top-level design of cross-cycle adaptive correction closed loop, which is an integration and sublimation of the two core innovations at the system level. The priority arbitration mechanism ensures that the support for the grid frequency has the highest priority in any situation, followed by voltage support, and finally the tracking of internal planned power. This design enables the port virtual storage (VESS) to contribute its maximum capacity without hesitation when the grid urgently needs support, which is of great significance to improving the emergency recovery capability of the grid and preventing the situation from escalating. The cross-cycle adaptive correction mechanism continuously evaluates the actual operation effect of the system (such as arbitration trigger frequency, ripple suppression effect, etc.), and adjusts the core parameters such as the degree of conservatism of power hard constraints, the parameters of virtual impedance, and the width of the zero-crossing window. This self-learning capability enables the system to continuously accumulate experience during use, adapt to seasonal changes in load characteristics or performance drift caused by equipment aging, and maintain itself in a near-optimal operating range for a long time, thereby maximizing its technical and economic benefits while ensuring absolute safety and significantly reducing the dependence on continuous manual tuning and maintenance.

[0189] The preferred embodiments of the application are described in detail above, but the application is not limited to the specific details of the above-described embodiments. Within the technical concept of the application, various equivalent transformations of the technical solutions of the application can be made, and these equivalent transformations all belong to the protection scope of the application.

Claims

1. A method for virtual energy storage aggregation and regulation of green electricity in ports, characterized in that, Includes the following steps: Acquire aggregated scenario and resource data, which includes: a virtual energy storage synchronization dataset, a baseline power trajectory, a set of equipment constraints, and a set of available vehicle and charging facility capabilities; Based on aggregated scenario and resource data, by calculating multi-band energy indicators and identifying system power transfer characteristics, power hard constraint mapping and sub-stake execution reference are generated. The energy indicators include: power quality indicators. Based on the mapping of aggregated scenario and resource data with power hard constraints, a local fast control configuration set is generated by solving the virtual impedance feasible region and allocating the compensation budget, and the aligned fast channel execution trajectory is obtained by performing compensation online. The system integrates the sub-pile execution reference, the local fast control configuration set, and the aligned fast channel execution trajectory, and performs priority arbitration under the constraints of the power hard constraint mapping to output the final execution power trajectory of the virtual energy storage.

2. The port green electricity virtual energy storage aggregation and control method according to claim 1, characterized in that, Generate power hard constraint mapping and stub execution reference, including: Based on the DC ripple limit of the devices included in the device constraint set of aggregated scenario and resource data, the peak power and the upper limit of the rate of change of the power domain are derived in reverse, forming a power hard constraint mapping. By using power hard constraint mapping, the baseline power trajectory in the aggregated scene and resource data is constrained and allocated to generate a stub execution reference.

3. The port green electricity virtual energy storage aggregation and control method according to claim 2, characterized in that, The peak power and the upper limit of the rate of change in the power domain are derived in reverse, including: Based on aggregated scenario and resource data, power change energy is calculated in three windows: low frequency band, twice the grid frequency band, and switch sideband, to obtain energy indicators for the three frequency bands. Based on aggregated scenario and resource data, the amplitude-frequency transfer characteristics of power change to DC voltage ripple are identified, and the power to DC ripple transfer characteristics are obtained. By combining the three-band energy index with the power-to-DC ripple transfer characteristics, and using the DC ripple limit as the boundary condition, the peak power and the upper limit of the rate of change are determined.

4. The port green electricity virtual energy storage aggregation and control method according to claim 2, characterized in that, The generation of power hard constraint mapping and stub execution reference also includes: Monitor the number of zero-crossings in the polymerization power to obtain the zero-crossing count; The system aggregates low-frequency energy and insulation monitoring alarms reflected by aggregated scenario and resource data to generate near-zero risk indicators. Based on the zero-point crossing count and the zero-crossing risk indicator, the polarity maintenance window duration and buffer bandwidth used to constrain the execution reference of the sub-stakes are dynamically adjusted.

5. The port green electricity virtual energy storage aggregation and control method according to claim 4, characterized in that, Also includes: At the boundary of the polarity-maintaining window, a two-stage smooth switching is performed, in which the rate of power change is slowed down to near zero by using trend feedforward. Enable second-order ramp control to complete the zero-crossing switching of power polarity, and extend the execution duration of the second-order ramp if an increase in the energy index of either frequency band is detected during the switching.

6. The port green electricity virtual energy storage aggregation and control method according to claim 1, characterized in that, Generate a local quick control configuration set, including: Based on aggregated scenario and resource data, the equivalent impedance frequency point set of the power grid is identified by injecting power disturbance sequences and collecting responses. Based on the preset frequency domain passivity condition and combined with the equivalent impedance frequency point set of the power grid, the feasible region of virtual impedance parameters that satisfy the condition is solved. A set of parameters is selected from the feasible domain of virtual impedance parameters to form a virtual impedance configuration, which is then used as part of the local fast control configuration set.

7. The port green electricity virtual energy storage aggregation and control method according to claim 6, characterized in that, The online execution compensation yields the aligned fast-track execution trajectory, including: In each local control cycle, the virtual impedance configuration is superimposed onto the power compensation control loop; Based on the equivalent impedance frequency set of the power grid, the passivity margin of the current cycle is assessed online, and a passivity compliance mark is generated. Based on the passive compliance marking, a safety arbitration is performed on the power compensation results to form the final compensation instruction.

8. The port green electricity virtual energy storage aggregation and control method according to claim 6, characterized in that, Generating a local quick control configuration set also includes: The total compensation bandwidth budget for the entire station is determined based on the power hard constraint mapping. Based on the available capabilities of each device in the aggregated scenario and resource data, the compensation participation factor for each device is calculated. Following the compensation participation factor, the total compensation bandwidth budget is allocated to each device to form a device-level gain cap budget, which is then incorporated into the local fast control configuration set.

9. The method according to claim 7, characterized in that, The fast-track execution trajectory after online execution compensation and alignment also includes: It collects communication metrics such as communication latency, jitter, and packet loss rate, and integrates them into a communication reliability score; Based on the communication credibility score and the preset threshold, the state machine state label is determined. The state machine state label includes any one of normal, out of step, conformal, and derating. Before executing the final compensation instruction, the power compensation gain is adaptively adjusted based on the state machine state labels.

10. The port green electricity virtual energy storage aggregation and control method according to claim 1, characterized in that, The process integrates the stub execution reference, the local fast control configuration set, and the aligned fast-channel execution trajectory, and performs priority arbitration under the constraints of power hard constraint mapping, including: When integrating the stub execution reference and the aligned fast-channel execution trajectory, it is necessary to detect in real time whether the synthesized power command exceeds the boundary of the power hard constraint mapping. If the boundary is exceeded, the sub-stake execution reference is trimmed according to the priority rule of frequency support first, voltage support second, and trajectory tracking third, to ensure that the fusion result conforms to the power hard constraint mapping.

Images