Black-start control method for grid-type energy storage based on multi-level adaptive inertia coupling
By using a multi-level adaptive inertia coupling method, the virtual inertia and power distribution of the energy storage unit are dynamically adjusted, which solves the problems of fixed inertia and frequency fluctuation during the black start of grid-type energy storage. This enables system frequency security assessment and path reliability verification, and improves the success rate of black start and power supply reliability.
Patent Information
- Application Number
- CN202611516620.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-28
- Publication Date
- 2026-11-17
AI Technical Summary
In existing technologies, the fixed virtual inertia setting during black start-up of grid-based energy storage cannot adapt to dynamic changes, resulting in sensitivity to frequency fluctuations. It also lacks system-level frequency security assessment and dynamic power allocation strategies, and lacks reliability assessment of the black start path, which can easily lead to black start failure.
By employing a multi-level adaptive inertia coupling method, the adaptive virtual inertia of each energy storage unit is calculated in real time. Combined with system frequency response capability assessment and dynamic power allocation, a complete chain control closed loop is realized, from single-unit inertia calculation to system frequency security assessment and black start path reliability verification. The virtual inertia and power allocation are dynamically adjusted to optimize the black start process.
It effectively suppresses frequency drops, extends the continuous power supply time of energy storage units, reduces the risk of dispatch misjudgment, improves the reliability and success probability of black start path, and ensures the safe and orderly restoration of power supply to the system.
Smart Images

Figure CN123121583A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of power system recovery control technology, and relates to a black start control method and system for grid-type energy storage based on multi-level adaptive inertia coupling. Background Technology
[0002] In the event of extreme natural disasters or cascading failures, the power system may experience a complete power outage (i.e., a complete blackout). Black start refers to the process of restoring the entire power grid to operation without external grid support, relying on the system's self-starting power sources to gradually build up voltage and frequency from zero, sequentially charging transmission lines, connecting loads, and ultimately restoring the entire grid. This process typically takes tens of minutes to several hours and is the last line of defense for power system security.
[0003] Traditional black-start power sources mainly include diesel generators and hydroelectric generators. Diesel generators typically take 5 to 30 minutes to start, and their output voltage and frequency fluctuate significantly in the initial stage, requiring a long warm-up and stabilization process. Hydroelectric generators start relatively quickly (about 2 to 5 minutes), but are limited by reservoir water levels and inflow conditions, and require long-distance transmission lines for charging when far from the load center, posing a risk of overvoltage.
[0004] The common drawback of the aforementioned traditional power supplies is their slow response speed (seconds to minutes). Furthermore, during islanded operation, due to the small system inertia, the frequency is extremely sensitive to load surges, making it prone to frequency overruns that could lead to black start failures.
[0005] Grid-forming energy storage is a type of energy storage converter that employs a voltage source control strategy. Unlike common grid-following energy storage, it does not require a phase-locked loop (PLL) to track the external grid voltage. Instead, it can autonomously establish the amplitude and phase of the voltage at the grid connection point, essentially "simulating" a synchronous generator at the electrical level. Its core control strategy is Virtual Synchronous Generator (VSG) control, which uses algorithms to simulate the rotor motion equations (i.e., the swing equations) of a synchronous generator, giving the energy storage converter virtual inertia and virtual damping characteristics.
[0006] The advantages of using grid-type energy storage for black start are: First, extremely fast response speed: the inverter switching frequency is in the kHz range, and the power regulation response time is in the millisecond range; Second, no rotating parts are required: AC power is directly output from the battery chemical energy through power electronic conversion; Third, the inertia can be flexibly set: the virtual inertia can be adjusted as needed by setting control parameters. Shi K et al., “Low-Voltage Ride-Through Control Strategy of Virtual SynchronousGenerator Based on Comprehensive Current Limiting”, IEEE Trans. PowerElectronics, 2018, 33(10): 8297-8308. This paper proposes a VSG control strategy, but its virtual inertia is a fixed value and does not involve dynamic adjustment based on SOC and frequency change rate. Wu W et al., “A Coordinated Control Strategy for Black Start of Microgrid with Multiple Distributed Generations”, IEEE Trans. Smart Grid, 2019, 10(3): 2629-2638. This paper involves black-starting various distributed power sources, but uses the equal incremental rate method for power allocation and does not consider the differences in energy storage SOC and inertia coupling. Sun W et al., “PowerSystem Restoration Framework Using Multi-Agent System”, IEEE Trans. PowerSystems, 2016, 31(2): 1084-1093. This paper uses a multi-agent framework for black-start path planning, but does not involve the virtual inertia control of grid-type energy storage and its coupling relationship with power allocation.
[0007] Therefore, it is evident that existing technologies have the following key problems: First, the virtual inertia is set in a fixed manner, which cannot adapt to the dynamically changing operating conditions during black start.
[0008] In current engineering practice, when grid-based energy storage participates in black start, the virtual inertia J of each energy storage converter is usually taken as a fixed value (such as the factory setting value) and is not adjusted throughout the black start process. However, black start is a highly dynamic process—as the load is gradually brought in, the system frequency fluctuates continuously, and the state of charge (SOC) of each energy storage unit also continuously decreases. The fixed inertia scheme cannot reflect these real-time changes, resulting in insufficient inertia to effectively suppress frequency drops when the system frequency deteriorates rapidly; when the SOC of a certain energy storage unit has dropped to a low level, continuing to bear power according to the fixed inertia may prematurely deplete its power and disconnect it, causing black start failure.
[0009] Second, there is a lack of system-level quantitative assessment methods for frequency security.
[0010] Before each load is applied during black start, dispatchers need to determine whether the current system has sufficient inertia to withstand the upcoming load surge. However, there is currently a lack of a quantitative indicator that comprehensively considers the real-time status of each energy storage unit, relying mainly on the experience-based judgment of dispatchers, which carries the risk of misjudgment.
[0011] Third, the power allocation strategy among multiple energy storage units is crude.
[0012] Existing technologies often use equal distribution or allocation based on rated capacity, without considering the differences in SOC, inertia contribution, and global constraints of system frequency safety margin among units. This results in units with low SOC and units with high SOC bearing similar power, accelerating the depletion of low SOC units.
[0013] Fourth, there is a lack of prior assessment of the overall reliability of the black start path.
[0014] Black start is a multi-step sequential operation process, and the success of each step depends on the successful completion of the previous step and the current availability of each energy storage unit. Existing technologies typically only assess the feasibility of a single step and lack end-to-end reliability quantification of the entire black start path (from the first step to the last step), making it impossible to predict the success probability of the entire scheme before startup.
[0015] In summary, existing technologies lack a collaborative control method that integrates adaptive virtual inertia of grid-type energy storage, system frequency response capability assessment, dynamic power allocation, and black-start path reliability verification into a complete closed loop. Summary of the Invention
[0016] The purpose of this invention is to solve the problems in the prior art: how to achieve dynamic adaptive adjustment of the virtual inertia of each energy storage unit during the black start process of multiple grid-forming energy storage units participating in the grid, and on this basis, to establish a complete chain calculation and control closed loop from single-unit inertia calculation to system frequency security assessment, then to power optimization allocation, and finally to black start path reliability verification. This invention provides a black start control method and system for grid-forming energy storage based on multi-level adaptive inertia coupling. Specifically, it involves a control method that, after a large-scale power outage, utilizes multiple grid-forming energy storage converters as black start power sources, and achieves safe and orderly power restoration through the series coordination of four stages: adaptive virtual inertia calculation, system frequency response capability assessment, dynamic power allocation, and black start path reliability verification.
[0017] To achieve the above objectives, the present invention employs the following technical solution: In a first aspect, the present invention provides a black-start control method for grid-type energy storage based on multi-level adaptive inertia coupling, comprising the following steps: Obtain the real-time state of charge and system frequency deviation rate of change for each energy storage unit, and calculate the adaptive virtual inertia of each energy storage unit at the current moment based on the real-time state of charge and system frequency deviation rate of change. The adaptive virtual inertia of all energy storage units is summed and aggregated to obtain the system-level equivalent inertia. Combined with the load power planned to be put into the current black start step and the preset frequency allowable deviation, the system frequency dynamic response capability index is calculated. The dynamic power allocation weight of each energy storage unit is calculated based on the adaptive virtual inertia of each energy storage unit at the current moment and the system frequency dynamic response capability index. The actual power command of each energy storage unit is generated according to the dynamic power allocation weight, the load power planned to be put into the current black start step and the system frequency dynamic response capability index. Based on dynamic power allocation weights and combined with the available power of each energy storage unit at each step of the black start path, the comprehensive reliability index of the entire black start path is recursively calculated; before executing the actual power command, the comprehensive reliability index is compared with a preset safety threshold, and the actual power command is executed based on the comparison result.
[0018] In one optional implementation, the adaptive virtual inertia of each energy storage unit at the current moment is limited by its reference virtual inertia. It increases with the increase of the absolute value of the rate of change of the system frequency deviation and decays exponentially with the decrease of its own state of charge. For the specific calculation formula, please refer to Formula 1 in the specific implementation.
[0019] In one optional implementation, the system frequency dynamic response capability index is the ratio of the maximum load power that can be withstood when the frequency deviation just reaches the allowable frequency deviation under the current system-level equivalent inertia, to the load power planned to be put into the current black start step. For the specific calculation formula, please refer to Formula 2.
[0020] In one optional implementation, the dynamic power allocation weight is the normalized value of the product of the adaptive virtual inertia and the state of charge of each energy storage unit (Formula 3); the actual power command is the product of the dynamic power allocation weight, the planned load power, and the safety gate coefficient (Formula 4), wherein the safety gate coefficient is the smaller value between the system frequency dynamic response capability index and 1.0; when the system frequency dynamic response capability index is less than 1, the planned load power of the current black start step is corrected to the maximum load power that the system can withstand before it is put into operation.
[0021] In one optional implementation, the comprehensive reliability index is obtained by multiplying the single-step reliability of each step in the black start path by a series reliability model; the single-step reliability is synthesized by correcting the power margin reliability of each energy storage unit with a redundancy correction index; the state of charge at each step is updated recursively according to the actual power command of the previous step and the duration of the step, and the specific calculation formulas are shown in Formulas 5 to 8.
[0022] In one optional implementation, when the overall reliability index is greater than or equal to a first preset threshold, the actual power command is executed; when the overall reliability index is greater than or equal to a second preset threshold and less than the first preset threshold, the parameters are adjusted and the preceding steps are re-executed; when the overall reliability index is less than the second preset threshold, the black start strategy is adjusted and the preceding steps are re-executed.
[0023] Secondly, the present invention provides a grid-type energy storage black-start control system based on multi-level adaptive inertia coupling, comprising: an adaptive virtual inertia calculation module, a system frequency dynamic response capability index calculation module, an actual power command calculation module, and a comprehensive reliability quantification module, each module being used to execute the corresponding steps of the method described in the first aspect.
[0024] Thirdly, the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the method described in the first aspect.
[0025] Fourthly, the present invention provides a terminal device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method described in the first aspect.
[0026] Fifthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method described in the first aspect.
[0027] Compared with the prior art, the present invention has the following beneficial effects: This invention discloses a black-start control method for grid-type energy storage based on multi-level adaptive inertia coupling. It introduces dual coupling of State of Charge (SOC) and frequency change rate, automatically reducing the virtual inertia of low-SOC units, allowing them to withstand less power surges and effectively extending their continuous power supply time during black-start. It transforms the qualitative judgment of the system's dynamic frequency response capability index into a precise numerical indicator, allowing dispatchers to make decisions simply by comparing the magnitudes of these indices, reducing the risk of misjudgment. By calculating the dynamic power allocation weight of each energy storage unit, units with high inertia and high SOC naturally receive a larger power share. Furthermore, it adds a quantification process for the comprehensive reliability index, enabling quantitative prediction of the end-to-end success probability of the entire scheme before the actual execution of black-start. The system can adjust the scheme in advance, avoiding secondary power outages caused by mid-process failures during black-start and improving the reliability of the black-start path. Attached Figure Description
[0028] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.
[0029] Figure 1 This is a schematic diagram of the black-start control method for grid-type energy storage based on multi-level adaptive inertia coupling according to the present invention.
[0030] Figure 2 This is a schematic diagram of the structure of a grid-type energy storage black-start control system based on multi-level adaptive inertia coupling provided in an embodiment of the present invention.
[0031] Figure 3 This is a schematic diagram of the structure of a terminal device provided in an embodiment of the present invention. Detailed Implementation
[0032] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.
[0033] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.
[0034] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.
[0035] In the description of the embodiments of the present invention, it should be noted that the terms "first," "second," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance.
[0036] The present invention will now be described in further detail with reference to the accompanying drawings: See Figure 1 This invention discloses a black-start control method and system for grid-forming energy storage based on multi-level adaptive inertia coupling. The core technical problem to be solved is how to achieve dynamic adaptive adjustment of the virtual inertia of each energy storage unit during the collaborative black-start process of multiple grid-forming energy storage units, and on this basis, establish a complete chain-like calculation and control closed loop from single-unit inertia calculation to system frequency security assessment, power optimization allocation, and finally black-start path reliability verification. Specifically, this embodiment relates to a control method that, after a large-scale power outage, utilizes multiple grid-forming energy storage converters as black-start power sources, and achieves safe and orderly power restoration through the series coordination of four stages: adaptive virtual inertia calculation, system frequency response capability assessment, dynamic power allocation, and black-start path reliability verification.
[0037] Example 1 To address the aforementioned technical problems, this invention discloses a black-start control method for grid-type energy storage based on multi-level adaptive inertia coupling, which includes... The system execution of a platform-structured grid-type energy storage unit includes the following four steps connected in series: Step S1: Adaptive Virtual Inertia Calculation (Single Unit Level): Data collection for each grid-type energy storage unit ( Real-time state of charge and system frequency deviation rate Based on the improved rotor rocking equation, the calculation of each energy storage unit at... Adaptive virtual inertia at time intervals The core idea of this step is that the virtual inertia of the energy storage unit should increase as the urgency of the system frequency increases (multiple supports), while decreasing as its remaining power decreases (self-protection).
[0038] Step S2: Calculation of System Frequency Dynamic Response Capability Index (System Level): All output from step S1 Adaptive virtual inertia of Taiwan energy storage unit Summation and aggregation are performed to obtain the system-level equivalent inertia. Furthermore, this is combined with the planned load power input during the current black start procedure. and frequency tolerance Calculate the system's frequency dynamic response capability index The physical meaning reflected by this index is: under the current inertia configuration, the load impact multiple that the system can safely withstand; a value greater than 1 indicates that it can withstand it, and a value less than 1 indicates that it cannot withstand it.
[0039] Step S3: Dynamic power allocation weight calculation (allocation level): Using the adaptive inertia of each element output in step S1 The system frequency response index output in step S2 Calculate the dynamic power allocation weight for each energy storage unit. Based on this, the actual power command for each unit is generated. The core idea of this step is: units with large inertia contributions and sufficient SOC should output more power, while the overall power should be reduced when the system frequency margin is insufficient.
[0040] Step S4: Comprehensive Reliability Verification of the Black Boot Path (Decision Level): Utilize the power allocation weights of each unit output in step S3 (Used to calculate the actual power command for each unit), combined with the available power of each unit. Calculate the single-step reliability of each black start step. Then, based on the series reliability model and redundancy correction mechanism, the entire black start path (total...) is calculated. Comprehensive reliability index of (step) This metric is used for the final decision: whether to execute a black boot as per the current plan.
[0041] The above four steps constitute a chain-like computation process, where the output of the previous step is directly used as the input of the next step, forming a complete computational loop.
[0042] Example 2 This embodiment elaborates on the calculation process of each step based on Embodiment 1, specifically including the following steps: The detailed implementation of step S1 includes: The objective of step S1.1 is to calculate a dynamic virtual inertia value for each grid-type energy storage unit in real time, so that the inertia value can reflect both the amount of support required by the system and the amount that the energy storage unit can still support.
[0043] Step S1.2: Original Formula and its Physical Background The rotor motion of a synchronous generator follows Newton's second law in a rotating system, known as the swing equation. The classical formulation of this equation can be found in the following literature: Source: Anderson PM, Fouad A A. Power System Control and Stability [M]. 3rd ed. Wiley-IEEE Press, 2017: Chapter 2, Equation (2.10).
[0044] The original swing equation is:
[0045] in, Represents the rotor's moment of inertia, in units of The physical meaning of is the magnitude of the rotor's inertia in resisting changes in rotational speed; Represents the actual angular velocity of the rotor, in units of ; This represents the mechanical power input to the rotor of a prime mover (such as a steam turbine or a water turbine), expressed in units of... ; This represents the electromagnetic power output by the generator, in units of... ; Indicates the damping coefficient, in units of This reflects the damping torque generated when the rotor speed deviates from the rated value; Indicates the rated angular velocity, in units of Corresponding to the rated frequency The relationship is .
[0046] The physical meaning of this equation can be simply understood as follows: the acceleration or deceleration of the rotor depends on the difference between the input power and the output power, and the inertia. The larger the value, the slower the change in rotational speed caused by the same power difference.
[0047] Step S1.3: Generalization from physical inertia to virtual inertia Grid-type energy storage has no physical rotating parts; its inertia is virtually realized in the converter controller through a VSG control algorithm. Specifically, the VSG controller solves the discrete form of the aforementioned swing equation in software and calculates the virtual rotational speed. This serves as a reference value for the converter output voltage frequency. When the system load increases ( When the power is increased, a power deficit appears on the right side of the equation, and the virtual speed increases. The frequency drop was simulated, which simulated the frequency drop process of a synchronous generator.
[0048] In this context, the equation It is no longer the actual inertia of the physical rotor, but a freely settable control parameter called virtual inertia. Increasing... It can slow down frequency changes (making the system more stable) and reduce It can accelerate frequency changes (making the system more sensitive).
[0049] The existing technology approach is to It is set to a fixed constant that does not change with operating conditions.
[0050] Step S1.4: Improved derivation of the present invention The core improvement of this invention lies in: making virtual inertia The system changes from a fixed constant to a time-varying variable that dynamically adjusts with the system frequency rate and the energy storage state of charge (SOC). .
[0051] The derivation process is as follows: The first step is to make ,in The baseline virtual inertia (fixed value) is used. This is a dynamic adjustment factor (time-varying, dimensionless). Substituting it into the swing equation:
[0052] The second step is to confirm. Design principles. Based on physical intuition: When the system frequency changes rapidly (whether rising or falling), it indicates that the isolated grid is experiencing a significant power imbalance impact. The virtual inertia should be increased accordingly to slow down the rate of frequency change, buying time for other adjustment methods. Therefore... Should be with (The absolute value of the rate of change of system frequency deviation) is positively correlated, that is, the more drastic the frequency change, the stronger the inertia support.
[0053] When the energy storage unit's State of Charge (SOC) is low, its remaining usable energy is limited. If it continues to operate with high inertia, it will deplete its power too quickly due to the continuous provision of inertia support, causing the unit to go offline and worsening the system's condition. Therefore... Should be with Positive correlation.
[0054] The third step is design. The specific mathematical expression is as follows. Considering both factors mentioned above, a multiplicative coupling structure is adopted:
[0055] In the formula: because Non-negative, "1+" ensures That is, the adaptive inertia is always no less than the reference value (ensuring the most basic inertia support capability). For the frequency change rate driving term: Divide by Achieve normalization (convert Hz / s to per-unit rate of change relative to the rated frequency). This is the gain coefficient (unit: s), used to adjust the sensitivity. SOC constraint terms: when When fully charged, the exponent term is: No constraints are generated; when When it decreases, As the exponential term increases, it decays rapidly, effectively suppressing the amplification effect of inertia. (Dimensionless) Control the decay rate.
[0056] Why use an exponential function? Instead of a simple linear function ? The reason lies in the discharge characteristics of lithium iron phosphate batteries: at a high SOC (0.3~1.0), the battery terminal voltage changes gradually, and the usable power is basically unaffected by the SOC; however, when the SOC drops below 0.3, the terminal voltage drops sharply, and the usable power shrinks rapidly. An exponential function can naturally simulate this nonlinear characteristic of slow decay at high SOC and rapid decay at low SOC, while a linear function cannot reflect this inflection point effect.
[0057] The fourth step is to Substituting back, we get the final formula: (Formula 1) Step S1.5: Verification of Dimensional Consistency For a dimensional analysis of the second term within the square brackets on the right side of the equation, see Table 1.
[0058] Table 1 Dimensional Analysis Reference Table
[0059] Second dimension = (dimensionless) ; Square brackets as a whole = ; The right side of the equation = Dimensions and Same, for .
[0060] Left side of the equation The dimensions are also .
[0061] Left and right have consistent dimensions .
[0062] Step S1.6: How to obtain each parameter (a) —Base virtual inertia These are the factory default parameters of the VSG controller, and their typical calculation method is as follows:
[0063] in: This is the equivalent inertial time constant (unit: seconds), typically ranging from 2 to 10 seconds. The selection of this value can refer to the typical inertial time constant of a synchronous generator. Reference: Kundur P. Power System Stability and Control [M]. McGraw-Hill, 1994: Table 3.1, which lists the equivalent inertial time constants for different types of synchronous generators. Value range (2.5~10s for steam turbine generators, 2~4s for hydro turbine generators). In this invention, [the value is taken as...]. As a typical value.
[0064] For the first Rated apparent power of the energy storage converter (unit: VA).
[0065] .
[0066] at rated power Power factor For example: ,but .
[0067] (b) : Frequency change rate coupling gain coefficient The results were obtained through offline simulation tuning, and the specific steps are as follows: 1. Build an islanded system model containing a single grid-type energy storage unit in MATLAB / Simulink or PSCAD / EMTDC simulation software; 2. Apply a step load disturbance (such as a sudden increase of 20% of the rated power) to the model and record different... The system frequency response curve for the range of values (scanned in 0.1s increments within the range of 0.1 to 5.0s); 3. Using the constraint that the lowest frequency point (nadir) is no lower than 49.0Hz and the frequency recovery time is no more than 10s, select the minimum value that satisfies the constraint. Value (to avoid excessive amplification of inertia leading to sluggish frequency response); 4. Typical tuning results are: .
[0068] (c) System frequency deviation rate of change (RoCoF), this parameter is acquired in real time by the frequency measurement module inside the energy storage converter controller. Specific acquisition method: The converter controller uses a sampling frequency (i.e., sampling period) Sample the voltage at the grid connection point; The instantaneous frequency of each sampling period is calculated using the zero-crossing detection method or the Discrete Fourier Transform (DFT). ; Frequency deviation ; The rate of change of frequency is calculated using the difference method: ; To reduce the impact of noise, a first-order low-pass filter is usually applied to the RoCoF signal in practical engineering, with a cutoff frequency of 10~20Hz.
[0069] (d) System rated frequency The rated frequency of China's power system is 50Hz.
[0070] (e) SOC attenuation sensitivity coefficient Calibration is achieved through battery discharge experiments. Specific steps: 1. Conduct constant power discharge experiments on the target model of lithium iron phosphate battery cells and record the maximum sustainable output power under different SOCs; 2. Using the maximum sustainable output power at SOC=1.0 as a benchmark, calculate the power availability ratio at each SOC; 3. Using exponential functions Least squares fitting was performed on the power availability curve; 4. Typical fitting results: At this point, when SOC=0.3, the exponential term This means that the inertia amplification capability of low SOC cells is compressed to about 12% of that at full charge, which is consistent with the decay trend of the actual battery power available ratio.
[0071] (f) State of charge The energy storage BMS reports the SOC to the EMS in real time. The BMS typically uses an ampere-hour integration method combined with an open-circuit voltage (OCV) calibration method to estimate the SOC. See: Piller S, Perrin M, Jossen A. Methods for state-of-chargedetermination and their applications [J]. Journal of Power Sources, 2001, 97-98: 1-9. The basic formula for the ampere-hour integration method is: ,in This refers to the battery's rated capacity (Ah). The discharge current is (A).
[0072] The SOC update cycle is typically 1 to 5 seconds, which meets the real-time requirements of the black start process (time scale is minutes) of this invention.
[0073] Furthermore, the output of step S1 is Adaptive virtual inertia value These values will be directly passed to step S2 as input.
[0074] The detailed implementation of step S2 includes: Step S2.1: The goal of step S2 is to aggregate the individual inertia of each energy storage unit into a global index that reflects the frequency tolerance of the entire system, providing a basis for power allocation and safety decisions in subsequent steps.
[0075] Step S2.2: Original Formula and its Physical Background (a) Equivalent inertia formula for multi-machine systems In power systems containing multiple generators, the overall system inertia level is typically characterized by the capacity-weighted average equivalent inertial time constant. The formula can be found here: Source: Anderson PM, Fouad A A. Power System Control and Stability [M]. 3rd ed. Wiley-IEEE Press, 2017: Chapter 4, Equation (4.23).
[0076]
[0077] The physical meaning of this formula is: the total kinetic energy of the system is the sum of the kinetic energies of each unit, aggregated with capacity as the weight.
[0078] (b) Prediction relationship of the lowest frequency point (Nadir) Under a step load disturbance, the lowest point of the system frequency is approximately inversely proportional to the system inertia and the disturbance power. This relationship can be derived from the energy conservation interpretation of the rocking equation; see: Source: Ulbig A, Borsche TS, Andersson G. Impact of Low RotationalInertia on Power System Stability and Operation [J]. IET Renewable PowerGeneration, 2014, 8(6): 729-738, formula (3).
[0079]
[0080] The physical meaning of this formula is: during the inertia support phase (before the first frequency modulation operation), the frequency deviation is approximately equal to the power deficit. Proportional to the total inertia of the system Inversely proportional.
[0081] Step S2.3: Improved derivation of the present invention The first step is to substitute the adaptive virtual inertia output from step S1 into the equivalent inertia calculation. Since the virtual inertia of grid-type energy storage is directly... ( Units are used, no need to go through and Indirect conversion, the system-level equivalent total inertia is directly:
[0082] The total kinetic energy stored in the system is (Unit: Joules, J).
[0083] The second step is to define the system frequency dynamic response capability index. Physical meaning: In the current system inertia Below, the frequency deviation is made to just reach the upper limit of the allowable range. Maximum load step power that can be withstood With respect to the currently planned load capacity The ratio of .
[0084]
[0085] Third step, derivation The expression is derived from the approximate energy conservation of the prediction relationship at the lowest frequency point: During the inertia-supported phase, the change in the system's kinetic energy equals the integral of the power deficit over time. Let the frequency start from... Offset to The time is Then there is an approximate relationship:
[0086] because (0.5Hz is much smaller than 50Hz), using a first-order approximation:
[0087] Substituting, we get:
[0088]
[0089] The inverse solution yields:
[0090] The fourth step is to Using the reference time constant Replacement. This represents the time required for the frequency to shift to its limit value, and is related to the transient response characteristics of the VSG control loop. In this invention, we take... This represents the typical settling time of the frequency transient response under VSG control, to eliminate the influence of unpredictable factors. Dependence.
[0091] Will Substitution Definition:
[0092] Expand : (Formula 2) Further verification of dimensional consistency is shown in Table 2.
[0093] Table 2 Verification Table of Dimensional Consistency
[0094] Combine the dimensions on the right side:
[0095]
[0096]
[0097]
[0098]
[0099] Left side of the equation It is dimensionless.
[0100] Left and right have consistent dimensions .
[0101] Furthermore, the methods for obtaining each parameter are shown in Table 3.
[0102] Table 3. Methods for obtaining each parameter
[0103] Furthermore, this embodiment discloses The physical meaning and criteria of this.
[0104] The physical meaning can be simply understood as the safety margin multiple that the system can withstand under load impact: The system's current inertia is sufficient to withstand the planned load power, and there is an additional margin, so the black start procedure can continue.
[0105] The system inertia margin is insufficient; if the load is applied as originally planned, the frequency deviation will exceed the allowable upper limit. It should be adjusted accordingly. Reduce the load capacity of this step (i.e., only deploy the maximum load that the system can currently withstand). Then proceed with the execution.
[0106] The system inertia is severely insufficient. Even if the load is reduced to the maximum extent, there is still a risk of exceeding the frequency limit, which may lead to low-frequency load shedding or even black start failure. This step should be paused and the black start plan should be adjusted.
[0107] Furthermore, the output of step S2 is the system frequency dynamic response capability index. (Dimensionless scalar), this value will be passed to step S3 as a global security constraint signal.
[0108] Detailed implementation of step S3: Step S3.1: The objective of step S3 is to rationally allocate the power share that each energy storage unit should bear in the current black start step, under the premise of ensuring system frequency safety, so that the stronger units output more power and the weaker units output less power, and automatically reduce the total power when the system frequency margin is insufficient.
[0109] Step S3.2: Original Formula and its Physical Background (a) Participation Factor Method In automatic generation control (AGC) and economic dispatch of power systems, the participation factor method is used to allocate the total power regulation demand of the system to each generating unit. Its basic formula can be found here: Source: Wood AJ, Wollenberg BF. Power Generation, Operation, and Control [M]. 2nd ed. Wiley, 2013: Chapter 3, Formula (3.14).
[0110]
[0111] in, For the first The participation factors of the Taiwanese unit satisfy Traditional participation factors are usually determined as a percentage of the unit's rated capacity or using the equal increment rate method, and are therefore fixed values.
[0112] Step S3.3: Improved derivation of the present invention The first step is to define the comprehensive contribution potential factor for each energy storage unit. .
[0113] Traditional participation factor methods allocate resources based solely on rated capacity, neglecting the dynamic state of energy storage. This invention couples the following two factors into a comprehensive contribution potential factor:
[0114] Design rationale: (From step S1): Reflects the inertia support capability of the unit under the current operating conditions. A large inertia means that the unit can provide a stronger inertial response when the frequency is disturbed, and is suitable for handling more power.
[0115] This reflects the unit's remaining energy margin. A high SOC means that the unit has sufficient "stock" to continuously supply power.
[0116] The product of the two It comprehensively represents the combined ability to both withstand (large inertia) and have surplus usable energy (high SOC).
[0117] Note: The output of step S2 We do not proceed directly to weight calculation here because the weight... It describes the relative proportional relationship between the units, while As a system-level indicator, its role is reflected in the constraint of the absolute value of the total power (see below). If the numerator and denominator are multiplied simultaneously by... After normalization, it will be canceled out and will not have a substantial impact. Therefore, this invention will Its role is placed in the power command calculation stage.
[0118] The second step is to normalize the comprehensive contribution potential factor to obtain the dynamic power allocation weights: (Formula 3) This can be directly verified by the definition: This satisfies the weight normalization requirement.
[0119] The third step is to calculate the actual power command for each unit. This involves importing the output from step S2. As a global safety gate: (Formula 4) Mechanism of action: when hour, The system has sufficient frequency margin, and all planned load power is allocated to each unit according to its weight. Total power command. .
[0120] when hour, The system frequency margin is insufficient. It can be seen that, This means that the EMS will adjust the load power to be deployed in this step of the plan to the maximum load that the system can currently safely withstand. Then, during the commissioning operation, the power commands for each energy storage unit are reduced accordingly based on their weights. In this way, the actual load power is automatically adjusted to a level that the system inertia can safely withstand, avoiding the risk of exceeding frequency limits due to forcibly applying all planned loads.
[0121] This design reflects the principle that allocation considers individuals (weights), while the total amount considers the system (weights). The hierarchical control concept.
[0122] Furthermore, this embodiment discloses a dimensional consistency verification method.
[0123] Weighting formula : molecular: Dimensions = ; Denominator: Dimensions = ; Ratio = (Dimensionless) = Left side Dimensions ; Power command formula : Dimensionless; : (i.e., W); Dimensionless; Dimensions on the right = = Left side Dimensions ; Left and right have consistent dimensions ; Furthermore, the methods for obtaining each parameter are shown in Table 4.
[0124] Table 4. Methods for obtaining each parameter
[0125] Furthermore, step S3 outputs two sets of data: 1. Dynamic power allocation weights for each unit (Dimensionless), pass to step S4; 2. Actual power command for each unit (Unit: W), directly sent to each energy storage converter controller for execution.
[0126] Detailed implementation of step S4.
[0127] The goal of step S4.1 is to: before performing a black boot, process the entire black boot path (from step 1 to step 2). The success rate of each step is quantitatively assessed end-to-end to provide dispatchers with a basis for "Go / No-Go" decision-making.
[0128] Step S4.2 Original Formula and its Physical Background (a) Reliability model of series system In reliability engineering, a series system is a structure where the failure of any component in the system leads to the failure of the entire system. The reliability calculation formula can be found here: Source: Ebeling C E. An Introduction to Reliability and MaintainabilityEngineering [M]. 2nd ed.Waveland Press, 2019: Chapter 4, Formula (4.3).
[0129]
[0130] in, For the first The reliability of each component (a probability value between 0 and 1).
[0131] (b) Exponential distribution failure rate model The reliability of a component at a given moment can be described by an exponential distribution:
[0132] in, This refers to the failure rate. See Chapter 3 of the same reference.
[0133] Step S4.3 Improved derivation of the present invention The first step is to define the power margin reliability of a single energy storage unit during a single black start-up step.
[0134] In the The time corresponding to each black start step , No. The challenge facing Taiwan's energy storage unit is whether it has enough available power to meet the power command allocated in step S3. Define the power margin ratio:
[0135] in: This represents the available power of the cell at the current SOC. Linearization is used here as an engineering simplification in the pre-reliability assessment: for lithium iron phosphate batteries, in Within this range, the available power is approximately proportional to the State of Charge (SOC). It should be noted that the exponential decay term in step S1 characterizes the active voltage drop of the inertia support capability under low SOC conditions (belonging to a control and protection strategy), which acts on a different stage than the linear estimation of available power here; the two are not contradictory. The residual introduced by this simplification is reflected in the redundancy correction exponent. This was incorporated into the Monte Carlo calibration process; The actual power command calculated in step S3.
[0136] Physical meaning: The available power is greater than the commanded power, indicating that the unit has sufficient power margin. The available power is exactly equal to the commanded power, and it is in the critical state of "full load". If the available power is insufficient to complete the instruction, the unit will have insufficient power margin.
[0137] The power margin ratio is mapped to reliability (a probability value between 0 and 1) using an exponential function: (Formula 5) Reasons for choosing the exponential function: when hour, The reliability approaches zero (completely unreliable); when hour, The critical state has a reliability of approximately 63% (taking into account the uncertainties of measurement errors and transient fluctuations). when hour, The reliability approaches 1 (completely reliable); The mapping relationship is monotonically increasing and smooth, which aligns with engineering intuition.
[0138] The second step is to calculate the system reliability of a single black start step.
[0139] In a single step, The energy storage units share power, similar to a parallel redundant structure—even if one unit's power is insufficient, other units can still compensate by reallocating power. However, the classic parallel reliability formula... Assuming that each component is completely independent and has a redundancy of 1 (meaning that any one component is sufficient), this does not fully meet the black start scenario.
[0140] This invention introduces a redundancy correction index. The reliability of the single-step system is corrected as follows: (Formula 6) Physical meaning: Equivalent to the classic series model, with no redundancy compensation capability. When all If the reliability of any one of the machines in Taichung is low, the overall system reliability will be directly affected.
[0141] This indicates that there is redundancy compensation capability between multiple energy storage units. Because... ,That The power of (n) Increases and decreases, therefore Follow Increase and rise: The larger the value, the stronger the redundancy compensation capability, and the more lenient the single-step reliability assessment result; The closer the value is to 1, the closer the evaluation is to the series case without redundancy compensation, and the more conservative the result. Actual values should be calibrated through Monte Carlo simulation (see the parameter acquisition section below) to ensure... The calculation results are consistent with the statistical success rate.
[0142] Typical values are determined through Monte Carlo simulation (see the parameter acquisition section below).
[0143] The third step is to calculate the overall reliability index of the entire black start path.
[0144] The black boot path includes The process involves several steps, which are interconnected; failure in any step results in the failure of the entire black boot process. Therefore: (Formula 7) in:
[0145] Notice Time-based SOC updates: Because black start is a gradual process, each energy storage unit updates its SOC at the specified time. The SOC of the step depends on the previous step. The cumulative discharge amount per step. The SOC update formula is: (Formula 8) in: For the first The duration of a step (i.e., the steady-state operating time from the start of the load in this step to the start of the next step) is typically 3 to 10 minutes. For the first Rated energy storage capacity of the energy storage unit (unit: Wh or J).
[0146] This means that the calculation in step S4 is recursive: from Begin by calculating each step step by step. , Simultaneously update the SOC until... .
[0147] Furthermore, this embodiment discloses a dimensional consistency verification method: Dimensions: molecular : ; denominator : ; ratio: (dimensionless) ; = Dimensionless ; = Dimensionless ; Dimensions: Dimensionless product = Dimensionless ; Dimensionless to the dimensionless power = dimensionless ; - = Dimensionless ; Dimensionless product = Dimensionless
[0148] Left and right have consistent dimensions .
[0149] Furthermore, the methods for obtaining each parameter are shown in Table 5.
[0150] Table 5. Methods for obtaining each parameter
[0151] Detailed description of the calibration method: 1. Build a simulation environment containing An islanded system model for grid-type energy storage; 2. Set the initial SOC and rated parameters for each energy storage unit; 3. In each step, a uniformly distributed perturbation of ±10% is randomly applied to the available power of each unit (to simulate uncertainties such as SOC estimation error and temperature influence). 4. Execute the complete During the black start process, the success of the black start is recorded (no frequency exceedance or energy storage overload occurred in any step). 5. Repeat the Monte Carlo simulation 10,000 times and calculate the success rate. ; 6. Adjustment The value that makes the formula calculate Simulation with Monte Carlo The mean square error is the smallest; 7. Regarding Taiwan energy storage, Typical configuration of the step, calibration results .
[0152] Further, see Table 6 for the decision criteria.
[0153] Table 6 Decision Criteria Reference Table
[0154] Furthermore, the output of step S4 is the comprehensive reliability index of the black start path. (Dimensionless numerical values between 0 and 1) are provided for dispatchers to make final decisions.
[0155] Based on the above steps, this embodiment summarizes the complete control flow.
[0156] Combination Figure 1 The complete execution flow of the method of this invention is divided into two stages: "pre-black start evaluation" and "real-time correction during black start execution", as detailed below: Step 0: Preparations before Black Boot 1. The EMS receives a power grid outage signal and activates the black start contingency plan; 2. Confirm participation in black boot All Taiwan-structured grid-type energy storage units are online and the converters are in standby mode; 3. Read the initial SOC from each BMS and the black start load sequence from the EMS database. Step, each step ); 4. Set global parameters .
[0157] Step 1: Activate grid-based energy storage to establish isolated grid voltage and frequency. 1. Each energy storage converter starts VSG control and establishes black start bus voltage with reference to rated voltage and rated frequency; 2. Wait for the voltage and frequency to stabilize (approximately 5-10 seconds).
[0158] Step 2: Pre-black start calculations (for each black start step) (Recursive execution) During the preliminary simulation phase, no actual load was applied. The frequency change rate for each step was a predicted value (obtainable from historical disturbance records or simulation estimates based on similar load applications). The state of charge was updated recursively according to Formula 8. For each step... Perform the following sub-steps: Sub-step 2a (corresponding to step S1): Read the predicted frequency change rate of the current black start step (During the real-time correction phase, the measured values of the converter controller are used.) Read the current status of each unit from the BMS. ; Calculate each unit according to Formula 1. .
[0159] Sub-step 2b (corresponding to step S2): Each Summation ; Calculate according to formula 2 ; Judgment: If (Severely insufficient) An alarm is issued, and it is recommended to suspend black boot.
[0160] Sub-step 2c (corresponding to step S3): Calculate each unit according to formulas 3 and 4. and This serves as a preliminary power command (not to be issued for the time being). Will The data is then distributed to each converter controller.
[0161] Sub-step 2d (corresponding to step S4): Calculate according to formulas 5 and 6. , ; Update each unit recursively according to Formula 8. .
[0162] Step 3: Verify the reliability of the black boot path (pre-execution decision) Complete all After the recursive pre-calculation of the steps, calculate according to Formula 7. ; Output conclusions and recommendations based on decision-making criteria.
[0163] Step 4: Perform a black boot Based on the decision conclusion of step 3, after confirmation by the dispatcher, the load input operations of each step are executed in sequence according to the plan, and the power commands of each step obtained from the pre-rehearsal are sent to the controllers of each energy storage converter.
[0164] In actual execution, after each load step is applied, sub-steps 2a~2c (i.e., steps S1~S3) are re-executed based on the latest measured rate of frequency change and state of charge, dynamically updating the power commands for subsequent steps; the comprehensive reliability index of step S4 is a pre-execution assessment and does not need to be repeatedly extrapolated during execution. This forms a complete control closed loop of "pre-execution decision-making and real-time correction during execution".
[0165] Numerical Examples To verify the feasibility of the above calculation process, a complete set of numerical examples is provided. The simulation object is a black start scenario of a 110kV distribution network: a total of N=4 grid-type energy storage units participate in the black start, each with a rated power of 20MW and a rated capacity of 40MWh, and an equivalent inertial time constant H=5s. The baseline virtual inertia Jbase,i=2026.4 kg·m² (i=1,2,3,4) is calculated by the formula; the black start path has a total of M=8 steps, and the duration of each step is Δtm=300s. The global parameters and initial state are shown in Table 8, and the black start load sequence and the predicted frequency change rate of each step are shown in Table 9.
[0166] Table 8 Parameter Table for Numerical Examples
[0167] Table 9 Black Start Load Sequence and Predicted Frequency Change Rate at Each Step
[0168] Taking the fourth step, which has the heaviest load, as an example, the calculation process and results according to formulas 1 to 8 are shown in Table 10. The state of charge in the first three steps has been updated recursively according to formula 8.
[0169] Table 10 Calculation details for step 4 (ΔPstep = 1.2MW, |d(Δf) / dt| = 0.25Hz / s)
[0170] In step 4, the system-level equivalent inertia Jeq(t4) = 8120.1 kg·m², the system frequency dynamic response capability index Φsys(t4) = 1.336 > 1, the safety gate coefficient min(Φsys,1.0) = 1.0, and the planned load is fully allocated to each unit according to its weight; the single-step reliability Rstep(t4) ≈ 1.0000. The calculation results of all 8 steps are summarized in Table 11.
[0171] Table 11 Summary of Calculation Results for Each Black Startup Step
[0172] From Formula 7, the overall reliability index RBS of the entire black start path is approximately 1.0000, which is greater than the first preset threshold of 0.95. Therefore, the reliability of the current scheme is deemed sufficient, and black start is executed according to the current scheme. In this example, the power margin at each step is sufficient, so ri is close to 1. When the planned load of a single step approaches the available power of a certain unit (such as when some units are taken out of operation or the SOC is too low), the corresponding ri will drop significantly, and RBS will fall into the middle range of Table 6, triggering adjustment mechanisms such as "reducing the single-step load" or "increasing the number of energy storage units". These will not be elaborated here.
[0173] Comparative verification: Using the existing technology's fixed inertia and proportional distribution scheme based on rated capacity (each unit has a weight of 0.25), the #4 energy storage unit (initial SOC=0.45) carries 131.25 kWh of electricity in the entire black start path; after adopting the method of this invention, the electricity carried by the #4 unit is reduced to 84.30 kWh, a reduction of approximately 36%, which effectively extends the continuous power supply time of the low SOC unit, verifying the coupling effect between steps S1 and S3.
[0174] Furthermore, the present invention also discloses other embodiments.
[0175] (1) Regarding alternatives to grid-type energy storage: This invention is not limited to lithium iron phosphate battery energy storage, but can also be applied to other types of grid-type energy storage systems such as vanadium redox flow batteries, sodium-ion batteries, and flywheel energy storage. For different types of energy storage, , These parameters need to be readjusted based on their power-capacity characteristics.
[0176] (2) Regarding the alternative to the VSG control strategy: The adaptive inertia calculation in step S1 of this invention is not limited to the classical VSG swing equation form, but can also be applied to a hybrid strategy of droop control and VSG control. Under the hybrid strategy, only the output of step S1 needs to be... Simply replace it with the corresponding virtual inertia parameter.
[0177] (3) Regarding the substitution of the constant term in the formula: the exponential function in Formula 1 It can also be replaced with other functional forms with similar decay characteristics, such as (In rational function form), the effect of slow decay at high SOC and rapid decay at low SOC is similar.
[0178] (4) Regarding the change in the number of steps in the black boot path: The value of is not limited to the 8 steps in the embodiment, and can be flexibly set according to the actual power grid scale and the number of loads to be restored. When it increases, The energy density will naturally decrease due to more steps connected in series; in this case, the number of energy storage units can be increased. or reduce To improve reliability.
[0179] (5) Regarding the method of obtaining the frequency change rate: In addition to the frequency measurement module inside the converter, It can also be obtained through the synchronous phasor measurement unit (PMU) in the wide area measurement system (WAMS), which is suitable for scenarios where multiple energy storage units are distributed in different geographical locations.
[0180] Furthermore, a comparison and summary of the present invention and the prior art is shown in Table 7.
[0181] Table 7 Summary of Comparison with Existing Technologies
[0182] Example 3 See Figure 2 This invention discloses a grid-type energy storage black-start control system based on multi-level adaptive inertia coupling, comprising: The adaptive virtual inertia calculation module is used to obtain the real-time state of charge and system frequency deviation change rate of each energy storage unit, and calculate the adaptive virtual inertia of each energy storage unit at the current moment based on the real-time state of charge and system frequency deviation change rate. The system frequency dynamic response capability index calculation module is used to adaptively sum and aggregate the adaptive virtual inertia of all energy storage units to obtain the system-level equivalent inertia. Combined with the load power planned to be put into the current black start step and the preset frequency allowable deviation, the system frequency dynamic response capability index is calculated. The actual power command calculation module is used to calculate the dynamic power allocation weight of each energy storage unit based on the adaptive virtual inertia of each energy storage unit at the current moment and the system frequency dynamic response capability index, and generate the actual power command of each energy storage unit according to the dynamic power allocation weight, the expected load power and the system frequency dynamic response capability index. The comprehensive reliability quantification module is used to calculate the comprehensive reliability index of the entire black start path based on the dynamic power allocation weight and the available power of each energy storage unit. Based on the comprehensive reliability index and the preset safety threshold, it selects whether to execute the actual power command.
[0183] Furthermore, the system to which this invention applies includes: Platform-based grid-type energy storage unit: Each energy storage unit ( It consists of a battery cluster, a battery management system (BMS), a DC / AC converter, and a VSG controller. The converter adopts a VSG control strategy and can autonomously establish the grid connection point voltage and frequency. The typical value is 2 to 8 units, in this embodiment... .
[0184] Black start bus: Each energy storage unit is connected to the same black start bus through its own step-up transformer.
[0185] Distribution substation (load node): The black starting bus gradually supplies power to the load through transmission lines and distribution switches. Each distribution transformer area is powered on. Each distribution transformer area corresponds to a step on the black start path. In this embodiment... .
[0186] Energy Management System (EMS): Responsible for executing the calculation process of this invention, collecting SOC data from each BMS, collecting frequency and frequency change rate data from the converter controller, determining the load input power at each step according to the black start scheme, and sending the calculated power command to each converter controller.
[0187] Compared with the prior art, the present invention has the following advantages: (1) Adaptive inertia to prevent premature depletion of low SOC units. By introducing dual coupling of SOC and frequency change rate in step S1, the virtual inertia of low SOC units is automatically reduced, allowing them to bear less power surge and effectively extending their continuous power supply time during black start. Taking the above numerical example, in the heaviest load step 4, the adaptive inertia of the #4 energy storage unit with SOC of 0.45 is only 1.0008 times the baseline value, while that of the #1 unit with SOC of 0.85 is 1.0025 times. After the cumulative effect of 8 black start steps, the amount of electricity carried by the #4 unit is reduced from 131.25 kWh in the fixed inertia equal distribution scheme to 84.30 kWh, a reduction of about 36%.
[0188] (2) Frequency security quantification eliminates the reliance on the experience of dispatchers. The output of step S2 is... This transforms the qualitative judgment of whether the system can withstand the next load increase into a precise numerical indicator, which dispatchers only need to compare. The relationship between the value and the value of 1.0 can be used to make decisions, reducing the risk of misjudgment.
[0189] (3) Power allocation balances efficiency and safety. In step S3, the weighting formula... This allows units with high inertia and high SOC to naturally obtain a larger share of power, while The function acts as a global safety gate, automatically reducing the total power command when the system frequency margin is insufficient, thus avoiding the contradiction of individual optimization but system insecurity.
[0190] (4) Path-level reliability prediction improves the success rate of black boot. The output of step S4 is... It can quantitatively predict the end-to-end success probability of the entire solution before the actual execution of the black start. When When the load is below the threshold, the system can adjust the plan in advance (such as adding energy storage units or reducing the single-step load) to avoid secondary power outages caused by failure during black start.
[0191] (5) No new hardware is required, and the engineering implementation cost is low. All the input signals required by this invention (SOC, frequency, rate of change of frequency, and planned load power) can be obtained from the existing sensors and communication interfaces of the existing energy storage BMS (battery management system), EMS (energy management system), and converter controller. No new hardware equipment is required; only the corresponding calculation module needs to be added to the control software of the energy storage converter.
[0192] See Figure 3This is a schematic diagram of the structure of a terminal device provided in an embodiment of the present invention. The terminal device of this embodiment includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps in the various method embodiments described above. Alternatively, when the processor executes the computer program, it implements the functions of each module / unit in the various device embodiments described above.
[0193] The computer program can be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention.
[0194] The terminal device can be a desktop computer, laptop computer, cloud server, or other device with strong computing power. The terminal device may include, but is not limited to, a processor and memory.
[0195] The processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc.
[0196] The memory can be used to store the computer program and / or module. The processor implements various functions of the terminal device by running or executing the computer program and / or module stored in the memory and calling the data stored in the memory.
[0197] If the modules / units integrated into the terminal device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.
[0198] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A black-start control method for grid-type energy storage based on multi-level adaptive inertia coupling, characterized in that, Includes the following steps: Obtain the real-time state of charge and system frequency deviation rate of change for each energy storage unit, and calculate the adaptive virtual inertia of each energy storage unit at the current moment based on the real-time state of charge and system frequency deviation rate of change. The adaptive virtual inertia of all energy storage units is summed and aggregated to obtain the system-level equivalent inertia. Combined with the load power planned to be put into the current black start step and the preset frequency allowable deviation, the system frequency dynamic response capability index is calculated. The dynamic power allocation weight of each energy storage unit is calculated based on the adaptive virtual inertia of each energy storage unit at the current moment and the system frequency dynamic response capability index. The actual power command of each energy storage unit is generated according to the dynamic power allocation weight, the load power planned to be put into the current black start step and the system frequency dynamic response capability index. Based on dynamic power allocation weights and combined with the available power of each energy storage unit at each step of the black start path, the comprehensive reliability index of the entire black start path is recursively calculated; before executing the actual power command, the comprehensive reliability index is compared with a preset safety threshold, and the actual power command is executed based on the comparison result.
2. The black-start control method for grid-type energy storage based on multi-level adaptive inertia coupling according to claim 1, characterized in that, The adaptive virtual inertia of each energy storage unit at the current moment is calculated by the following formula: in, Represents the baseline virtual inertia; This represents the coupling gain coefficient based on the rate of change of frequency. Represents the absolute value of the rate of change of the system frequency deviation; Indicates the system's rated frequency; Indicates the SOC attenuation sensitivity coefficient; This represents the state of charge of the i-th energy storage unit at the current moment.
3. The black-start control method for grid-type energy storage based on multi-level adaptive inertia coupling according to claim 1, characterized in that, The system-level equivalent inertia is calculated using the following formula: in, This represents the adaptive virtual inertia of each energy storage unit at the current moment. This indicates the total number of grid-connected energy storage units participating in black start; The system frequency dynamic response capability index is calculated using the following formula: in, Indicates the system's rated angular velocity; This indicates the maximum allowed frequency deviation for a black-start isolated network; Indicates the system's rated frequency; This represents the typical settling time of the frequency transient response under VSG control; This indicates the load power engaged in the current black start step; This represents the adaptive virtual inertia of each energy storage unit at the current moment.
4. The black-start control method for grid-type energy storage based on multi-level adaptive inertia coupling according to claim 1, characterized in that, The dynamic power allocation weight of each energy storage unit is calculated using the following formula: in, This represents the adaptive virtual inertia of the i-th energy storage unit at the current moment; This represents the state of charge of the i-th energy storage unit at time t; The actual power command of each energy storage unit is calculated using the following formula: in, This represents the dynamic power allocation weight of the i-th energy storage unit; This indicates the total load power planned for the current black start step; This represents the system's frequency dynamic response capability index.
5. The black-start control method for grid-type energy storage based on multi-level adaptive inertia coupling according to claim 1, characterized in that, The comprehensive reliability index of the entire black start path, calculated based on dynamic power allocation weights and the available power of each energy storage unit, includes: By combining the dynamic power allocation weight of each energy storage unit with the available power of each energy storage unit, the single-step reliability of each step in the black start path is calculated. A series reliability model is introduced to calculate the overall reliability index of the entire black start path; The overall reliability index of the entire black start path is calculated using the following formula: in, This represents the single-step reliability of the m-th step in the black-start path; The reliability of the i-th energy storage unit in the m-th step is represented by the following formula: in, This indicates the total number of steps in the black boot path; Indicates the redundancy correction index; It indicates that the i-th energy storage unit is in the i-th... The reliability of each step; Indicates the first Taiwan's energy storage in the first The state of charge at each step, Indicates the first The planned load power for each black start step.
6. The black-start control method for grid-type energy storage based on multi-level adaptive inertia coupling according to claim 5, characterized in that, The selection of whether to execute the actual power command based on comprehensive reliability indicators and preset safety thresholds includes: When the comprehensive reliability index is greater than or equal to the first preset threshold, the current scheme is determined to be reliable, and the actual power command is executed. When the comprehensive reliability index is greater than or equal to the second preset threshold and less than the first preset threshold, it is determined that the reliability is insufficient and needs to be adjusted. The current scheme is not executed, the parameters are adjusted and the previous steps are re-executed to obtain the updated actual power command. When the overall reliability index is less than the second preset threshold, the reliability is deemed to be seriously insufficient. The current black boot scheme is not executed, and the black boot strategy is adjusted before the previous steps are re-executed.
7. A grid-type energy storage black-start control system based on multi-level adaptive inertia coupling, characterized in that, include: The adaptive virtual inertia calculation module is used to obtain the real-time state of charge and system frequency deviation change rate of each energy storage unit, and calculate the adaptive virtual inertia of each energy storage unit at the current moment based on the real-time state of charge and system frequency deviation change rate. The system frequency dynamic response capability index calculation module is used to sum and aggregate the adaptive virtual inertia of all energy storage units to obtain the system-level equivalent inertia. Combined with the load power planned to be put into the current black start step and the preset frequency allowable deviation, the system frequency dynamic response capability index is calculated. The actual power command calculation module is used to calculate the dynamic power allocation weight of each energy storage unit based on the adaptive virtual inertia of each energy storage unit at the current moment and the system frequency dynamic response capability index, and generate the actual power command of each energy storage unit according to the dynamic power allocation weight, the expected load power and the system frequency dynamic response capability index. The comprehensive reliability quantification module is used to calculate the comprehensive reliability index of the entire black start path based on the dynamic power allocation weight and the available power of each energy storage unit. Based on the comprehensive reliability index and the preset safety threshold, it selects whether to execute the actual power command.
8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-6.
9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-6.
10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1-6.
