Energy system intelligent collaborative operation method and device

Abstract

The specification provides an energy system intelligent collaborative operation method and device. According to the resource fluctuation prediction data in the preset time domain, the target virtual inertia is determined through the preset mapping function; according to the preset system operation cost target, the droop control coefficient of each energy storage unit is determined; according to the system energy state and the preset system energy state reference benchmark, the system energy state deviation is multiplied by the droop control coefficient of each energy storage unit to determine the power regulation amount of each energy storage unit; the system virtual inertia is used to check the target virtual inertia to obtain a verification result; according to the verification result, when the system virtual inertia is less than the target virtual inertia, the preset optimization model is used to update the droop control coefficient of each energy storage unit; and the operation benchmark of each energy storage unit is adjusted according to the updated droop control coefficient. The real-time and reliability of the system regulation response are effectively improved, and the active recovery of the system dynamic stability margin is realized.

Description

Technical Field

[0001] This specification pertains to the field of intelligent control of energy systems, and in particular relates to a method and apparatus for intelligent collaborative operation of energy systems. Background Technology

[0002] With the large-scale integration of heterogeneous units such as wind, solar, and hydrogen storage, energy systems often adopt centralized optimization scheduling methods to achieve coordinated operation. However, existing technologies suffer from insufficient real-time and reliability of regulation response, as well as difficulties in proactive recovery.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This specification provides a method and apparatus for intelligent collaborative operation of an energy system, which solves the technical problems of insufficient real-time performance and reliability of the adjustment response of the energy system during collaborative operation, and the difficulty in restoring the dynamic stability margin of the system due to the inability to actively adjust the operating reference based on the virtual inertia kernel result of the system.

[0005] This specification provides a method for intelligent collaborative operation of an energy system, including: Acquire operational data of each energy storage unit within the energy system, as well as resource volatility prediction data within a preset time domain; Based on the operational data, the corresponding system energy state and system virtual inertia are determined respectively; wherein, the virtual inertia is used to characterize the system dynamic stability margin; Based on the resource volatility prediction data within the preset time domain, the target virtual inertia is determined through a preset mapping function; based on the preset system operating cost target, the droop control coefficient of each energy storage unit is determined. Based on the system energy state and the preset system energy state reference benchmark, the system energy state deviation is determined; the system energy state deviation is multiplied by the droop control coefficient of each energy storage unit to determine the power adjustment amount of each energy storage unit, so as to drive each energy storage unit to perform autonomous and coordinated adjustment. During the autonomous and coordinated adjustment process, the target virtual inertia is verified using the system's virtual inertia to obtain the verification result; Based on the verification results, when the virtual inertia of the system is less than the target virtual inertia, the droop control coefficient of each energy storage unit is updated using a preset optimization model; wherein, the preset optimization model includes at least an inertia recovery constraint. The operating reference of each energy storage unit is adjusted according to the updated droop control coefficient; wherein, the operating reference is the target output base point of each energy storage unit in the energy balance state, and the operating reference is used to adjust the state of charge of the energy storage unit so that the virtual inertia of the system is restored to a level not lower than the target virtual inertia.

[0006] In one embodiment, determining the corresponding system energy state and system virtual inertia based on the operational data includes: The operational data is normalized to obtain multiple normalized energy storage state variables; wherein, the operational data includes the state of charge data of the electrochemical energy storage unit, the hydrogen storage pressure data of the hydrogen energy storage unit, and the demand response potential data of the adjustable load energy storage unit. The corresponding system energy state is determined by weighting and summing the multiple normalized energy storage state variables using preset weighting coefficients. Based on the operating data of each energy storage unit and the preset inertia constant, calculate the state-dependent inertia contribution factor that each energy storage unit can provide in the current state. The inertia contribution factors of each energy storage unit are linearly summed and aggregated to determine the corresponding system virtual inertia.

[0007] In one embodiment, determining the target virtual inertia based on the resource volatility prediction data within the preset time domain using a preset mapping function includes: Obtain meteorological forecast data within the prediction time domain, and determine the total power output prediction curve within the prediction time domain based on the meteorological forecast data; Calculate the standard deviation of the total output prediction curve in the prediction time domain, and determine the standard deviation as the resource volatility prediction data; The standard deviation is substituted into the preset mapping function as the independent variable, and the target virtual inertia is calculated and determined using the preset mapping function. The preset mapping function is a monotonically non-decreasing function, so that the target virtual inertia increases as the resource volatility prediction data increases.

[0008] In one embodiment, substituting the standard deviation as an independent variable into the preset mapping function and using the preset mapping function to calculate and determine the target virtual inertia includes: The target virtual inertia is determined according to the following formula:

[0009] in, The target virtual inertia, To predict the length of the time domain, The length of the prediction time domain starting from time t The standard deviation of the total output forecast within, It is a monotonically non-decreasing function.

[0010] In one embodiment, determining the droop control coefficient of each energy storage unit based on a preset system operating cost target includes: Construct a multi-objective optimization function with the total cost of the energy system in the prediction time domain as the optimization objective; wherein, the total cost includes the operating cost of each energy storage unit, the degradation cost of the equipment, and the compensation cost of demand response load; Establish the constraints of the multi-objective optimization function; wherein the constraints include at least the power regulation range limit term, the power ramp rate limit term, and the operating safety range limit term for each energy storage unit; The multi-objective optimization function that satisfies the constraints is optimized using a preset optimization algorithm to determine the droop control coefficient of each energy storage unit. Based on the constraints, the multi-objective optimization function is optimized and solved to determine the droop control coefficient of each energy storage unit.

[0011] In one embodiment, updating the droop control coefficient of each energy storage unit using a preset optimization model includes: Based on the system energy state corresponding to each energy storage unit and the inertia contribution factor, the stability adjustment priority of each energy storage unit is determined. Based on the stability adjustment priority, determine the weight coefficients in the objective function of the preset optimization model; The target virtual inertia is substituted as a parameter into the inertia recovery constraint in the preset optimization model so that the sum of the inertia provided by each energy storage unit is greater than or equal to the target virtual inertia. Using an optimization algorithm, under the premise of satisfying the inertia recovery constraint, the objective function containing the weight coefficients is optimized to determine the correction amount of each droop control coefficient; Based on the correction amount of each of the aforementioned droop control coefficients, the updated droop control coefficients are determined.

[0012] In one embodiment, adjusting the operating baseline of each energy storage unit according to the updated droop control coefficient includes: The baseline adjustment amount corresponding to each of the energy storage units is determined based on the updated droop control coefficient; wherein, the energy storage unit includes an electrochemical energy storage unit and a hydrogen energy storage unit; Based on the reference adjustment amount corresponding to the electrochemical energy storage unit, the power operation reference of the electrochemical energy storage unit is shifted from the output point corresponding to the current state of charge to the preset energy state equilibrium point, so as to correct the power adjustment space of the electrochemical energy storage unit. Based on the reference adjustment amount corresponding to the hydrogen energy storage unit, the operating mode reference of the hydrogen energy storage unit is switched from the current shutdown reference or restricted operating reference to the corresponding standby operating reference or fully adjustable operating reference, so as to release the inertia support potential of the hydrogen energy storage unit.

[0013] This specification provides an intelligent collaborative operation device for an energy system, comprising: The data acquisition module is used to acquire the operating data of each energy storage unit in the energy system, as well as the resource volatility prediction data within a preset time domain; The first determining module is used to determine the corresponding system energy state and system virtual inertia based on the operating data; wherein the virtual inertia is used to characterize the system dynamic stability margin. The second determining module is used to determine the target virtual inertia based on the resource volatility prediction data within the preset time domain through a preset mapping function; and to determine the droop control coefficient of each energy storage unit based on the preset system operating cost target. The power regulation module is used to determine the system energy state deviation based on the system energy state and the preset system energy state reference benchmark; and to multiply the system energy state deviation with the droop control coefficient of each energy storage unit to determine the power regulation amount of each energy storage unit, so as to drive each energy storage unit to perform autonomous and coordinated regulation. The result determination module is used to verify the target virtual inertia using the system virtual inertia during the autonomous coordinated adjustment process, and obtain the verification result. The coefficient update module is used to update the droop control coefficient of each energy storage unit using a preset optimization model based on the verification results when the virtual inertia of the system is less than the target virtual inertia; wherein, the preset optimization model includes at least an inertia recovery constraint; The reference adjustment module is used to adjust the operating reference of each energy storage unit according to the updated droop control coefficient; wherein, the operating reference is the target output base point of each energy storage unit in the energy balance state, and the operating reference is used to adjust the state of charge of the energy storage unit so that the virtual inertia of the system is restored to a level not lower than the target virtual inertia.

[0014] This specification also provides a computer-readable storage medium storing computer instructions that, when executed, implement a method for intelligent collaborative operation of an energy system.

[0015] Based on the intelligent collaborative operation method for an energy system provided in this specification, the following steps are taken: First, operational data of each energy storage unit within the energy system and resource volatility prediction data within a preset time domain are acquired. Second, based on the operational data, the corresponding system energy state and system virtual inertia are determined. The virtual inertia characterizes the system's dynamic stability margin. Third, based on the resource volatility prediction data within the preset time domain, a target virtual inertia is determined using a preset mapping function. Fourth, based on a preset system operating cost target, the droop control coefficient for each energy storage unit is determined. Fifth, based on the system energy state and a preset system energy state reference benchmark, the system energy state deviation is determined. Sixth, the system energy state deviation is multiplied by the droop control coefficient of each energy storage unit to determine the droop control coefficient of each energy storage unit. The power regulation of the energy storage units drives each energy storage unit to perform autonomous coordinated regulation. During the autonomous coordinated regulation process, the system virtual inertia is used to verify the target virtual inertia to obtain a verification result. According to the verification result, when the system virtual inertia is less than the target virtual inertia, the droop control coefficient of each energy storage unit is updated using a preset optimization model. The preset optimization model includes at least an inertia recovery constraint. The operating reference of each energy storage unit is adjusted according to the updated droop control coefficient. The operating reference is the target output base point of each energy storage unit in the energy balance state, and the operating reference is used to adjust the state of charge of the energy storage units so that the system virtual inertia is restored to a level not lower than the target virtual inertia. In this way, by using the system energy state and droop control coefficient to determine the power adjustment amount of each unit, each unit is driven to perform autonomous and coordinated adjustment, which effectively improves the real-time performance and reliability of the system adjustment response. By using the closed-loop verification of the system virtual inertia against the target virtual inertia, and adjusting the operating benchmark of each unit according to the updated droop control coefficient, it is ensured that the adjusted system virtual inertia is not lower than the target virtual inertia, thus realizing the active recovery of the system's dynamic stability margin. Attached Figure Description

[0016] To more clearly illustrate the embodiments of this specification, the accompanying drawings used in the embodiments will be briefly introduced below. The drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating an embodiment of an intelligent collaborative operation method for an energy system provided in this specification. Figure 2 This is a schematic diagram of the electronic device structure provided in one embodiment of this specification; Figure 3 This is a schematic diagram of the structural composition of an intelligent collaborative operation device for an energy system, provided in one embodiment of this specification. Figure 4 This is a flowchart illustrating another intelligent collaborative method for energy systems provided in one embodiment of this specification; Figure 5 This is a schematic diagram of a core state variable construction module provided in one embodiment of this specification; Figure 6 This is a schematic diagram of an operation rule optimization module provided in one embodiment of this specification; Figure 7 This is a schematic diagram of an autonomous collaborative control module provided in one embodiment of this specification; Figure 8 This is a schematic diagram of an inertia margin verification and reshaping module provided in one embodiment of this specification. Detailed Implementation

[0018] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0019] See Figure 1 As shown in the embodiments of this specification, an intelligent collaborative operation method for an energy system is provided, wherein the method is specifically applied to the server side. In specific implementation, the method may include the following: S101: Acquire operational data of each energy storage unit within the energy system, as well as resource volatility prediction data within a preset time domain; S102: Based on the operating data, determine the corresponding system energy state and system virtual inertia respectively; wherein, the virtual inertia is used to characterize the system dynamic stability margin; S103: Based on the resource volatility prediction data within the preset time domain, determine the target virtual inertia through a preset mapping function; based on the preset system operating cost target, determine the droop control coefficient of each energy storage unit; S104: Determine the system energy state deviation based on the system energy state and the preset system energy state reference benchmark; multiply the system energy state deviation with the droop control coefficient of each energy storage unit to determine the power adjustment amount of each energy storage unit, so as to drive each energy storage unit to perform autonomous coordinated adjustment; S105: During the autonomous and coordinated adjustment process, the target virtual inertia is checked using the system virtual inertia to obtain the verification result; S106: Based on the verification results, when the virtual inertia of the system is less than the target virtual inertia, the droop control coefficient of each energy storage unit is updated using a preset optimization model; wherein, the preset optimization model includes at least an inertia recovery constraint. S107: Adjust the operating reference of each energy storage unit according to the updated droop control coefficient; wherein, the operating reference is the target output base point of each energy storage unit in the energy balance state, and the operating reference is used to adjust the state of charge of the energy storage unit so that the virtual inertia of the system is restored to a level not lower than the target virtual inertia.

[0020] The aforementioned operational data can be physical state parameters fed back in real time by each energy storage unit, specifically including the state of charge (SOC) of electrochemical energy storage, the real-time pressure and tank capacity of the hydrogen energy storage system, and the real-time power deviation and demand response potential of the controllable load. These data constitute the raw input of the system's perception layer and are the basis for subsequent calculations of the system's global state indicators.

[0021] The aforementioned system energy state can be a scalar indicator reflecting the overall energy reserve adequacy of the integrated energy system. It is usually obtained by weighted aggregation of variables such as the normalized state of charge and pressure of each unit.

[0022] The aforementioned virtual inertia can be a technical indicator characterizing the ability of an energy system to maintain its stability when subjected to dynamic disturbances. Its value depends on the current operating point of each power electronic device in the system, the available kinetic energy reserves, and the response speed of the control algorithm.

[0023] The aforementioned resource volatility prediction data can be a statistical description of the output characteristics of distributed wind, solar and other uncertain resources within a future preset time domain (such as 15 minutes) based on weather forecasts and historical output models.

[0024] The aforementioned pre-defined mapping function can be a mathematical model that establishes the correspondence between external disturbance risks and internal safety thresholds. It is typically expressed as a monotonically non-decreasing function with volatility standard deviation as the independent variable and target virtual inertia as the dependent variable. The function's purpose is to dynamically define an adaptive safety defense red line for the system based on environmental predictions.

[0025] The aforementioned target virtual inertia can be the minimum virtual inertia level that must be maintained to cope with anticipated resource fluctuations. As a reference benchmark for the safe operation of the system, it is the instruction input in the closed-loop verification process and directly determines whether the system needs to switch from "normal economic mode" to "stability reshaping mode".

[0026] The aforementioned droop control coefficient can be defined as the proportional gain that defines the intensity of the energy deviation in the response system of each unit, reflecting the output weight of each unit in the coordinated regulation task. This coefficient is obtained through multi-objective optimization, taking into account the operating cost of each unit, the equipment degradation rate, and its potential contribution to the overall stability of the system.

[0027] The aforementioned preset system energy state reference benchmark can be the center point of the energy distribution that is expected to be maintained under ideal operating conditions (for example, setting SES to 0.5).

[0028] The aforementioned system energy state deviation can be the algebraic difference between the real-time SES and the reference baseline. It characterizes the degree of deviation of the system's current energy reserve from the ideal state and is the only global feedback signal in the distributed control logic, used to trigger the coordinated adjustment behavior of each heterogeneous unit.

[0029] The aforementioned power adjustment amount can be the instantaneous output adjustment command obtained by each energy storage unit based on the product of the local droop coefficient and the global energy deviation.

[0030] The aforementioned preset optimization model can be a mathematical framework for reconfiguring system control parameters in emergency situations, typically employing constrained quadratic programming or multi-objective programming algorithms. Its core task is to redistribute the regulation responsibilities of each unit with minimal parameter disturbance cost while ensuring system inertia recovery.

[0031] The aforementioned inertia recovery constraint can be a hard boundary condition in the optimization model, and its mathematical expression is that the sum of the inertia contributions of each unit after updating the coefficients is not less than the target virtual inertia.

[0032] The aforementioned operating benchmark can be the intercept term of the droop control curve of each energy storage unit on the vertical axis, representing the physical output baseline of the equipment under a deviation-free state. By adjusting the benchmark, the physical operating range of the equipment is essentially changed (such as pulling the energy storage from the edge of full charge back to the center area), thereby fundamentally reshaping the dynamic support capability of the system.

[0033] In some embodiments, determining the droop control coefficient of each energy storage unit based on a preset system operating cost target may further include: Obtain the unit regulation cost parameters for each of the energy storage units, wherein the unit regulation cost parameters include unit power degradation cost, unit energy conversion loss cost, and operation and maintenance conversion cost; Calculate the comprehensive cost weighting factor for each energy storage unit based on the unit adjustment cost parameter of each energy storage unit. Obtain the preset total droop adjustment gain of the energy system, and determine the allocation ratio of each energy storage unit to the total droop adjustment gain according to the comprehensive cost weighting factor of each energy storage unit, wherein the allocation ratio is negatively correlated with the comprehensive cost weighting factor; Multiply the allocation ratio by the total droop adjustment gain to obtain the droop control coefficient corresponding to each energy storage unit.

[0034] Based on the above embodiments, the online computing pressure on the distributed controller is significantly reduced, and the latency risk of complex optimization algorithms running on low-performance embedded chips is avoided. By clearly defining the negative correlation between cost weight and droop coefficient, it is ensured that the system's adjustment actions spontaneously tend to protect high-value, low-efficiency loss assets during long-term operation. While ensuring real-time adjustment performance, the system's economic efficiency is maximized, and the logic is simple and robust.

[0035] In some embodiments, the step of multiplying the system energy state deviation with the droop control coefficient of each energy storage unit to determine the power regulation of each energy storage unit may specifically include: The initial adjustment command value of each energy storage unit is obtained by multiplying the system energy state deviation with the droop control coefficient corresponding to each energy storage unit. Determine whether the absolute value of the system energy state deviation is greater than a preset adjustment dead zone threshold; If the absolute value of the system energy state deviation is greater than the adjustment dead zone threshold, then the real-time adjustability margin of each energy storage unit is determined based on the current operating power of each energy storage unit and the preset output power extreme value. The initial adjustment command value is limited using the real-time adjustable margin to determine the power adjustment amount of each energy storage unit.

[0036] In some embodiments, during the autonomous coordinated adjustment process, the target virtual inertia is verified using the system virtual inertia to obtain a verification result. Specifically, this may include: Obtain the system virtual inertia determined at the current moment, and retrieve the target virtual inertia determined by the mapping function at the corresponding moment; Calculate the inertia deviation between the system virtual inertia and the target virtual inertia; Determine whether the inertia deviation value is less than a preset safety threshold. If the inertia deviation value is less than the safety threshold, then determine that the system has entered an inertia-deficient state. The cumulative duration or trigger frequency of the inertia deficiency state within a preset observation window is statistically analyzed, and the verification result is generated based on the cumulative duration or trigger frequency; wherein, the verification result is used to indicate whether to activate the preset optimization model to update the droop control coefficient.

[0037] Based on the above embodiments, the robustness of the energy system control strategy is significantly improved, effectively avoiding false triggers caused by instantaneous sensor errors or unsteady random disturbances, thereby reducing unnecessary control parameter adjustments. By accurately locking in the verification results, it is ensured that the system only calls the optimization model for reconstruction when truly facing a stability crisis. This not only reduces the ineffective consumption of computing resources but also maintains the stability of the system during adjustment, guaranteeing the orderly switching of energy storage units between normal economic operating conditions and extreme defense operating conditions.

[0038] In some embodiments, the operating benchmark serves as the target output base point for each energy storage unit in the energy equilibrium state, acting as a physical state "anchor." By real-time correction of this benchmark value, the power command bias of the energy storage unit in steady state is changed, thereby driving the state of charge of the electrochemical energy storage unit to shift to a range with greater adjustment flexibility, or guiding the hydrogen energy storage unit into a pressure range with greater response potential. This feedback adjustment mechanism based on benchmark displacement not only realizes the adaptive reconstruction of the state of charge, but also ensures from the underlying physical support level that the total virtual inertia of the system can be quickly and stably restored to a safe threshold level not lower than the preset target virtual inertia.

[0039] In some embodiments, the method for determining the corresponding system energy state and system virtual inertia based on the operating data may further include the following: S1: Normalize the operating data to obtain multiple normalized energy storage state variables; wherein, the operating data includes the state of charge data of the electrochemical energy storage unit, the hydrogen storage pressure data of the hydrogen energy storage unit, and the demand response potential data of the adjustable load energy storage unit. S2: Use preset weighting coefficients to perform a weighted summation of the multiple normalized energy storage state variables to determine the corresponding system energy state; S3: Based on the operating data of each energy storage unit and the preset inertia constant, calculate the state-dependent inertia contribution factor that each energy storage unit can provide in the current state. S4: Linearly sum and aggregate the inertia contribution factors of each energy storage unit to determine the corresponding system virtual inertia.

[0040] The aforementioned preset inertia constant can be a static physical parameter that is predetermined during the design or configuration phase of each energy storage unit and is used to characterize its inherent virtual inertia support capability.

[0041] From a physical perspective, this constant represents the fundamental support capability of an energy storage device under rated operating conditions, determined by its topology, the physical characteristics of the energy storage medium, and the converter control bandwidth. For non-rotating resources such as electrochemical or hydrogen energy storage, the preset inertia constant is essentially a nominal value simulating the rotating inertia of a traditional synchronous generator, reflecting the nominal virtual inertia intensity that the unit can provide to the system under ideal conditions. This constant serves as the baseline for calculations in the system and does not change with real-time environmental fluctuations. Its role is to set a standardized starting point for support evaluation for each adjustable unit. By coupling with subsequently dynamically determined state correction coefficients, the static physical properties of the equipment are transformed into dynamic support capabilities deeply bound to the current operating state.

[0042] Specifically, the system first collects raw operational data from various heterogeneous energy storage units in real time via an underlying communication gateway. For electrochemical energy storage units, their State of Charge (SOC) data is extracted; for hydrogen energy storage units, their hydrogen storage pressure data is extracted; and for adjustable load units, their current demand response potential data is extracted. Subsequently, a pre-defined normalization function is used to standardize the data, which have inconsistent physical dimensions. Through a mapping relationship, the SOC, pressure, and potential values ​​are converted into dimensionless energy storage state variables within the range [0, 1]. This process eliminates the property differences between different energy media, laying the foundation for subsequent unified quantitative evaluation.

[0043] Next, a pre-defined weighting coefficient table is invoked, and weighting coefficients are assigned to multiple normalized energy storage state variables based on the energy capacity, response speed, and importance of each energy storage unit. Through weighted summation, the state information of each unit is aggregated into a scalar indicator reflecting the overall energy reserve adequacy, namely the State of Energy (SES). This SES can intuitively reflect the degree of energy deviation from the ideal operating point at the current moment, providing a unified feedback signal for the upper-level control algorithm.

[0044] Meanwhile, the preset inertia constant for each energy storage unit is obtained through table lookup or real-time calculation. This constant characterizes the inherent physical support capability of the equipment under rated conditions. To reflect the dynamic changes in equipment capability during actual operation, a state-dependent evaluation logic is introduced: combining real-time operating data of each unit (such as the proximity of extreme values ​​of SOC), a nonlinear mapping function is used to correct the inertia constant, thereby calculating the state-dependent inertia contribution factor that each energy storage unit can actually provide under current physical constraints. This factor reflects the instantaneous strength of the equipment in resisting power disturbances under current energy margins.

[0045] Finally, the inertia contribution factors generated by each energy storage unit are linearly summed and aggregated to output the current virtual inertia. This aggregated index is no longer a static parameter, but a dynamic variable that fluctuates in real time with the physical state of each unit. By transforming the support capacity of dispersed heterogeneous resources into a globally unified virtual inertia, real-time perception of dynamic stability margin is achieved, thus providing a precise trigger criterion for subsequent adjustment of the droop control coefficient based on stability gaps.

[0046] By employing normalization and weighted aggregation methods, this approach effectively addresses the challenge of uniformly measuring heterogeneous energy resources across the energy dimension, achieving deep coupling and logical alignment of multiple resources such as electricity, hydrogen, and load at the control level. By introducing a "state-dependent" inertia calculation mechanism, this method can objectively reflect the true disturbance rejection capability under different energy reserve levels in real time, avoiding the evaluation biases produced by traditional fixed inertia models when equipment is at energy extremes (such as battery depletion or insufficient hydrogen storage pressure). This significantly improves the predictability and accuracy of dynamic stability margin perception, providing a solid physical reference benchmark for subsequent proactive state reshaping and inertia recovery actions, and ensuring the operational safety of integrated energy systems under complex conditions.

[0047] In some embodiments, the step of calculating the inertia contribution factor of each energy storage unit in its current state based on the operating data of each energy storage unit and a preset inertia constant may specifically include: Based on the operating data of each energy storage unit, a state correction coefficient corresponding to each energy storage unit is determined; wherein, the state correction coefficient is used to characterize the gain or attenuation characteristics of the virtual inertia support capability of each energy storage unit under real-time physical conditions. Each state correction coefficient is multiplied by its corresponding preset inertia constant to obtain the inertia contribution factor of each energy storage unit in the current state.

[0048] Specifically, the system first uses real-time operating data of each energy storage unit collected by the global state information acquisition module. Through built-in mapping logic, a corresponding state correction coefficient is determined for each unit. This coefficient is designed as a "real-time weight" to characterize the true strength of the virtual inertia support capability provided by the device under different physical states. Specifically, when the energy storage unit's physical state is in the efficient adjustment range, this coefficient exhibits gain characteristics, strengthening its support contribution; while when the unit approaches its physical operating limit or its adjustment capability tends to saturate, the coefficient exhibits decay characteristics, thus accurately reflecting the decline in its support capability.

[0049] Furthermore, for electrochemical energy storage units, their state correction coefficient is deeply coupled with the current state of charge (SOC). A safe redundancy range for SOC is preset. When the battery's SOC is in the middle of this range, a higher SOC is assigned to reflect its sufficient power throughput potential. Once the SOC shifts to extremely high or low levels and approaches the battery's physical cutoff boundary, the SOC of the unit is automatically lowered according to the degree of shift, using a preset exponential or piecewise function. This approach effectively avoids the logical bias of mistakenly identifying battery packs on the adjustment edge as having full support capability during inertia verification.

[0050] For hydrogen energy storage units, the state correction factor is mainly determined by the hydrogen storage pressure and the corresponding response characteristics. In this embodiment, the pressure level inside the hydrogen storage tank is monitored in real time and compared with the rated operating pressure. Since the output response speed of hydrogen energy electrolyzers or fuel cells varies physically under different pressure environments, a preset pressure mapping curve is used to convert the current pressure data into the corresponding state correction factor. For example, when the pressure is in the rated response sensitivity range, an enhanced correction factor is assigned to ensure that the calculated inertia contribution can truly reflect the upper limit of the physical support of hydrogen energy under the current operating conditions.

[0051] Finally, the dynamic state correction coefficients determined above are multiplied by the pre-set inertia constants of each energy storage unit. The inertia constants are inherent physical parameters determined during the design phase of the equipment based on its rated capacity and simulated rotational inertia, representing its support baseline under ideal conditions. By multiplying this "static baseline" by the aforementioned "dynamic correction coefficients," the inertia contribution factor of each energy storage unit in its current state is obtained. This calculation method achieves a scientific transformation from static physical attributes to real-time support capabilities, enabling the subsequently generated total virtual inertia to accurately capture the stability margin evolution of distributed resources due to state fluctuations, providing a physically deterministic data foundation for energy security defense.

[0052] In some embodiments, the step of constructing the system energy state of overall system energy adequacy includes: normalizing the state of charge data of the electrochemical energy storage unit, the hydrogen storage pressure data of the hydrogen energy storage unit, and the demand response potential data of the adjustable load unit to obtain multiple normalized energy storage state variables. Then, by weighted summing the multiple normalized energy storage state variables, the system energy state is calculated, which can be expressed as:

[0053] in, Represents time; The system energy state at time t; Normalized state of charge data for electrochemical energy storage units; Normalized hydrogen storage pressure data for hydrogen energy storage units; Normalized adjustable potential data for demand response load; These are the weighting coefficients for the three energy storage forms, and their sum is 1.

[0054] In some embodiments, the step of dynamically reconstructing the system virtual inertia of the system dynamic stability margin based on the operating state data specifically includes: firstly, calculating the state-dependent inertia contribution factor for each unit in the integrated energy system, wherein the calculation is based on the operating state data of each unit and a preset maximum inertia constant for each unit, the maximum inertia constant being predetermined based on the physical parameters and maximum power regulation capability of the unit; subsequently, aggregating the state-dependent inertia contribution factors of each unit to obtain the system virtual inertia. The aggregation method is linear summation:

[0055] in: It is the set of all adjustable units in the system that can contribute inertia; The inertia contribution factor of element i at time t; The total virtual inertia of the system at time t; in a more specific embodiment, the contribution factor of the inertia of the electrochemical energy storage unit. The calculation method is as follows:

[0056] in, The inertia contribution factor of the electrochemical energy storage unit at time t; The maximum inertia constant that the electrochemical energy storage unit can contribute is determined in advance based on the rated capacity, maximum charge / discharge rate and other physical parameters of the energy storage unit, as well as the maximum power regulation capability determined therefrom. The real-time state of charge at time t; As a reference state of charge, it is usually set to 0.5, representing an intermediate charge level; This is the highest state of charge allowed by the regulations; To calculate the current state of charge Relative to the ideal reference point Normalization bias; The whole constitutes a state evaluation function that takes values ​​within an interval. Its value is the maximum value of 1 when the charged state is the reference point, and gradually decreases to 0 when the charged state deviates from the reference point.

[0057] In some embodiments, the method for determining the target virtual inertia based on the resource volatility prediction data within the preset time domain using a preset mapping function may further include the following: S1: Obtain meteorological forecast data within the prediction time domain, and determine the total power output prediction curve within the prediction time domain based on the meteorological forecast data; S2: Calculate the standard deviation of the total output prediction curve in the prediction time domain, and determine the standard deviation as the resource volatility prediction data; S3: Substitute the standard deviation as the independent variable into the preset mapping function, and use the preset mapping function to calculate and determine the target virtual inertia; The preset mapping function is a monotonically non-decreasing function, so that the target virtual inertia increases as the resource volatility prediction data increases.

[0058] Specifically, high-precision weather forecast data for the energy system's location is first acquired in real time via a communication link. The time resolution is typically set to the minute level, covering key parameters such as average wind speed, instantaneous wind speed, solar radiation intensity, and ambient temperature for the next 15 minutes to 1 hour. The control unit then calls upon pre-stored energy conversion models (such as the Pv characteristic curve of wind turbines and the IV mathematical model of photovoltaic arrays) to map the weather forecast parameters to corresponding power prediction points, thereby fitting and generating a total output prediction curve within the prediction time domain. This curve visually describes the evolution trend and potential abrupt changes in the system's total power supply over a future period.

[0059] Subsequently, the system utilizes statistical analysis methods to perform in-depth analysis of the total power output prediction curve. Specifically, the system extracts power sampling points within the prediction time domain using a sliding window algorithm and calculates the standard deviation of these sampling points relative to the mean. In this embodiment, the standard deviation is... Defined as resource volatility forecast data, it is used to quantify the uncertainty of future meteorological resources. If the standard deviation is large, it indicates that the power fluctuations are frequent and severe within the forecast time domain, and the system may face a serious risk of instantaneous power imbalance; conversely, if the standard deviation is small, it represents a stable meteorological environment and a lower risk of system operation.

[0060] Finally, the system will calculate the standard deviation. As the independent variable, it is substituted into the preset mapping function J. target = f( The target virtual inertia is determined through analytical calculations. In this embodiment, the preset mapping function is designed as a monotonically non-decreasing function, such as a piecewise linear function, an exponential function, or a sigmoid function with saturation characteristics. This mapping mechanism ensures that the target virtual inertia J is determined. target It can dynamically improve as the intensity of resource fluctuation prediction increases. When weather forecasts indicate that extreme fluctuations are about to occur, the system automatically increases the defense level (i.e., increases the target virtual inertia) through this mapping logic, providing an accurate reference target for subsequent inertia recovery and benchmark adjustment.

[0061] In some embodiments, the method of substituting the standard deviation as an independent variable into the preset mapping function and using the preset mapping function to calculate and determine the target virtual inertia may further include the following: The target virtual inertia is determined according to the following formula:

[0062] in, The target virtual inertia, To predict the length of the time domain, The length of the prediction time domain starting from time t The standard deviation of the total output forecast within, It is a monotonically non-decreasing function.

[0063] In some embodiments, the method for determining the droop control coefficient of each energy storage unit based on a preset system operating cost target may further include the following: S1: Construct a multi-objective optimization function with the total cost of the energy system in the prediction time domain as the optimization objective; wherein, the total cost includes the operating cost of each energy storage unit, the degradation cost of the equipment, and the compensation cost of demand response load; S2: Establish the constraints of the multi-objective optimization function; wherein the constraints include at least the power adjustment range limit term, the power ramp rate limit term, and the operating safety range limit term for each energy storage unit; S3: Use a preset optimization algorithm to find the optimal solution for the multi-objective optimization function that satisfies the constraints, and determine the droop control coefficient of each energy storage unit. S4: Based on the constraints, the multi-objective optimization function is optimized to determine the droop control coefficient of each energy storage unit.

[0064] Specifically, a multi-objective optimization function is first constructed using a pre-defined economic evaluation index system, with the goal of minimizing the total operating cost of energy within the future forecast time domain. This total cost function is a comprehensive index: operating cost is calculated from the energy conversion losses of each energy storage unit and real-time electricity / gas prices; degradation cost is converted into economic expenditure by establishing a cycle life decay model for energy storage equipment, converting physical losses such as charge / discharge depth and cycle count into economic expenditures; and compensation cost for demand response load is calculated based on a pre-defined incentive agreement, determining the economic compensation payable to users for shifting or reducing load. In this way, multi-dimensional operating costs are quantified into a unified economic benchmark.

[0065] Subsequently, a rigorous set of physical and safety constraints was established for the multi-objective optimization function to ensure that the optimization process does not deviate from the limits of hardware operation. Specifically: the power regulation range constraint specifies the maximum / minimum charge / discharge power of each energy storage unit to prevent equipment damage due to overload; the power ramp-up rate constraint limits the span of power change per unit time to ensure that the regulation process conforms to the dynamic response characteristics of various energy storage media; and the safe operating range constraint focuses on monitoring the state of charge (SOC) of electrochemical energy storage and the tank pressure of hydrogen energy storage, forcing them to operate within the preset safety margin range. These constraints constitute the search boundary for the optimization solution, ensuring the physical feasibility of the control strategy.

[0066] Finally, a pre-defined optimization algorithm (such as particle swarm optimization, genetic algorithm, or interior point method) is invoked to efficiently iteratively solve the optimization model that satisfies the above constraints. During the optimization process, the algorithm uses the droop control coefficient k of each energy storage unit... i As a decision variable, by continuously adjusting the coefficient distribution, the search aims to minimize the total cost F. total The optimal combination of control gains is minimized. After solving, the calculated optimal droop control coefficients are distributed to the controllers of each unit as the gain benchmark for subsequent autonomous coordinated adjustment. This process achieves a precise mapping from macroeconomic benefit objectives to underlying control parameters.

[0067] Based on the above embodiments, a deep integration of energy dispatching objectives and real-time control characteristics is achieved. Through refined modeling across multiple cost dimensions, the overall lifecycle operational economy is significantly improved. By explicitly introducing equipment degradation costs into the optimization function, this method effectively avoids over-utilization of specific energy storage units, significantly extending the lifespan of core hardware while meeting regulation requirements. Furthermore, through hard constraints on power range and ramp rate, the generated droop control coefficients are ensured to possess high robustness and safety in real-world physical environments. This effectively addresses the risks of regulation over-limits or equipment damage that often occur with traditional empirical matching methods under complex operating conditions, achieving a dual optimization of economical operation and stable operation.

[0068] In some embodiments, the step of determining the optimal droop control coefficient for each energy storage unit is performed by the central controller, which constructs a total cost estimate for the integrated energy system over a future time domain. Multi-objective optimization function with minimization as the objective:

[0069] in: To predict discrete time points within a time domain of length T; The operating cost of the system may include the cost of purchasing electricity from an external power grid or the revenue from selling electricity to it, as well as the fuel cost of hydrogen consumed by the fuel cell. The degradation cost of the equipment may specifically include: the life loss cost of the electrochemical energy storage unit due to charge-discharge cycles, which can be quantified as a function related to its throughput; The compensation costs required to invoke demand response load; These are the weighting factors for various costs, used to adjust the priority among different optimization objectives.

[0070] In some embodiments, the method of updating the droop control coefficient of each energy storage unit using a preset optimization model may further include the following: S1: Determine the stability adjustment priority of each energy storage unit based on the system energy state corresponding to each energy storage unit and the inertia contribution factor; S2: Determine the weight coefficients in the objective function of the preset optimization model based on the stability adjustment priority; S3: Substitute the target virtual inertia as a parameter into the inertia recovery constraint in the preset optimization model, so that the sum of the inertia provided by each energy storage unit is greater than or equal to the target virtual inertia; S4: Using an optimization algorithm, under the premise of satisfying the inertia recovery constraint, perform optimization calculation on the objective function containing the weight coefficients to determine the correction amount of each of the droop control coefficients; S5: Determine the updated droop control coefficients based on the correction amount of each droop control coefficient.

[0071] Specifically, when it is determined that the virtual inertia is insufficient to cope with current resource fluctuations, the system autonomously invokes a pre-set optimization model in the background to reconstruct the control law. First, it retrieves the real-time energy state and inertia contribution factor of each energy storage unit. Specifically, it couples the energy reserve adequacy with the instantaneous support capacity to determine the stability adjustment priority of each unit. This priority reflects the equipment's "willingness" and "ability" to participate in adjustment under the current physical state: that is, equipment with higher energy levels and greater inertia increments due to unit coefficient changes is assigned a higher priority level.

[0072] Subsequently, based on the determined stability adjustment priority, differentiated weight coefficients are configured for the objective function of the preset optimization model. In this embodiment, priority and weight coefficients are negatively correlated: for units with high priority, by reducing their weight ratio in the objective function, these units, guided by the mathematical principle of "minimizing adjustment costs," can undertake larger parameter corrections with less computational penalty. This mechanism realizes a "the capable do more" resource scheduling logic, ensuring that the adjustment potential of advantageous resources is maximized while ensuring safety.

[0073] In the constraint processing stage, the target virtual inertia determined in the previous steps is used as a numerical parameter and input into the inertia recovery constraint term of the preset optimization model in real time. This action mathematically sets a hard safety "red line" for the optimization calculation, requiring that the sum of the virtual inertia provided by all energy storage units participating in the adjustment must be greater than or equal to the target threshold after updating the coefficients. At the same time, a power adjustment range constraint term is constructed by combining the rated power, current limit and other physical boundaries of each unit, thereby locking the safety requirements and hardware physical limits together within the feasible region of optimization.

[0074] Finally, a pre-defined optimization algorithm (such as the interior-point method or the effective set method) is used to solve the configured objective function. Under the dual constraints of inertia recovery and physical power, the algorithm calculates the correction amount for the droop control coefficient corresponding to each energy storage unit. This correction amount is algebraically superimposed with the initially determined droop control coefficient, ultimately outputting the updated droop control coefficient. This deviation-based update method achieves a smooth transition from the economic operating benchmark to the safety support benchmark, avoiding the secondary oscillation impact caused by large jumps in control parameters.

[0075] In some embodiments, the preset optimization model can be determined according to the following formula:

[0076] Where F is the objective function, These are the weighting coefficients. For decision variables, it refers to the correction amount of the droop control coefficient of the i-th energy storage unit.

[0077] The inertia recovery constraint can be determined according to the following formula:

[0078] in, Let be the droop control coefficient determined for the i-th energy storage unit in the initial state (economic mode). Let be the inertia contribution factor of the i-th energy storage unit. The target virtual inertia.

[0079] In some embodiments, the virtual inertia of the system is checked in a closed loop. When the virtual inertia of the system is detected to be lower than the target virtual inertia, a state reshaping is triggered. The state reshaping restores the virtual inertia of the system by re-executing the step of determining the optimal droop control coefficients of each unit.

[0080] The state reshaping is performed under a high-priority constraint: ensuring that the droop control coefficients redefined through state reshaping enable the system's expected total virtual inertia. Restored to a value no less than the target virtual inertia The level that satisfies .

[0081] In some embodiments, the method of adjusting the operating baseline of each energy storage unit according to the updated droop control coefficient may further include the following: S1: Determine the baseline adjustment amount corresponding to each energy storage unit based on the updated droop control coefficient; wherein, the energy storage unit includes an electrochemical energy storage unit and a hydrogen energy storage unit. S2: Based on the reference adjustment amount corresponding to the electrochemical energy storage unit, shift the power operation reference of the electrochemical energy storage unit from the output point corresponding to the current state of charge to the preset energy state equilibrium point, so as to correct the power adjustment space of the electrochemical energy storage unit. S3: Based on the reference adjustment amount corresponding to the hydrogen energy storage unit, switch the operating mode reference of the hydrogen energy storage unit from the current shutdown reference or restricted operating reference to the corresponding standby operating reference or fully adjustable operating reference, so as to release the inertia support potential of the hydrogen energy storage unit.

[0082] Specifically, after obtaining the updated droop control coefficients, the gain adjustment is not only performed at the control algorithm level, but also simultaneously triggers a reshaping of the underlying physical resource operating baseline. First, based on the deviation between the updated droop control coefficients and the initial coefficients, and combined with the power sensitivity index of each energy storage unit, the baseline adjustment amount for each unit is calculated. This adjustment amount serves as a physical execution command to correct the output starting point of each unit in the energy equilibrium state, ensuring that the change in control gain can match the margin of physical output to jointly complete the inertia support task.

[0083] For electrochemical energy storage units, a "power reference shift" operation is performed using a defined reference adjustment amount. Specifically, if the current state of charge (SOC) of the battery cell is at a high or low adjustment edge (e.g., close to 90% or 10%), its original output point may limit the adjustment space. In this case, the power operating reference is shifted from the current output point to a preset energy state equilibrium point (e.g., the rated equilibrium point corresponding to 50% SOC) according to the reference adjustment amount. This displacement guides the SOC back to a safe range with bidirectional adjustment flexibility by forcibly changing the battery's steady-state output reference, thereby physically correcting and widening the power adjustment space of electrochemical energy storage and avoiding adjustment saturation during inertia recovery.

[0084] For hydrogen energy storage units, a "mode transition" operation is performed using a reference adjustment. Considering the differences in the response characteristics of hydrogen energy (such as electrolyzers or fuel cells) under different modes, the operating mode reference of the hydrogen energy storage unit is forcibly switched from the current shutdown reference or restricted operating reference to the corresponding standby operating reference or fully adjustable operating reference, based on the instructions generated by the reference adjustment. This switching action effectively removes the physical constraints on the hydrogen energy equipment under normal economic operating conditions, allowing it to transition from a "low-activity" or "off" state to a "high-sensitivity" support state. Through this change in the mode reference, the previously locked energy throughput capacity of the hydrogen energy unit is instantly released, thereby providing sufficient inertia support potential at the physical level and ensuring that the full virtual inertia can quickly recover to above the safe target value.

[0085] Based on the above embodiments, by establishing a control architecture that combines periodic optimization and autonomous collaborative control, the system's robustness to communication delays and its rapid response to power disturbances are improved. The high-frequency autonomous collaborative control layer can autonomously adjust power based on system energy state deviation and pre-optimized droop control coefficients, without waiting for real-time commands from the central controller. This reduces dependence on high-speed communication networks and ensures the system's instantaneous response performance during disturbances. By constructing a unified system energy state and virtual inertia, intelligent collaboration and resource complementarity among heterogeneous units such as wind, solar, hydrogen storage, and demand response are achieved. By quantifying the states of energy storage and load units with different physical characteristics into two unified indicators at the system level—energy adequacy and dynamic stability margin—the central controller can optimize operating rules at the overall system level. Through differentiated allocation of droop coefficients, it guides each unit to automatically and collaboratively assume power balance responsibilities according to its own state. By introducing the setting of target virtual inertia and the closed-loop verification and state reshaping mechanism of system virtual inertia, the forward-looking and proactive maintenance capability of system dynamic stability is enhanced. Through the prediction of the volatility of renewable energy, the dynamic stability margin that the system needs to maintain is set in advance. Through periodic verification and state reshaping when necessary, it is ensured that the system always has sufficient inertia reserve to cope with potential high-power disturbances, thereby transforming passive disturbance response into proactive margin management.

[0086] As can be seen from the above, the intelligent collaborative operation method for an energy system provided in this specification involves acquiring the operating data of each energy storage unit within the energy system, as well as resource volatility prediction data within a preset time domain; determining the corresponding system energy state and system virtual inertia based on the operating data; wherein the virtual inertia is used to characterize the system's dynamic stability margin; determining the target virtual inertia through a preset mapping function based on the resource volatility prediction data within the preset time domain; determining the droop control coefficient of each energy storage unit based on a preset system operating cost target; determining the system energy state deviation based on the system energy state and a preset system energy state reference benchmark; and multiplying the system energy state deviation by the droop control coefficient of each energy storage unit. The power regulation amount of each energy storage unit is determined to drive each energy storage unit to perform autonomous coordinated regulation. During the autonomous coordinated regulation process, the system virtual inertia is used to verify the target virtual inertia to obtain a verification result. According to the verification result, when the system virtual inertia is less than the target virtual inertia, the droop control coefficient of each energy storage unit is updated using a preset optimization model. The preset optimization model includes at least an inertia recovery constraint. The operating reference of each energy storage unit is adjusted according to the updated droop control coefficient. The operating reference is the target output base point of each energy storage unit in the energy balance state, and the operating reference is used to adjust the state of charge of the energy storage unit so that the system virtual inertia is restored to a level not lower than the target virtual inertia. In this way, by using the system energy state and droop control coefficient to determine the power adjustment amount of each unit, each unit is driven to perform autonomous and coordinated adjustment, which effectively improves the real-time performance and reliability of the system adjustment response. By using the closed-loop verification of the system virtual inertia against the target virtual inertia, and adjusting the operating benchmark of each unit according to the updated droop control coefficient, it is ensured that the adjusted system virtual inertia is not lower than the target virtual inertia, thus realizing the active recovery of the system's dynamic stability margin.

[0087] See Figure 2 As shown in the embodiments of this specification, a specific electronic device is also provided, wherein the electronic device includes a network communication port 201, a processor 202 and a memory 203, and the above structures are connected by internal cables so that the various structures can perform specific data interaction.

[0088] Specifically, the network communication port 201 can be used to acquire the operating data of each energy storage unit in the energy system, as well as the resource volatility prediction data within a preset time domain.

[0089] The processor 202 is specifically configured to determine the corresponding system energy state and system virtual inertia based on the operating data; wherein the virtual inertia is used to characterize the system's dynamic stability margin; determine the target virtual inertia through a preset mapping function based on the resource volatility prediction data within the preset time domain; determine the droop control coefficient of each energy storage unit based on the preset system operating cost target; determine the system energy state deviation based on the system energy state and the preset system energy state reference benchmark; and multiply the system energy state deviation by the droop control coefficient of each energy storage unit to determine the power adjustment amount of each energy storage unit, thereby driving each energy storage unit to perform autonomous operation. Coordinated adjustment; during the autonomous coordinated adjustment process, the system virtual inertia is used to verify the target virtual inertia to obtain a verification result; based on the verification result, when the system virtual inertia is less than the target virtual inertia, the droop control coefficient of each energy storage unit is updated using a preset optimization model; wherein, the preset optimization model includes at least an inertia recovery constraint; the operating reference of each energy storage unit is adjusted according to the updated droop control coefficient; wherein, the operating reference is the target output base point of each energy storage unit in the energy balance state, and the operating reference is used to adjust the state of charge of the energy storage unit so that the system virtual inertia is restored to a level not lower than the target virtual inertia.

[0090] The memory 203 can be used to store the corresponding instruction program.

[0091] Based on the above method, the relevant structural performance of electronic devices can be effectively utilized to improve the data processing speed of electronic devices and efficiently realize a method for intelligent collaborative operation of energy systems.

[0092] In this embodiment, the network communication port 201 can be a virtual port bound to different communication protocols, thereby enabling the sending or receiving of different data. For example, the network communication port can be a port responsible for web data communication, a port responsible for FTP data communication, or a port responsible for email data communication. Furthermore, the network communication port can also be a physical communication interface or communication chip. For example, it can be a wireless mobile network communication chip, such as GSM or CDMA; it can also be a Wi-Fi chip; or it can be a Bluetooth chip.

[0093] In this embodiment, the processor 202 can be implemented in any suitable manner. For example, the processor can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers, etc. This specification is not limiting.

[0094] In this embodiment, the memory 203 may include a hierarchy. In a digital system, anything that can store binary data can be a memory. In an integrated circuit, a circuit with storage function but no physical form is also called a memory, such as RAM, FIFO, etc. In a system, a storage device with a physical form is also called a memory, such as a memory stick, TF card, etc.

[0095] This specification also provides a computer-readable storage medium based on the above-described intelligent collaborative operation method for an energy system. This medium acquires operational data of each energy storage unit within the energy system, as well as resource volatility prediction data within a preset time domain. Based on the operational data, it determines the corresponding system energy state and system virtual inertia, whereby the virtual inertia characterizes the system's dynamic stability margin. Based on the resource volatility prediction data within the preset time domain, it determines the target virtual inertia through a preset mapping function. Based on a preset system operating cost target, it determines the droop control coefficient for each energy storage unit. Based on the system energy state and a preset system energy state reference benchmark, it determines the system energy state deviation. Finally, it compares the system energy state deviation with the droop control coefficient of each energy storage unit. The power adjustment amount of each energy storage unit is determined by multiplication to drive each energy storage unit to perform autonomous coordinated adjustment. During the autonomous coordinated adjustment process, the system virtual inertia is used to verify the target virtual inertia to obtain a verification result. According to the verification result, when the system virtual inertia is less than the target virtual inertia, the droop control coefficient of each energy storage unit is updated using a preset optimization model. The preset optimization model includes at least an inertia recovery constraint. The operating reference of each energy storage unit is adjusted according to the updated droop control coefficient. The operating reference is the target output base point of each energy storage unit in the energy balance state, and the operating reference is used to adjust the state of charge of the energy storage unit so that the system virtual inertia is restored to a level not lower than the target virtual inertia.

[0096] In this embodiment, the storage medium includes, but is not limited to, Random Access Memory (RAM), Read-Only Memory (ROM), Cache, Hard Disk Drive (HDD), or Memory Card. The memory can be used to store computer program instructions. The network communication unit can be an interface configured according to standards specified in the communication protocol for network connection communication.

[0097] In this embodiment, the specific functions and effects implemented by the program instructions stored in the computer-readable storage medium can be explained in comparison with other embodiments, and will not be repeated here.

[0098] See Figure 3 At the software level, embodiments of this specification also provide an intelligent collaborative operation device for an energy system, which may specifically include the following structural modules: The data acquisition module 301 is used to acquire the operating data of each energy storage unit in the energy system, as well as the resource volatility prediction data within a preset time domain; The first determining module 302 is used to determine the corresponding system energy state and system virtual inertia based on the operating data; wherein the virtual inertia is used to characterize the system dynamic stability margin. The second determining module 303 is used to determine the target virtual inertia based on the resource volatility prediction data in the preset time domain through a preset mapping function; and to determine the droop control coefficient of each energy storage unit based on the preset system operating cost target. The power regulation module 304 is used to determine the system energy state deviation based on the system energy state and the preset system energy state reference benchmark; and to multiply the system energy state deviation with the droop control coefficient of each energy storage unit to determine the power regulation amount of each energy storage unit, so as to drive each energy storage unit to perform autonomous and coordinated regulation. The result determination module 305 is used to verify the target virtual inertia using the system virtual inertia during the autonomous coordinated adjustment process, and obtain the verification result. The coefficient update module 306 is used to update the droop control coefficient of each energy storage unit using a preset optimization model based on the verification results when the virtual inertia of the system is less than the target virtual inertia; wherein the preset optimization model includes at least an inertia recovery constraint. The reference adjustment module 307 is used to adjust the operating reference of each energy storage unit according to the updated droop control coefficient; wherein, the operating reference is the target output base point of each energy storage unit in the energy balance state, and the operating reference is used to adjust the state of charge of the energy storage unit so that the virtual inertia of the system is restored to a level not lower than the target virtual inertia.

[0099] In some embodiments, the first determining module 302, in its specific implementation, normalizes the operating data to obtain multiple normalized energy storage state variables; wherein, the operating data includes the state of charge data of the electrochemical energy storage unit, the hydrogen storage pressure data of the hydrogen energy storage unit, and the demand response potential data of the adjustable load energy storage unit; the multiple normalized energy storage state variables are weighted and summed using preset weighting coefficients to determine the corresponding system energy state; based on the operating data of each energy storage unit and the preset inertia constant, the state-dependent inertia contribution factor that each energy storage unit can provide in the current state is calculated respectively; the inertia contribution factors of each energy storage unit are linearly summed and aggregated to determine the corresponding system virtual inertia.

[0100] In some embodiments, the second determining module 303 described above specifically implements the following steps: acquiring meteorological forecast data within the prediction time domain; determining the total power output prediction curve within the prediction time domain based on the meteorological forecast data; calculating the standard deviation of the total power output prediction curve within the prediction time domain; determining the standard deviation as the resource volatility prediction data; substituting the standard deviation as an independent variable into the preset mapping function; and calculating the target virtual inertia using the preset mapping function; wherein the preset mapping function is a monotonically non-decreasing function, so that the target virtual inertia increases with the increase of the resource volatility prediction data. The step of substituting the standard deviation as an independent variable into the preset mapping function and calculating the target virtual inertia using the preset mapping function includes: The target virtual inertia is determined according to the following formula:

[0101] in, The target virtual inertia, To predict the length of the time domain, The length of the prediction time domain starting from time t The standard deviation of the total output forecast within, It is a monotonically non-decreasing function.

[0102] In some embodiments, the second determining module 303, in its specific implementation, constructs a multi-objective optimization function with the total cost of the energy system in the prediction time domain as the optimization objective; wherein the total cost includes the operating cost of each energy storage unit, the degradation cost of the equipment, and the compensation cost of demand response load; establishes constraints on the multi-objective optimization function; wherein the constraints include at least a power adjustment range limit term, a power ramp rate limit term, and an operating safety range limit term for each energy storage unit; uses a preset optimization algorithm to optimize and solve the multi-objective optimization function that satisfies the constraints, and determines the droop control coefficient of each energy storage unit; and optimizes and solves the multi-objective optimization function based on the constraints to determine the droop control coefficient of each energy storage unit.

[0103] In some embodiments, the coefficient update module 306, in its specific implementation, determines the stability adjustment priority of each energy storage unit based on the system energy state corresponding to each energy storage unit and the inertia contribution factor; determines the weight coefficients in the objective function of the preset optimization model based on the stability adjustment priority; substitutes the target virtual inertia as a parameter into the inertia recovery constraint in the preset optimization model so that the sum of inertia provided by each energy storage unit is greater than or equal to the target virtual inertia; uses an optimization algorithm to perform optimization calculation on the objective function containing the weight coefficients, under the premise of satisfying the inertia recovery constraint, to determine the correction amount of each droop control coefficient; and determines the updated droop control coefficients based on the correction amount of each droop control coefficient.

[0104] In some embodiments, the reference adjustment module 307, in specific implementation, adjusts the power operation reference of the electrochemical energy storage unit from the output point corresponding to the current state of charge to a preset energy state equilibrium point according to the adjustment direction determined by the updated droop control coefficient; and switches the operation mode reference of the hydrogen energy storage unit from the current shutdown reference or restricted operation reference to the corresponding standby operation reference or fully adjustable operation reference according to the state adjustment command determined by the updated droop control coefficient.

[0105] It should be noted that the units, devices, or modules described in the above embodiments can be implemented by computer chips or physical entities, or by products with certain functions. For ease of description, the above devices are described by dividing them into various modules according to their functions. Of course, in implementing this specification, the functions of each module can be implemented in the same software and / or hardware, or modules that implement the same function can be implemented by a combination of sub-modules or sub-units, etc. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and there may be other division methods in actual implementation. For example, units or components can be combined or integrated into another system, or some features can be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection between the devices or units shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0106] As can be seen from the above, based on the intelligent collaborative operation device for an energy system provided in the embodiments of this specification, the device acquires the operating data of each energy storage unit in the energy system, as well as the resource volatility prediction data within a preset time domain; according to the operating data, it determines the corresponding system energy state and system virtual inertia respectively; wherein, the virtual inertia is used to characterize the dynamic stability margin of the system; according to the resource volatility prediction data within the preset time domain, it determines the target virtual inertia through a preset mapping function; according to the preset system operating cost target, it determines the droop control coefficient of each energy storage unit; according to the system energy state and the preset system energy state reference benchmark, it determines the system energy state deviation; and it multiplies the system energy state deviation with the droop control coefficient of each energy storage unit. The power regulation amount of each energy storage unit is determined to drive each energy storage unit to perform autonomous coordinated regulation. During the autonomous coordinated regulation process, the system virtual inertia is used to verify the target virtual inertia to obtain a verification result. According to the verification result, when the system virtual inertia is less than the target virtual inertia, the droop control coefficient of each energy storage unit is updated using a preset optimization model. The preset optimization model includes at least an inertia recovery constraint. The operating reference of each energy storage unit is adjusted according to the updated droop control coefficient. The operating reference is the target output base point of each energy storage unit in the energy balance state, and the operating reference is used to adjust the state of charge of the energy storage unit so that the system virtual inertia is restored to a level not lower than the target virtual inertia.

[0107] In a specific scenario example, the intelligent collaborative operation method and device for energy systems provided in this specification can be applied to solve the technical problems of insufficient real-time performance and reliability of the adjustment response during collaborative operation of energy systems, and the difficulty in restoring the system's dynamic stability margin due to the inability to actively adjust the operating benchmark based on the system's virtual inertia kernel results. In specific implementations, the following may also be included.

[0108] In some embodiments, see Figure 4 As shown, an intelligent collaborative energy operation system is provided, whose core architecture consists of interconnected modules: a global state information acquisition module 100, a core state variable construction module 200, an operation rule optimization module 300, an autonomous collaborative control module 400, and an inertia margin verification and reshaping module 500. The global state information acquisition module 100 is responsible for real-time sensing of the raw operation data of each heterogeneous unit within the integrated energy system and outputting processed operation state data to the core state variable construction module 200 through a built-in preprocessing algorithm. The core state variable construction module 200 reconstructs the features in two dimensions by normalizing and weighting the above data: on the one hand, it outputs the real-time energy state reflecting the adequacy of energy reserves to the autonomous collaborative control module 400; on the other hand, it outputs the real-time virtual inertia reflecting instantaneous support capability to the inertia margin verification and reshaping module 500.

[0109] Meanwhile, the operation rule optimization module 300, based on resource volatility prediction results, determines the target virtual inertia and the initial droop control coefficient through an internally preset mapping function and economic optimization model. The droop control coefficient is sent to the autonomous collaborative control module 400 to drive the integrated energy system to perform proportional adjustments based on energy deviation; the target virtual inertia is sent to the inertia margin verification and reshaping module 500 as a dynamic red line for safe operation. The inertia margin verification and reshaping module 500 performs a closed-loop verification between the received real-time virtual inertia and the target virtual inertia to generate a verification result indicating whether the stability margin meets the standard.

[0110] When the inertia margin verification and reshaping module 500 determines that the current inertia level is insufficient, it sends a trigger state reshaping signal to the operation rule optimization module 300. Upon receiving this signal, the operation rule optimization module 300 immediately activates a preset optimization model, dynamically updates the droop control coefficients of each energy storage unit, and simultaneously adjusts the operating baseline of each unit. The updated control parameters are then sent back to the autonomous collaborative control module 400, thereby reshaping the operating state of the integrated energy system at the physical level. This closed-loop optimization mechanism based on inertia feedback ensures that the autonomous collaborative control module 400 can not only make economic adjustments based on energy deviations, but also, in the event of a stability crisis, achieve forced recovery of the inertia level to the target value through the coordinated efforts of modules 500 and 300.

[0111] In some embodiments, see Figure 5 As shown, the specific internal logic of the core state variable construction module 200 is disclosed. The core state variable construction module 200 receives the processed running state data and inputs it in parallel into two feature reconstruction branches to achieve a quantitative description of the state in the energy dimension and the stability dimension, respectively.

[0112] In the energy state construction branch, this module first extracts energy storage-related variables from the operational status data and performs normalization on them. The normalization process aims to eliminate the differences in physical dimensions between different energy storage media (such as electricity, hydrogen, and heat), transforming them into dimensionless variables with a unified scale. Subsequently, a preset weight matrix is ​​invoked, and the normalized variables are weighted and summed according to the energy capacity and adjustment priority of each unit. The real-time energy state output by this process can centrally reflect the overall energy surplus or deficit, providing a global feedback reference for subsequent autonomous and coordinated regulation.

[0113] Meanwhile, in the virtual inertia construction branch, this module focuses on evaluating the physical capability to withstand transient disturbances. For each connected adjustable unit, based on its real-time output state and physical constraints, the state-dependent inertia contribution factor that each unit can provide at the current moment is calculated. These contribution factors accurately reflect the evolution of the distributed resource's support capability due to changes in its own energy state (such as power level or pressure magnitude). Subsequently, by aggregating and summing the contribution factors of all adjustable units, a real-time virtual inertia reflecting the overall dynamic stability margin is finally generated. This indicator directly connects to the subsequent inertia verification process, ensuring real-time perception of stability risks by the control system.

[0114] In some embodiments, see Figure 6 As shown, the specific internal structure and logical flow of the operation rule optimization module 300 are disclosed. This module, through two parallel computing branches, realizes in-depth pre-setting of control rules from two dimensions: security defense requirements and economic operation indicators.

[0115] In the target virtual inertia determination branch, module 300 first retrieves the volatility prediction data of renewable energy output, which quantitatively describes the potential disturbance intensity caused by resource fluctuations within a preset time domain. Subsequently, it uses a preset mapping function to convert this volatility index into the corresponding target virtual inertia. This logic ensures that the safety red line can be dynamically adjusted according to predicted changes in the external environment, providing an adaptive reference benchmark for subsequent inertia verification.

[0116] In the droop control coefficient optimization branch, module 300 simultaneously acquires status and cost data. This branch first constructs a multi-objective optimization function based on the real-time operating cost of each energy storage unit, equipment degradation cost, and load-side compensation cost. Then, under the premise of strictly satisfying physical constraints such as the power adjustment range, power ramp-up rate, and safe operating interval of each unit, a preset optimization algorithm is invoked to find the optimal solution. The calculation results ultimately output the optimal droop control coefficient set, which defines the output weight of each heterogeneous resource in the conventional autonomous collaborative adjustment process, achieving the optimal distribution of operating costs.

[0117] In some embodiments, see Figure 7 As shown, the specific internal architecture and closed-loop execution logic of the autonomous collaborative control module 400 are disclosed. The autonomous collaborative control module 400, as the underlying execution core, receives real-time status and control parameters from the upper-level modules to achieve precise allocation of power output from each heterogeneous energy storage unit.

[0118] The module first retrieves the current real-time energy state and compares it with a preset energy state reference benchmark to calculate the energy state deviation, reflecting the overall energy gap. This deviation signal is then sent to the local controller of each energy storage unit. Combined with the droop control coefficient set issued by the operation rule optimization module 300, each unit autonomously calculates its power adjustment. This distributed computing model enables each unit to spontaneously determine its share of regulation based on its own adjustment gain, and ultimately generate and issue power adjustment commands to drive the underlying hardware to perform physical actions.

[0119] Meanwhile, the autonomous collaborative control module 400 also constructs a complete physical feedback loop. It aggregates the power regulation calculated by each unit and combines it with real-time uncontrollable source load output and the current real-time virtual inertia, all acting on the energy state reference. This process simulates the combined impact of distributed resource regulation, uncontrollable random disturbances, and inertia support on the overall energy level evolution, feeding it back into the real-time energy state calculation at the next moment. Through this dynamic cyclical feedback, module 400 ensures that the regulation action is not only a response based on instantaneous deviations, but also exists within a closed-loop autonomous regulation system that considers physical inertia and disturbance cancellation, maintaining the dynamic balance of the energy level.

[0120] In some embodiments, see Figure 8 As shown, the internal logic judgment and execution flow of the inertia margin verification and reshaping module 500 are disclosed. As the ultimate defense for safe operation, this module realizes a logical closed loop from state perception to proactive reshaping through a periodic verification mechanism.

[0121] The inertia margin verification and reshaping module 500 receives the real-time virtual inertia output by the core state variable construction module and retrieves the target virtual inertia at the current moment. The pre-set periodic verification logic within the module continuously compares these two key indicators. When the verification determines that the real-time virtual inertia is not lower than the target virtual inertia, the current operation is maintained; however, once it determines that the real-time virtual inertia is less than the target virtual inertia, i.e., it is in the "no" judgment branch, it will immediately execute the "yes" branch logic and formally issue a signal to trigger state reshaping.

[0122] After triggering state reshaping, module 500 forcibly invokes the operation rule optimization module 300, incorporating high-priority constraints, namely the aforementioned inertia recovery hard constraint, during the optimization process. Driven by this high-priority constraint, the droop control coefficient is re-optimized, and state adjustment operations are performed. Specifically, to achieve the reshaped physical operating state, a series of example operations are executed: on the one hand, the energy storage unit is driven back to the reference state of charge, adjusting its output benchmark to remove it from the regulation saturation zone caused by extremely high or low charge levels; on the other hand, the hydrogen energy device is driven into a fully adjustable operating mode, removing its standby limitation in economic mode. Finally, new droop control coefficients and target virtual inertia are generated and output, thereby ensuring that energy is physically restored to the expected safe level.

[0123] While this specification provides the steps of operation for the methods described in the embodiments or flowcharts, more or fewer steps may be included based on conventional or non-inventive means. The order of steps listed in the embodiments is merely one possible order of execution among many steps and does not represent the only possible order. In actual device or client product execution, the methods shown in the embodiments or drawings may be executed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment, or even a distributed data processing environment). The terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, product, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, product, or apparatus. Without further limitations, the presence of other identical or equivalent elements in a process, method, product, or apparatus that includes said elements is not excluded. The terms "first," "second," etc., are used to denote names and do not indicate any particular order.

[0124] Those skilled in the art will also know that, besides implementing the controller using purely computer-readable program code, the same functions can be achieved by logically programming the method steps, making the controller function as logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers (PLCs), and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the devices within it used to implement various functions can also be considered structures within that hardware component. Alternatively, the devices used to implement various functions can be considered as both software modules implementing the method and structures within a hardware component.

[0125] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this specification can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solutions of this specification can essentially be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, mobile terminal, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments of this specification.

[0126] Although this specification has been described by way of examples, those skilled in the art will recognize that many variations and modifications are possible without departing from the spirit of this specification, and it is intended that the appended claims cover such variations and modifications without departing from the spirit of this specification.

Claims

1. A method for intelligent collaborative operation of an energy system, characterized in that, include: Acquire operational data of each energy storage unit within the energy system, as well as resource volatility prediction data within a preset time domain; Based on the operational data, the corresponding system energy state and system virtual inertia are determined respectively; wherein, the virtual inertia is used to characterize the system dynamic stability margin; Based on the resource volatility prediction data within the preset time domain, the target virtual inertia is determined through a preset mapping function; based on the preset system operating cost target, the droop control coefficient of each energy storage unit is determined. Based on the system energy state and the preset system energy state reference benchmark, the system energy state deviation is determined; the system energy state deviation is multiplied by the droop control coefficient of each energy storage unit to determine the power adjustment amount of each energy storage unit, so as to drive each energy storage unit to perform autonomous and coordinated adjustment. During the autonomous and coordinated adjustment process, the target virtual inertia is verified using the system's virtual inertia to obtain the verification result; Based on the verification results, when the virtual inertia of the system is less than the target virtual inertia, the droop control coefficient of each energy storage unit is updated using a preset optimization model; wherein, the preset optimization model includes at least an inertia recovery constraint. The operating reference of each energy storage unit is adjusted according to the updated droop control coefficient; wherein, the operating reference is the target output base point of each energy storage unit in the energy balance state, and the operating reference is used to adjust the state of charge of the energy storage unit so that the virtual inertia of the system is restored to a level not lower than the target virtual inertia.

2. The method according to claim 1, characterized in that, The step of determining the corresponding system energy state and system virtual inertia based on the operational data includes: The operational data is normalized to obtain multiple normalized energy storage state variables; wherein, the operational data includes the state of charge data of the electrochemical energy storage unit, the hydrogen storage pressure data of the hydrogen energy storage unit, and the demand response potential data of the adjustable load energy storage unit. The corresponding system energy state is determined by weighting and summing the multiple normalized energy storage state variables using preset weighting coefficients. Based on the operating data of each energy storage unit and the preset inertia constant, the inertia contribution factor of each energy storage unit in the current state is calculated respectively. The inertia contribution factors of each energy storage unit are linearly summed and aggregated to determine the corresponding system virtual inertia.

3. The method according to claim 1, characterized in that, The step of determining the target virtual inertia based on the resource volatility prediction data within the preset time domain and through a preset mapping function includes: Obtain meteorological forecast data within the prediction time domain, and determine the total power output prediction curve within the prediction time domain based on the meteorological forecast data; Calculate the standard deviation of the total output prediction curve in the prediction time domain, and determine the standard deviation as the resource volatility prediction data; The standard deviation is substituted into the preset mapping function as the independent variable, and the target virtual inertia is calculated and determined using the preset mapping function. The preset mapping function is a monotonically non-decreasing function, so that the target virtual inertia increases as the resource volatility prediction data increases.

4. The method according to claim 3, characterized in that, The step of substituting the standard deviation as an independent variable into the preset mapping function and using the preset mapping function to calculate and determine the target virtual inertia includes: The target virtual inertia is determined according to the following formula: in, The target virtual inertia, To predict the length of the time domain, The length of the prediction time domain starting from time t The standard deviation of the total output forecast within, It is a monotonically non-decreasing function.

5. The method according to claim 1, characterized in that, The step of determining the droop control coefficient for each energy storage unit based on a preset system operating cost target includes: Construct a multi-objective optimization function with the total cost of the energy system in the prediction time domain as the optimization objective; wherein, the total cost includes the operating cost of each energy storage unit, the degradation cost of the equipment, and the compensation cost of demand response load; Establish the constraints of the multi-objective optimization function; wherein the constraints include at least the power regulation range limit term, the power ramp rate limit term, and the operating safety range limit term for each energy storage unit; The multi-objective optimization function that satisfies the constraints is optimized using a preset optimization algorithm to determine the droop control coefficient of each energy storage unit. Based on the constraints, the multi-objective optimization function is optimized and solved to determine the droop control coefficient of each energy storage unit.

6. The method according to claim 5, characterized in that, The step of updating the droop control coefficient of each energy storage unit using a preset optimization model includes: Based on the system energy state corresponding to each energy storage unit and the inertia contribution factor, the stability adjustment priority of each energy storage unit is determined. Based on the stability adjustment priority, determine the weight coefficients in the objective function of the preset optimization model; The target virtual inertia is substituted as a parameter into the inertia recovery constraint in the preset optimization model so that the sum of the inertia provided by each energy storage unit is greater than or equal to the target virtual inertia. Using an optimization algorithm, under the premise of satisfying the inertia recovery constraint, the objective function containing the weight coefficients is optimized to determine the correction amount of each droop control coefficient; Based on the correction amount of each of the aforementioned droop control coefficients, the updated droop control coefficients are determined.

7. The method according to claim 2, characterized in that, The adjustment of the operating baseline of each energy storage unit based on the updated droop control coefficient includes: The baseline adjustment amount corresponding to each of the energy storage units is determined based on the updated droop control coefficient; wherein, the energy storage unit includes an electrochemical energy storage unit and a hydrogen energy storage unit; Based on the reference adjustment amount corresponding to the electrochemical energy storage unit, the power operation reference of the electrochemical energy storage unit is shifted from the output point corresponding to the current state of charge to the preset energy state equilibrium point, so as to correct the power adjustment space of the electrochemical energy storage unit. Based on the reference adjustment amount corresponding to the hydrogen energy storage unit, the operating mode reference of the hydrogen energy storage unit is switched from the current shutdown reference or restricted operating reference to the corresponding standby operating reference or fully adjustable operating reference, so as to release the inertia support potential of the hydrogen energy storage unit.

8. An intelligent collaborative operation device for an energy system, characterized in that, include: The data acquisition module is used to acquire the operating data of each energy storage unit in the energy system, as well as the resource volatility prediction data within a preset time domain; The first determining module is used to determine the corresponding system energy state and system virtual inertia based on the operating data; wherein the virtual inertia is used to characterize the system dynamic stability margin. The second determining module is used to determine the target virtual inertia based on the resource volatility prediction data within the preset time domain through a preset mapping function; and to determine the droop control coefficient of each energy storage unit based on the preset system operating cost target. The power regulation module is used to determine the system energy state deviation based on the system energy state and the preset system energy state reference benchmark; and to multiply the system energy state deviation with the droop control coefficient of each energy storage unit to determine the power regulation amount of each energy storage unit, so as to drive each energy storage unit to perform autonomous and coordinated regulation. The result determination module is used to verify the target virtual inertia using the system virtual inertia during the autonomous coordinated adjustment process, and obtain the verification result. The coefficient update module is used to update the droop control coefficient of each energy storage unit using a preset optimization model based on the verification results when the virtual inertia of the system is less than the target virtual inertia; wherein, the preset optimization model includes at least an inertia recovery constraint; The reference adjustment module is used to adjust the operating reference of each energy storage unit according to the updated droop control coefficient; wherein, the operating reference is the target output base point of each energy storage unit in the energy balance state, and the operating reference is used to adjust the state of charge of the energy storage unit so that the virtual inertia of the system is restored to a level not lower than the target virtual inertia.

9. An electronic device, characterized in that, It includes a processor and a memory for storing processor-executable instructions, wherein the processor, when executing the instructions, implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores computer instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1 to 7.

Images