A multi-time scale based multi-element energy storage frequency emergency support method and system

Abstract

The application discloses a multi-time-scale-based multi-element energy storage frequency emergency support method and system, relates to the technical field of multi-element energy storage frequency emergency support, and realizes the optimal balance of response speed and support time length through multi-time-scale coordinated control, wherein the ultra-fast energy storage device is responsible for millisecond-level transient impact, the medium-speed energy storage device is responsible for second-level frequency fluctuation, and the slow-speed energy storage device is responsible for minute-level frequency recovery; the energy storage device resource configuration is optimized, the overuse of high-performance devices and the idling of low-performance devices are avoided, and the overall utilization efficiency of energy storage is improved; the self-adaptive control mechanism enables the system to dynamically adjust the control strategy according to disturbance characteristics, maintains effective frequency support capability under extreme disturbance, and improves the transient stability of energy storage and power supply reliability; through optimized distribution and closed-loop adjustment, the operation loss and maintenance cost of the energy storage device are minimized under the premise of ensuring support effect, and the economy of frequency support is improved.

Description

Technical Field

[0001] This application relates to the field of emergency frequency support technology for multi-energy storage, specifically to a method and system for emergency frequency support of multi-energy storage based on multiple time scales. Background Technology

[0002] Frequency stability in power systems is crucial for ensuring the safe operation of the power grid. With the large-scale integration of new energy sources into the grid, system inertia is decreasing, and frequency disturbances are becoming increasingly prominent. Energy storage technology, due to its rapid response characteristics, has become an important means of emergency frequency support. In existing technologies, energy storage participation in frequency regulation often adopts a control strategy with a uniform time scale or simply superimposes multiple energy storage devices, lacking refined coordination for dynamic characteristics at different time scales. When dealing with complex frequency disturbances, traditional methods often struggle to simultaneously meet the multiple requirements of millisecond-level instantaneous impact suppression, second-level fluctuation smoothing, and minute-level recovery support, leading to a contradiction between rapid response and sustained support.

[0003] Regarding the above-mentioned solutions, the inventors of this application have discovered that the above-mentioned technology has at least the following technical problems:

[0004] 1. Current frequency support tasks typically use fixed time scales, lacking the ability to adaptively adjust to the dynamic characteristics of frequency disturbances; the time constants and power requirements of each time scale cannot be adjusted in real time according to changes in disturbance characteristics, resulting in static and rigid control strategies that are difficult to achieve dynamic optimization.

[0005] 2. Current power allocation often fails to achieve global optimization, lacking comprehensive consideration of support costs, equipment status, and system constraints. Its multi-objective optimization capability is insufficient, resulting in either poor support performance or excessively high operating costs. It fails to effectively balance support performance with equipment lifespan, and the optimization objectives lack a state of charge balance term, which can easily lead to overcharging and over-discharging of energy storage equipment, accelerating equipment aging while providing frequency support. The allocation scheme lacks sufficient power balance constraints, equipment power limit constraints, and state of charge constraints, which may lead to infeasibility or safety hazards in actual implementation, such as equipment overload. Summary of the Invention

[0006] To address the aforementioned technical shortcomings, the purpose of this application is to provide a method and system for emergency frequency support of multi-timescale multi-electrode energy storage.

[0007] To solve the above-mentioned technical problems, this application adopts the following technical solution: In its first aspect, this application provides a multi-timescale multi-element energy storage frequency emergency support method, which includes the following steps: S1. Collect real-time frequency signals of power, extract frequency features and generate frequency feature vectors.

[0008] S2. Based on the frequency feature vector, analyze and derive the sub-tasks at each time scale, and then obtain the time constant and power requirement at each time scale.

[0009] S3. Based on the real-time status parameters of each energy storage device and the power demand at each time scale, calculate the matching degree matrix between each energy storage device and the time scale.

[0010] S4. Based on the matching degree matrix and the power demand at each time scale, the output power of each energy storage device at each time scale is analyzed and obtained.

[0011] S5. Based on the optimized allocation results, generate control command sets for each energy storage device and send them to the corresponding energy storage controller to perform frequency support operations.

[0012] S6. Evaluate the frequency support effect in real time and make adaptive adjustments based on the evaluation results.

[0013] Preferably, the step of acquiring the real-time frequency signal of the power, extracting frequency features, and generating a frequency feature vector includes: S201, acquiring the real-time frequency signal of the power, and calculating the frequency deviation and frequency change rate, wherein: through the calculation formula Resulting in frequency deviation In the formula Represented as a real-time frequency signal of electricity. Expressed as the rated frequency of the power supply; calculated using the formula derive the rate of change of frequency In the formula This represents the sampling time interval.

[0014] S202, Based on the frequency deviation and frequency change rate, extract frequency features, including amplitude features, change rate features and duration features.

[0015] S203 combines the feature parameters to generate a frequency feature vector.

[0016] Preferably, the step of analyzing and obtaining sub-tasks at each time scale based on the frequency feature vector, and then obtaining the time constant and power requirement at each time scale, includes: S301, calculating the time constant of the ultra-short-term time scale based on the frequency change rate in the frequency feature vector, wherein the ultra-short-term time scale corresponds to a millisecond-level rapid frequency impact support task.

[0017] S302, based on the frequency deviation in the frequency feature vector, calculate the time constant of the short-term time scale, which corresponds to the second-level frequency fluctuation suppression task.

[0018] S303, based on the duration of the disturbance in the frequency feature vector, calculate the time constant of the intermediate time scale, which corresponds to the minute-level frequency recovery support task.

[0019] S304, based on the calculated time constants of the ultra-short-term time scale, the short-term time scale, the medium-term time scale, and the frequency characteristic vector, determines the required support power for each time scale, and then derives the power demand vector.

[0020] Preferably, determining the required support power for each time scale, and thus deriving the power demand vector, includes: calculating a set of formulas. The support power value on the ultra-short time scale was calculated. Support power value on short-term time scale Support power values ​​on a medium-term timescale ,in It is represented as a time constant for ultra-short time scales. It is represented as a time constant on a short-term time scale. This is represented as a time constant on a medium-term timescale. , and These are represented as power coefficients for the ultra-short-term, short-term, and medium-term time scales, respectively; the power demand vector is represented as... .

[0021] Preferably, the step of calculating the matching degree matrix between each energy storage device and the time scale based on the real-time status parameters of each energy storage device and the power demand at each time scale includes: S501, obtaining the real-time operating status parameters of each energy storage device, wherein the real-time operating status parameters include the state of charge and response time of each energy storage device.

[0022] S502, Calculate the time characteristic matching degree component based on the time constant of each time scale and the response time of each energy storage device.

[0023] S503, Calculate the state characteristic matching degree component based on the deviation between the state of charge and the optimal state of charge of each energy storage device.

[0024] S504 constructs a matching degree matrix between energy storage devices and time scales based on time characteristic matching degree components and state characteristic matching degree components.

[0025] Preferably, constructing the matching degree matrix between the energy storage device and the time scale based on the time characteristic matching degree component and the state characteristic matching degree component includes: calculating the matching degree matrix using the formula... The elements of the matching degree matrix are obtained. Elements of the matching degree matrix Represented as the first Energy storage devices in the first Overall matching degree across time scales; Indicates the first i Energy storage devices in the first j A time-specific matching degree component at each time scale; Indicates the first i Energy storage devices in the first j The state characteristic matching degree component at each time scale.

[0026] Preferably, the step of analyzing and determining the output power of each energy storage device at each time scale based on the matching degree matrix and the power demand at each time scale includes: S701, selecting a preset number of energy storage devices with the highest matching degree for each time scale based on the matching degree matrix, forming a set of candidate energy storage devices for each time scale.

[0027] S702, establish an optimization model with the goal of minimizing total support cost. The objective function of the optimization model includes the energy storage equipment output cost term and the state of charge balance term.

[0028] S703 sets power balance constraints, upper and lower limits of energy storage device power constraints, and state of charge constraints, forming a complete set of constraints.

[0029] S704 solves the optimization model to obtain the power output allocation scheme of each energy storage device at each time scale, and then calculates the total power output and charging / discharging mode of each energy storage device.

[0030] Preferably, the step of generating control command sets for each energy storage device based on the optimized allocation results and sending them to the corresponding energy storage controllers to perform frequency support operations includes: generating original control commands containing target power values, control durations, and charging / discharging modes based on the output power allocation schemes of each energy storage device at each time scale, converting them into standardized control commands conforming to communication protocol standards, and sending them to the corresponding energy storage device controllers respectively; each energy storage device controller receives and parses the standardized control commands, converts them into executable control signals, and then drives the power converter to perform corresponding charging / discharging operations to realize the frequency support function.

[0031] Preferably, the real-time evaluation of frequency support effect and adaptive adjustment based on the evaluation results include: S901, real-time monitoring of key performance indicators of the frequency support process, and calculation of frequency recovery time, maximum frequency deviation and comprehensive utilization rate of energy storage equipment.

[0032] S902, based on the key performance indicators, calculate the comprehensive effect evaluation score of the current control cycle using the effect evaluation function.

[0033] S903, compare the comprehensive effect evaluation score with the preset target value to determine the adjustment direction and amount of each parameter.

[0034] S904 applies the adjusted parameters to the time-scale decomposition and optimization allocation process of the next control cycle.

[0035] In a second aspect, this application provides a system for a multi-timescale, multi-element energy storage frequency emergency support method, comprising: preferably, a frequency feature extraction module, used to acquire real-time frequency signals of electricity, extract frequency features, and generate a frequency feature vector.

[0036] The multi-timescale task decomposition module analyzes and derives sub-tasks for each time scale based on the frequency feature vector, thereby obtaining the time constant and power requirement for each time scale.

[0037] The energy storage matching degree calculation module calculates the matching degree matrix between each energy storage device and the time scale based on the real-time status parameters of each energy storage device and the power demand at each time scale.

[0038] The power optimization and allocation module analyzes and derives the output power of each energy storage device at each time scale based on the matching degree matrix and the power demand at each time scale.

[0039] The control command generation and distribution module generates control command sets for each energy storage device based on the optimized allocation results, and sends them to the corresponding energy storage controllers to perform frequency support operations.

[0040] The effect evaluation and adaptive adjustment module evaluates the frequency support effect in real time and makes adaptive adjustments based on the evaluation results.

[0041] The beneficial effects of this application are as follows: 1. This application provides a method and system for emergency frequency support based on multi-timescale multi-element energy storage. Through multi-timescale coordinated control, ultra-fast energy storage devices are responsible for millisecond-level instantaneous impacts, medium-speed energy storage devices are responsible for second-level frequency fluctuations, and slow-speed energy storage devices are responsible for minute-level frequency recovery, achieving an optimal balance between response speed and support duration; optimizing the resource allocation of energy storage devices avoids the overuse of high-performance devices and the idleness of low-performance devices, improving the overall utilization efficiency of energy storage; the adaptive control mechanism enables the system to dynamically adjust the control strategy according to the disturbance characteristics, maintaining effective frequency support capability under extreme disturbances, improving the transient stability of energy storage and power supply reliability; through optimized allocation and closed-loop adjustment, the operating losses and maintenance costs of energy storage devices are minimized while ensuring the support effect, improving the economic efficiency of frequency support.

[0042] 2. This application discretizes the continuous frequency support task into three time scales with distinct meanings and different time constants. This decomposition method is based on the dynamic characteristics of frequency disturbances for adaptive adjustment, rather than using a fixed time scale division. When the frequency disturbance characteristics change, the time constants and power requirements of each time scale can be adjusted in real time, realizing the dynamic optimization of the control strategy.

[0043] 3. This application achieves globally optimal power allocation by comprehensively considering support costs, equipment status, and system constraints through a two-layer optimization model. It achieves optimal power allocation under multiple objectives, minimizing operating costs while ensuring support effectiveness. It balances support effectiveness with equipment lifespan by including a state-of-charge balance term in the optimization objective function, avoiding overcharging and over-discharging of energy storage devices and maintaining the health of the equipment while providing frequency support. It ensures the feasibility and safety of allocation by using power balance constraints, equipment power limit constraints, and state-of-charge constraints to guarantee the feasibility and safety of the allocation scheme in actual implementation, avoiding safety hazards such as equipment overload. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is a flowchart illustrating the implementation steps of the method described in this application.

[0046] Figure 2 This is a schematic diagram of the system structure connection of this application. Detailed Implementation

[0047] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0048] Please see Figure 1 As shown, this application provides a multi-timescale multi-electrode energy storage frequency emergency support method in the first aspect, including: S1, collecting real-time frequency signals of power, extracting frequency features and generating frequency feature vectors.

[0049] In a specific embodiment, the step of acquiring the real-time frequency signal of the power supply, extracting frequency features, and generating a frequency feature vector includes: S201, acquiring the real-time frequency signal of the power supply, and calculating the frequency deviation and frequency change rate, wherein: through the calculation formula... Resulting in frequency deviation In the formula Represented as a real-time frequency signal of electricity. Expressed as the rated frequency of the power supply; calculated using the formula derive the rate of change of frequency In the formula This represents the sampling time interval.

[0050] S202, Based on the frequency deviation and frequency change rate, extract frequency features, including amplitude features, change rate features and duration features.

[0051] S203 combines the feature parameters to generate a frequency feature vector.

[0052] It should be noted that the process of acquiring real-time frequency signals of the power system is achieved through frequency measurement devices deployed at key nodes of the power system. These frequency measurement devices collect frequency data at a fixed sampling frequency, with the sampling time interval set to 10-100 milliseconds based on the dynamic characteristics of the power system. Furthermore, when calculating the frequency deviation, the rated frequency of the power system is typically 50Hz or 60Hz, with the specific value determined according to the standards of the power grid. The frequency deviation reflects the degree of deviation between the current frequency and the rated frequency; a positive deviation indicates that the frequency is higher than the rated value, and a negative deviation indicates that the frequency is lower than the rated value.

[0053] It should be noted that the frequency change rate is calculated using the first-order difference method, obtained by dividing the difference between the real-time frequency signal at the current sampling moment and the real-time frequency signal at the previous sampling moment by the sampling time interval. The frequency change rate characterizes the speed of frequency change; a larger absolute value indicates a more rapid frequency change. Further, the extracted frequency features include: amplitude features denoted as... And take it as the maximum absolute value of the frequency deviation, that is The rate of change characteristic is denoted as And take it as the maximum absolute value of the rate of change of frequency, that is The duration feature is denoted as Furthermore, the duration characteristic is the continuous duration during which the frequency deviation exceeds the threshold.

[0054] It should be noted that the process of combining the feature parameters to generate a frequency feature vector involves arranging the three feature parameters in a specific order to form a three-dimensional vector, i.e. Furthermore, the generation of the frequency feature vector also considers the normalization of feature parameters, dividing each feature parameter by its corresponding reference value to eliminate the influence of dimensions and facilitate subsequent algorithm processing. The normalized frequency feature vector is: in It is represented as the frequency deviation reference value, which is the maximum value of the allowable frequency deviation range of the system; The reference value for the rate of change of frequency is determined based on the system's inertial time constant; This serves as the time reference value, typically taken as the duration of a typical disturbance.

[0055] S2. Based on the frequency feature vector, analyze and derive the sub-tasks at each time scale, and then obtain the time constant and power requirement at each time scale.

[0056] It should be noted that the various time scales include ultra-short-term time scales, short-term time scales, and medium-term time scales.

[0057] In a specific embodiment, the step of analyzing and deriving sub-tasks at each time scale based on the frequency feature vector, and then obtaining the time constant and power requirement at each time scale, includes: S301, calculating the time constant of the ultra-short-term time scale based on the frequency change rate in the frequency feature vector, wherein the ultra-short-term time scale corresponds to a millisecond-level rapid frequency impact support task.

[0058] S302, based on the frequency deviation in the frequency feature vector, calculate the time constant of the short-term time scale, which corresponds to the second-level frequency fluctuation suppression task.

[0059] S303, based on the duration of the disturbance in the frequency feature vector, calculate the time constant of the intermediate time scale, which corresponds to the minute-level frequency recovery support task.

[0060] S304, based on the calculated time constants of the ultra-short-term time scale, the short-term time scale, the medium-term time scale, and the frequency characteristic vector, determines the required support power for each time scale, and then derives the power demand vector.

[0061] It should be noted that the calculation of the time constant for the ultra-short-term time scale is based on the frequency change rate component. The ultra-short-term time scale primarily addresses instantaneous frequency surges in power systems caused by generator tripping, heavy load switching, etc., and its time constant typically ranges from 10 to 500 milliseconds. The specific calculation formula is as follows: ,in It is represented as a time constant for ultra-short time scales. This is expressed as the weighting coefficient of the rate of change of frequency on the ultra-short-term time scale. The reference time constant, denoted as the ultra-short-term time scale, and the frequency change rate weighting coefficient of the ultra-short-term time scale, along with the reference time constant, are adjustable parameters that can be adaptively adjusted through step S6. Furthermore, the absolute value of the frequency change rate reflects the drastic degree of frequency change. When the frequency change rate is large, it indicates that the system is experiencing a severe frequency shock, requiring a shorter time scale for rapid response; therefore, the value of the time constant for the ultra-short-term time scale will decrease accordingly. Conversely, when the frequency change rate is small, the value of the time constant for the ultra-short-term time scale will be appropriately increased to avoid unnecessary rapid response.

[0062] It should be noted that the calculation of the time constant for the short-term time scale is based on the frequency deviation component. The short-term time scale primarily addresses frequency fluctuations in the power system caused by load fluctuations, changes in renewable energy output, etc., and its time constant typically ranges from 1 to 30 seconds. The specific calculation formula is as follows: ,in It is represented as a time constant on a short-term time scale. This is expressed as the frequency deviation weighting coefficient on a short-term time scale. The reference time constant is represented as a short-term time scale, and both the short-term time scale frequency deviation weighting coefficient and the reference time constant are adjustable parameters. Furthermore, the absolute value of the frequency deviation reflects the degree to which the frequency deviates from its rated value. When the frequency deviation is large, it indicates that the system frequency has seriously deviated from the normal range, requiring stronger support; therefore, the value of the short-term time scale time constant will be adjusted accordingly. When the frequency deviation is small, the value of the short-term time scale time constant will be appropriately relaxed to avoid over-control.

[0063] It should be noted that the calculation of the time constant for the intermediate time scale is based on the duration component of the disturbance. The intermediate time scale primarily addresses frequency deviations in the power system that require a relatively long recovery time, and its time constant typically ranges from 1 to 10 minutes. The specific calculation formula is as follows: ,in Time constants expressed as medium-term time scales This is represented by the duration weighting coefficient on a medium-term time scale. This is expressed as a reference time constant for a medium-term time scale. Furthermore, the duration of the disturbance reflects the persistence of the frequency event. A longer duration indicates that the frequency event has persistent characteristics, requiring support strategies for a longer time scale; a shorter duration suggests a possible transient disturbance, with relatively less need for support strategies for a medium-term time scale.

[0064] Furthermore, time-scale decomposition parameters and The initial values ​​are set based on the typical operating conditions of the power system and the characteristics of the energy storage equipment, and are adjusted online in step S6 according to the actual support effect. This parameter adaptive mechanism ensures that the time-scale decomposition algorithm can adapt to different operating scenarios and disturbance types.

[0065] In one specific embodiment, determining the required support power at each time scale, and thus deriving the power demand vector, includes: calculating a set of formulas. The support power value on the ultra-short time scale was calculated. Support power value on short-term time scale Support power values ​​on a medium-term timescale ,in It is represented as a time constant for ultra-short time scales. It is represented as a time constant on a short-term time scale. This is represented as a time constant on a medium-term timescale. , and These are represented as power coefficients for the ultra-short-term, short-term, and medium-term time scales, respectively; the power demand vector is represented as... .

[0066] It should be noted that the power coefficients of the ultra-short-term time scale, the short-term time scale, and the medium-term time scale are related to system parameters such as the system inertia constant and damping coefficient. The integral of the frequency deviation reflects the accumulated frequency deviation energy. The power demand vector contains the required supporting power values ​​at each time scale, providing a quantitative basis for subsequent power allocation to energy storage devices.

[0067] This application discretizes the continuous frequency support task into three time scales with distinct meanings and different time constants. This decomposition method is based on the dynamic characteristics of frequency disturbances for adaptive adjustment, rather than using a fixed time scale division. When the characteristics of frequency disturbances change, the time constants and power requirements of each time scale can be adjusted in real time, realizing dynamic optimization of the control strategy.

[0068] S3. Based on the real-time status parameters of each energy storage device and the power demand at each time scale, calculate the matching degree matrix between each energy storage device and the time scale.

[0069] In a specific embodiment, the step of calculating the matching degree matrix between each energy storage device and the time scale based on the acquired real-time status parameters of each energy storage device and the power demand at each time scale includes: S501, acquiring the real-time operating status parameters of each energy storage device, the real-time operating status parameters including the state of charge and response time of each energy storage device.

[0070] S502, Calculate the time characteristic matching degree component based on the time constant of each time scale and the response time of each energy storage device.

[0071] S503, Calculate the state characteristic matching degree component based on the deviation between the state of charge and the optimal state of charge of each energy storage device.

[0072] S504 constructs a matching degree matrix between energy storage devices and time scales based on time characteristic matching degree components and state characteristic matching degree components.

[0073] It should be noted that the process of acquiring real-time operating status parameters of each energy storage device is achieved through an energy storage monitoring system. These devices include novel energy storage, flywheel energy storage, and battery energy storage. The response time of novel energy storage characterizes the time required from receiving control commands to outputting the target power, typically ranging from 1 to 10 milliseconds. The electric heating stage can achieve a rapid electrical response at the millisecond level, while the thermal energy output system is affected by thermal inertia, typically taking the order of minutes. Its core advantage lies not in competing with lithium batteries in terms of short-time response speed, but in its large-scale, long-cycle energy storage capacity and its ability to provide stable and flexible regulation in new power systems. The response time of flywheel energy storage typically ranges from 10 to 100 milliseconds; the response time of battery energy storage typically ranges from 100 to 1000 milliseconds. State of charge (SOC) indicates the current energy storage level of the energy storage device, ranging from 0 to 100%. Furthermore, the aforementioned novel energy storage, flywheel energy storage, and battery energy storage constitute the core components of the energy storage system. New energy storage technologies have the fastest response speed and the highest power density, but lower energy density, making them suitable for tasks requiring instantaneous high power support. Flywheel energy storage has a relatively fast response speed and medium energy density, making it suitable for tasks requiring short-term frequency fluctuation suppression. Battery energy storage has a high energy density but a relatively slow response speed, making it suitable for tasks requiring longer-term frequency recovery support.

[0074] It should be noted that the calculation of the time characteristic matching degree component is based on the degree of matching between the time constants of each time scale and the response time of each energy storage device. This is achieved through the calculation formula... The result is the Energy storage devices in the first Time characteristic matching degree component at each time scale ,in Represented as the first Response time of energy storage devices Represented as the first The time constant of each time scale This represents the weighting coefficient of the time characteristic matching degree component. The time characteristic matching degree component reflects the degree of matching between the response characteristics of the energy storage device and the dynamic demand over a time scale. The closer the reciprocal of the energy storage device's response duration is to the reciprocal of the time constant of the time scale, the higher the matching degree. Furthermore, each time scale includes ultra-short-term, short-term, and medium-term time scales. The time constant of the time scale... Including time constants of ultra-short time scales The time constant of the short-term time scale , is the time constant of the medium-term time scale. The response time of the energy storage device. This includes the response time of new energy storage, the response time of flywheel energy storage, and the response time of battery energy storage.

[0075] It should be noted that the calculation of the state characteristic matching degree component is based on the deviation between the current state of charge and the optimal state of charge of each energy storage device. This is achieved through the calculation formula. The result is the Energy storage devices in the first Each time-scale state characteristic matching degree component; where Represented as the first The current state of charge of energy storage devices Indicates the first The optimal state of charge for each time scale. This represents the weighting coefficient of the state characteristic matching degree component. The state characteristic matching degree component reflects the degree of matching between the current energy storage level of the energy storage device and the task requirements of the time scale. Furthermore, the optimal state of charge (SOC) is set according to the task characteristics of each time scale. For ultra-short-term time scales, the optimal SOC is usually set in the range of 50%–70% to ensure that the new energy storage has sufficient charging and discharging capacity to cope with instantaneous shocks; for short-term time scales, the optimal SOC is usually set in the range of 60%–80%; for medium-term time scales, the optimal SOC is usually set in the range of 40%–60% to ensure that the battery energy storage has a longer continuous support capability.

[0076] It should be noted that the weighting coefficients of the time characteristic matching degree component... Weight coefficients corresponding to the matching degree components of state characteristics The setting reflects the relative importance of time characteristics and state characteristics. Typically, A larger value emphasizes the importance of matching response time, because frequency support tasks have high requirements for time characteristics; The value is relatively small, but it cannot be ignored, to ensure that the energy storage device operates under a suitable state of charge; at the same time ,and and The S6 step allows for adaptive adjustments based on actual operational results.

[0077] In a specific embodiment, constructing the matching degree matrix between the energy storage device and the time scale based on the time characteristic matching degree component and the state characteristic matching degree component includes: calculating the matching degree matrix using the formula... The elements of the matching degree matrix are obtained. Elements of the matching degree matrix Represented as the first Energy storage devices in the first Overall matching degree across time scales; Indicates the first i Energy storage devices in the first j A time-specific matching degree component at each time scale; Indicates the first i Energy storage devices in the first j The state characteristic matching degree component at each time scale.

[0078] It should be noted that the matching degree matrix is ​​a 3×3 matrix, with rows corresponding to the three types of energy storage devices and columns corresponding to the three time scales. Each element in the matrix... Indicates the first Energy storage devices in the first The overall matching degree across time scales is calculated, with smaller matching values ​​indicating a higher degree of matching. This matching degree matrix provides a basis for equipment selection in subsequent power optimization and allocation, and energy storage-time scale combinations with high matching degrees will be given priority in the optimization.

[0079] S4. Based on the matching degree matrix and the power demand at each time scale, the output power of each energy storage device at each time scale is analyzed and obtained.

[0080] In a specific embodiment, the step of analyzing the output power of each energy storage device at each time scale based on the matching degree matrix and the power demand at each time scale includes: S701, selecting a preset number of energy storage devices with the highest matching degree for each time scale based on the matching degree matrix, forming a set of candidate energy storage devices for each time scale.

[0081] S702, establish an optimization model with the goal of minimizing total support cost. The objective function of the optimization model includes the energy storage equipment output cost term and the state of charge balance term.

[0082] S703 sets power balance constraints, upper and lower limits of energy storage device power constraints, and state of charge constraints, forming a complete set of constraints.

[0083] S704 solves the optimization model to obtain the power output allocation scheme of each energy storage device at each time scale, and then calculates the total power output and charging / discharging mode of each energy storage device.

[0084] It should be noted that this involves selecting a preset number of energy storage devices with the highest matching degree for each time scale. For very short time scales, 1-2 energy storage devices with the highest matching degree are selected, typically with newer energy storage technologies showing the highest matching degree. For short time scales, 2 energy storage devices with the highest matching degree are selected, typically with flywheel energy storage and newer energy storage technologies showing a high matching degree. For medium-term time scales, 1-2 energy storage devices with the highest matching degree are selected, typically with battery energy storage showing the highest matching degree. The preset number is dynamically adjusted based on the actual energy storage system configuration and mission requirements.

[0085] Furthermore, the construction rules for the candidate energy storage device set are as follows: for the first... The time scale, from the matching degree matrix at the time scale, the _th time scale, from the _th time scale, Select matrix elements in the column The first with the smallest matching score One energy storage device, of which Represented as the first The number of candidate devices for each time scale. This selection mechanism ensures that each time scale is supported by the energy storage device most suitable for undertaking the task at that scale.

[0086] It should be noted that the process of establishing the optimization model is as follows: The optimization model aims to minimize the total support cost, and the objective function contains two main parts: the energy storage equipment output cost term and the state of charge balance term. The specific form of the objective function is: ,in Represented as the first Energy storage devices in the first Power output over a timescale Represented as the first Energy storage devices in the first Output cost coefficient at each time scale The weighting coefficients representing the state of charge balance Represented as the first The current state of charge of energy storage devices This is represented as the reference state of charge. Furthermore, the output cost coefficient is determined based on the matching degree matrix elements. The higher the matching degree of the energy storage device-time scale combination, the lower the output cost coefficient, which makes the optimization process tend to allocate power to energy storage devices with high matching degree.

[0087] It should be noted that the state-of-charge balance term Its function is to maintain the operation of each energy storage device within a reasonable state of charge range, preventing some devices from over-discharging or over-charging. The reference state of charge is usually set at 50% to ensure that the energy storage devices have a balanced charging and discharging capability.

[0088] It should be noted that the process of setting constraints includes three main categories of constraints. The first category is power balance constraints, which require that the total output power at each time scale equals the power demand at that time scale: in Indicates the first A collection of candidate energy storage devices across multiple time scales. Indicates the first Power demand at each time scale. Furthermore, the second type of constraint is the upper and lower limit constraint on the power of energy storage devices, ensuring that the output of each energy storage device is within its technically permissible range: in and They represent the first Minimum and maximum permissible output power of energy storage devices.

[0089] It should be noted that the third type of constraint is the state of charge constraint, which ensures that the state of charge of each energy storage device is within the safe operating range after optimization. ,in and They represent the first Minimum state of charge and maximum permissible state of charge for energy storage devices Indicates the control time interval. Indicates the first Rated capacity of energy storage devices.

[0090] It should be noted that, in solving the optimization model, since the objective function is quadratic and the constraints are linear, this optimization problem is a convex optimization problem, and a global optimum exists. The optimal solution obtained represents the optimal power output allocation of each energy storage device at different time scales.

[0091] It should be noted that this involves calculating the total output power and charging / discharging mode of each energy storage device. The output power is based on the optimal output power allocation scheme. Calculate the first The total output power of the energy storage device is: The charging / discharging mode is determined based on the sign of the total output power: when When, it indicates the discharge mode; when When, it indicates the charging mode; when When, it indicates standby mode.

[0092] Furthermore, the control duration of each energy storage device is determined based on the time constant of the corresponding time scale. For energy storage devices that simultaneously undertake tasks at multiple time scales, their control duration is the weighted sum of the control durations at each scale.

[0093] This application achieves globally optimal power allocation by comprehensively considering support costs, equipment status, and system constraints through a two-layer optimization model. It achieves optimal power allocation under multiple objectives, minimizing operating costs while ensuring support effectiveness. It balances support effectiveness with equipment lifespan by including a state-of-charge (POC) balance term in the optimization objective function, avoiding overcharging and over-discharging of energy storage devices and maintaining the health of the equipment while providing frequency support. It ensures the feasibility and safety of allocation by using power balance constraints, equipment power limit constraints, and POC constraints to guarantee the feasibility and safety of the allocation scheme in actual implementation, avoiding safety hazards such as equipment overload.

[0094] S5. Based on the optimized allocation results, generate control command sets for each energy storage device and send them to the corresponding energy storage controller to perform frequency support operations.

[0095] In one specific embodiment, the step of generating control command sets for each energy storage device based on the optimized allocation results and sending them to the corresponding energy storage controllers to perform frequency support operations includes: generating raw control commands containing target power values, control durations, and charging / discharging modes based on the output power allocation schemes of each energy storage device at each time scale, converting them into standardized control commands conforming to communication protocol standards, and sending them to the corresponding energy storage device controllers; each energy storage device controller receives and parses the standardized control commands, converts them into executable control signals, and then drives the power converter to perform corresponding charging / discharging operations to achieve the frequency support function.

[0096] It should be noted that the process of generating the initial control commands is based on the optimal power output allocation scheme. The specific calculation formula is as follows: ,in Indicates the first Energy storage devices in the first The output power of the optimal power allocation scheme at each time scale. Indicates the first The time constant of each time scale Represented as the first Target power values ​​for energy storage devices Represented as the first The control duration of energy storage devices. Furthermore, the first... The charging and discharging modes of energy storage devices are denoted as: The sign of the target power value is used to determine the power level. hour, ;when hour, ;when hour, For energy storage devices that simultaneously undertake tasks at various time scales, the control duration... It is the power-weighted average of the control duration at each time scale.

[0097] It should be noted that converting the control commands to standardized control commands conforming to communication protocol standards involves formatting and encoding the original control commands. The formatting process converts the original control commands into command frames with a standard structure, including: frame header, device address, command type, target power value, control duration, charging / discharging mode, checksum, and frame trailer. The encoding process uses binary encoding to convert each parameter into a binary data stream. Furthermore, the communication protocol for the standardized control commands is selected based on the actual configuration of the energy storage system. Commonly used communication protocols include Modbus RTU, IEC 61850, and DNP3.0. For ultra-short-term control commands requiring high real-time performance, high-speed communication protocols and priority transmission mechanisms are used to ensure command transmission delays are less than 10 milliseconds; for short- and medium-term control commands, standard communication protocols are used, with relatively relaxed transmission delay requirements.

[0098] It should be noted that standardized control commands are sent to the corresponding energy storage device controllers via a communication network. This communication network employs a layered architecture, including a station control layer, a bay layer, and a process layer. Standardized control commands generated by the central controller are first sent to the station control layer server, and then distributed to the corresponding process layer controllers of each energy storage device via bay layer switches. For critical, ultra-short-term control commands, a point-to-point direct communication method is used to bypass intermediate switching links and reduce transmission latency.

[0099] Furthermore, the communication network employs a redundant design, including primary and backup communication channels and an automatic switching mechanism. When the primary communication channel fails, the system automatically switches to the backup communication channel to ensure reliable transmission of control commands. Simultaneously, the control command transmission utilizes an acknowledgment-retransmission mechanism: the receiver sends an acknowledgment signal upon successful reception of the command, and the sender retransmits the command if no acknowledgment is received within a timeout period.

[0100] It should be noted that each energy storage device controller receives and parses standardized control commands. The controller continuously listens to the communication port and begins receiving data when a valid command frame header is detected. After reception, the controller performs integrity checks on the command frame, including frame length verification and checksum verification. If the verification passes, the controller parses each field in the command frame, extracting the target power value, control duration, and charging / discharging mode parameters. Furthermore, the parsed control parameters undergo a rationality check, including: whether the target power value is within the device's allowable range, whether the control duration is reasonable, and whether the charging / discharging mode is compatible with the current device state. If abnormal parameters are found, the controller will refuse to execute the command and send an error report to the central controller, requesting the regeneration of the control command.

[0101] It should be noted that the process of driving the power converter to perform charging and discharging operations involves the following steps. Based on the analyzed control parameters, the energy storage device controller generates a PWM (Pulse Width Modulation) control signal or an analog control signal to drive the power converter (such as a DC / AC converter or DC / DC converter) to perform corresponding power regulation. In discharge mode, the controller controls the power converter to convert the DC power from the energy storage device into AC power for injection into the grid; in charging mode, the controller controls the power converter to convert the AC power from the grid into DC power to charge the energy storage device. Furthermore, the power converter control employs a closed-loop control strategy, monitoring the deviation between the actual output power and the target power in real time, and adjusting the control signal through a PID control algorithm to ensure that the output power quickly and accurately tracks the target power value. For fast-response tasks with ultra-short time scales, a feedforward-feedback composite control strategy is adopted to improve response speed and control accuracy.

[0102] It is important to note the safety protection mechanisms during execution. The controller monitors the operating status of the energy storage device and power converter in real time, including parameters such as voltage, current, and temperature. When abnormal conditions such as overvoltage, overcurrent, or overtemperature are detected, the controller immediately activates the protection program, gradually reducing power or stopping operation according to preset safety strategies to ensure equipment safety. Simultaneously, the controller reports the abnormal status information to the central controller in real time, providing a basis for subsequent control strategy adjustments.

[0103] S6. Evaluate the frequency support effect in real time and make adaptive adjustments based on the evaluation results.

[0104] In a specific embodiment, the real-time evaluation of frequency support effect and adaptive adjustment based on the evaluation results include: S901, real-time monitoring of key performance indicators of the frequency support process, and calculation of frequency recovery time, maximum frequency deviation, and comprehensive utilization rate of energy storage equipment.

[0105] S902, based on the key performance indicators, calculate the comprehensive effect evaluation score of the current control cycle using the effect evaluation function.

[0106] S903, compare the comprehensive effect evaluation score with the preset target value to determine the adjustment direction and amount of each parameter.

[0107] S904 applies the adjusted parameters to the time-scale decomposition and optimization allocation process of the next control cycle.

[0108] It should be noted that this refers to the process of real-time monitoring of key performance indicators. The frequency recovery time is the time elapsed from the occurrence of the frequency disturbance until the frequency recovers to within the allowable deviation range (typically ±0.05Hz). The maximum frequency deviation is the maximum absolute value of the frequency deviating from the rated value during the entire frequency support process. The comprehensive utilization rate of the energy storage device is the ratio of the total energy actually used for frequency support to the total energy that the energy storage system can provide, calculated using the following formula: That Represented as the first Energy storage devices at all times The output power, Represented as the first The total energy that energy storage devices can provide under the current state of charge. This is expressed as the control cycle duration. Furthermore, the sampling and calculation of key performance indicators are performed in units of control cycles. After each control cycle, the system automatically calculates the frequency recovery time, maximum frequency deviation, and overall utilization rate of the energy storage equipment within that cycle. The length of the control cycle is set according to the system's dynamic characteristics, typically between 1 and 5 minutes, ensuring a complete assessment of the overall effectiveness of a frequency support process.

[0109] It should be noted that the process of calculating the comprehensive performance evaluation score using the performance evaluation function is as follows. This performance evaluation function comprehensively considers three aspects: frequency recovery speed, frequency deviation suppression effect, and energy storage utilization efficiency, and is calculated using the following formula. Result in an overall performance evaluation score ,in This is expressed as a reference recovery time, and is usually set to 2-3 times the system inertia time constant; Indicates recovery time; The reference frequency deviation is typically set to 1.5 times the allowable deviation of the system frequency. Indicates frequency deviation; For reference purposes, utilization is typically set at 70%–80%. Indicates utilization rate; , and These are the weighting coefficients corresponding to frequency recovery time, frequency deviation, and the comprehensive utilization rate of energy storage equipment, respectively, and they satisfy the following conditions: .

[0110] Furthermore, the weighting coefficients for frequency recovery time, frequency deviation, and comprehensive utilization rate of energy storage devices reflect the relative importance of different performance indicators. In emergency frequency support scenarios, the weighting coefficient for frequency deviation is typically the largest (0.4-0.5), emphasizing the frequency deviation suppression effect; the weighting coefficient for frequency recovery time is (0.3-0.4), emphasizing recovery speed; and the weighting coefficient for comprehensive utilization rate of energy storage devices is relatively small (0.1-0.2), considering energy storage utilization efficiency while ensuring support effectiveness.

[0111] It should be noted that the process of comparing the overall performance evaluation score with the preset target value is as follows. The preset target value is set based on system operating requirements and performance expectations, typically ranging from 0.8 to 1.2. The comparison result produces three states: when the overall performance evaluation score is greater than the preset target value, it indicates that the current control effect is better than expected, and the parameter adjustment direction is fine-tuning optimization; when the overall performance evaluation score is equal to the preset target value, it indicates that the control effect meets expectations, and the parameters remain stable; when the overall performance evaluation score is less than the preset target value, it indicates that the control effect has not met expectations, and the parameters need significant adjustment.

[0112] Furthermore, the parameter adjustment amount is calculated based on the deviation of the effect evaluation score and the current parameter value. The formula for calculating the adjustment amount is: , This is expressed as a parameter adjustment amount. This is represented by the deviation in the performance evaluation score. This is represented as a preset target value. This represents the current parameter value. This is represented as the learning rate coefficient, which controls the adjustment range of the parameter, typically set between 0.05 and 0.2. A larger learning rate is used when the results are significantly lower than expected; a smaller learning rate is used when the results are close to expected.

[0113] It should be noted that the parameters are adaptively adjusted; the adjustment of the frequency change rate weighting coefficient and the reference time constant for each time scale is based on the performance of the frequency recovery time. When the frequency recovery time is too long, the frequency change rate weighting coefficient of the ultra-short time scale is reduced (to improve the response speed of the ultra-short time scale) and the reference time constant of the ultra-short time scale is increased (to prolong the ultra-short support duration); when the maximum value of the frequency deviation is too large, the frequency change rate weighting coefficient of the short-term time scale is increased (to strengthen the power allocation of the short-term time scale) and the reference time constant of the short-term time scale is adjusted (to optimize the short-term support strategy).

[0114] Furthermore, optimize the weighting coefficients of the time characteristic matching degree component. Weight coefficients corresponding to the matching degree components of state characteristics The adjustment is based on the overall utilization rate of energy storage equipment. When the overall utilization rate of energy storage equipment is too low, it indicates that the energy storage equipment is not being fully utilized, and the adjustment is reduced. (Reduce the weight of time characteristic matching) or increase it (Increase the weight of state characteristic matching); when the overall utilization rate of energy storage equipment is too high but the support effect is poor, then increase... (Reduce the weight of time characteristic matching) or decrease .

[0115] It should be noted that the adjusted parameters are immediately applied to the time-scale decomposition algorithm (step S3) and energy storage matching degree calculation (step S4) of the next control cycle. The system continuously monitors the support effect of each control cycle and dynamically adjusts the parameters based on the effect evaluation results, forming a closed-loop optimization control cycle of "execution-evaluation-adjustment-re-execution".

[0116] Furthermore, the closed-loop optimization control has learning and memory functions. The system records parameter settings and corresponding performance evaluation scores in historical control cycles, establishing a parameter-performance mapping database. When encountering similar frequency disturbance scenarios, the system can quickly recall historically optimal parameter combinations, improving control response speed. Simultaneously, the system employs a sliding window mechanism, retaining only data from the most recent certain period to prevent outdated historical data from affecting current adjustment decisions.

[0117] It should be noted that closed-loop optimization control also includes an anomaly handling mechanism. When the performance evaluation score remains below a threshold for multiple consecutive control cycles, the system determines that the current control strategy has failed and automatically switches to standby control mode. Standby control mode employs conservative parameter settings and simplified control logic to ensure basic frequency support functionality is still provided even under the worst-case scenario. Simultaneously, the system issues an alarm to operators, prompting them to request manual intervention or strategy maintenance.

[0118] Please see Figure 2 As shown, in its second aspect, this application provides a system for a multi-timescale, multi-element energy storage frequency emergency support method.

[0119] The system 100 of the multi-timescale multi-element energy storage frequency emergency support method described in this invention can be installed in an electronic device. Depending on the functions implemented, the system 100 may include a frequency feature extraction module 101, a multi-timescale task decomposition module 102, an energy storage matching degree calculation module 103, a power optimization allocation module 104, a control command generation and issuance module 105, and an effect evaluation and adaptive adjustment module 106. The modules described in this invention can also be called units, referring to a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0120] In this embodiment, the functions of each module / unit are as follows:

[0121] The frequency feature extraction module is used to collect real-time frequency signals of power, extract frequency features, and generate frequency feature vectors.

[0122] The multi-timescale task decomposition module analyzes and derives sub-tasks for each time scale based on the frequency feature vector, thereby obtaining the time constant and power requirement for each time scale.

[0123] The energy storage matching degree calculation module calculates the matching degree matrix between each energy storage device and the time scale based on the real-time status parameters of each energy storage device and the power demand at each time scale.

[0124] The power optimization and allocation module analyzes and derives the output power of each energy storage device at each time scale based on the matching degree matrix and the power demand at each time scale.

[0125] The control command generation and distribution module generates control command sets for each energy storage device based on the optimized allocation results, and sends them to the corresponding energy storage controllers to perform frequency support operations.

[0126] The effect evaluation and adaptive adjustment module evaluates the frequency support effect in real time and makes adaptive adjustments based on the evaluation results.

[0127] This application provides a method and system for emergency frequency support based on multi-timescale multi-element energy storage. Through multi-timescale coordinated control, ultra-fast energy storage devices handle millisecond-level instantaneous impacts, medium-speed energy storage devices handle second-level frequency fluctuations, and slow-speed energy storage devices handle minute-level frequency recovery, achieving an optimal balance between response speed and support duration. Optimized energy storage device resource allocation avoids overuse of high-performance devices and idleness of low-performance devices, improving the overall utilization efficiency of energy storage. An adaptive control mechanism enables the system to dynamically adjust control strategies based on disturbance characteristics, maintaining effective frequency support capability under extreme disturbances, thus improving the transient stability and power supply reliability of energy storage. Through optimized allocation and closed-loop adjustment, the system minimizes the operating losses and maintenance costs of energy storage devices while ensuring support effectiveness, improving the economics of frequency support.

[0128] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0129] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0130] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0131] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0132] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0133] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for emergency frequency support of multi-timescale multi-element energy storage, characterized in that, include: S1. Collect real-time frequency signals of power, extract frequency features and generate frequency feature vectors; S2. Based on the frequency feature vector, analyze and derive the sub-tasks at each time scale, and then obtain the time constant and power requirement at each time scale. S3. Based on the real-time status parameters of each energy storage device and the power demand at each time scale, calculate the matching degree matrix between each energy storage device and the time scale. S4. Based on the matching degree matrix and the power demand at each time scale, the output power of each energy storage device at each time scale is analyzed and obtained. S5. Generate a set of control instructions for each energy storage device based on the optimized allocation results, and send it to the corresponding energy storage controller to perform frequency support operations; S6. Real-time evaluation of frequency support effectiveness, and adaptive adjustment based on evaluation results; Based on the acquired real-time status parameters of each energy storage device and the power demand at each time scale, the matching degree matrix between each energy storage device and the time scale is calculated, including: S501, obtain the real-time operating status parameters of each energy storage device, including the state of charge and response time of each energy storage device; S502, Calculate the time characteristic matching degree component based on the time constant of each time scale and the response time of each energy storage device; S503, Calculate the state characteristic matching degree component based on the deviation between the state of charge and the optimal state of charge of each energy storage device; S504, based on the time characteristic matching degree component and the state characteristic matching degree component, constructs the matching degree matrix between the energy storage device and the time scale; The construction of the matching degree matrix between the energy storage device and the time scale based on the time characteristic matching degree component and the state characteristic matching degree component includes: Through calculation formula The elements of the matching degree matrix are obtained. Elements of the matching degree matrix Represented as the first Energy storage devices in the first Overall matching degree across time scales; Indicates the first i Energy storage devices in the first j The time characteristic matching degree component of each time scale; Indicates the first i Energy storage devices in the first j The state characteristic matching degree component at each time scale.

2. The method for emergency frequency support based on multi-timescale multi-element energy storage according to claim 1, characterized in that, The process of acquiring real-time frequency signals of electricity, extracting frequency features, and generating frequency feature vectors includes: S201, acquires the real-time frequency signal of the power supply, and calculates the frequency deviation and frequency change rate, where: through the calculation formula... Resulting in frequency deviation In the formula Represented as a real-time frequency signal of electricity. Expressed as the rated frequency of the power supply; calculated using the formula derive the rate of change of frequency In the formula This is represented as the sampling time interval; S202, Based on the frequency deviation and frequency change rate, extract frequency features, including amplitude features, change rate features and duration features; S203 combines the feature parameters to generate a frequency feature vector.

3. The method for emergency frequency support based on multi-timescale multi-element energy storage according to claim 1, characterized in that, Based on the frequency feature vector, the sub-tasks at each time scale are analyzed and obtained, thereby yielding the time constant and power requirement at each time scale, including: S301, Based on the frequency change rate in the frequency feature vector, the time constant of the ultra-short time scale is calculated, and the ultra-short time scale corresponds to the millisecond-level rapid frequency impact support task. S302, Based on the frequency deviation in the frequency feature vector, the time constant of the short-term time scale is calculated, and the short-term time scale corresponds to the second-level frequency fluctuation suppression task; S303, calculate the time constant of the intermediate time scale based on the duration of the disturbance in the frequency feature vector, wherein the intermediate time scale corresponds to the minute-level frequency recovery support task; S304, based on the calculated time constants of the ultra-short-term time scale, the short-term time scale, the medium-term time scale, and the frequency characteristic vector, determines the required support power for each time scale, and then derives the power demand vector.

4. The method for emergency frequency support based on multi-timescale multi-element energy storage according to claim 3, characterized in that, The process of determining the required support power at each time scale, and thus deriving the power demand vector, includes: Through the calculation formula group The support power value on the ultra-short time scale was calculated. Support power value on short-term time scale Support power values ​​on a medium-term timescale ,in It is represented as a time constant for ultra-short time scales. It is represented as a time constant on a short-term time scale. This is expressed as a time constant on a medium-term timescale. and These are represented as power coefficients for the ultra-short-term, short-term, and medium-term time scales, respectively; the power demand vector is represented as... .

5. The method for emergency frequency support based on multi-timescale multi-element energy storage according to claim 1, characterized in that, The analysis, based on the matching degree matrix and power demand at each time scale, yields the output power of each energy storage device at each time scale, including: S701, Based on the matching degree matrix, select a preset number of energy storage devices with the highest matching degree for each time scale to form a set of candidate energy storage devices for each time scale; S702, establish an optimization model with the goal of minimizing total support cost. The objective function of the optimization model includes the energy storage equipment output cost term and the state of charge balance term. S703 sets power balance constraints, upper and lower limit constraints on the power of energy storage devices, and state of charge constraints to form a complete set of constraints. S704 solves the optimization model to obtain the power output allocation scheme of each energy storage device at each time scale, and then calculates the total power output and charging / discharging mode of each energy storage device.

6. The method for emergency frequency support based on multi-timescale multi-element energy storage according to claim 1, characterized in that, The process of generating control command sets for each energy storage device based on the optimized allocation results and sending them to the corresponding energy storage controllers to perform frequency support operations includes: Based on the power output allocation scheme of each energy storage device at each time scale, a raw control command containing the target power value, control duration, and charging / discharging mode is generated and converted into a standardized control command conforming to the communication protocol standard. The commands are then sent to the corresponding energy storage device controllers. Each energy storage device controller receives and parses the standardized control command, converts it into an executable control signal, and then drives the power converter to perform the corresponding charging / discharging operation to achieve the frequency support function.

7. The method for emergency frequency support based on multi-timescale multi-element energy storage according to claim 1, characterized in that, The real-time evaluation of frequency support effectiveness, and adaptive adjustment based on the evaluation results, includes: S901 monitors key performance indicators of the frequency support process in real time and calculates frequency recovery time, maximum frequency deviation and comprehensive utilization rate of energy storage equipment. S902, Based on the key performance indicators, the comprehensive effect evaluation score of the current control cycle is calculated using the effect evaluation function; S903, compare the comprehensive effect evaluation score with the preset target value to determine the adjustment direction and amount of each parameter; S904 applies the adjusted parameters to the time-scale decomposition and optimization allocation process of the next control cycle.

8. A system for implementing the multi-timescale, multi-element energy storage frequency emergency support method according to any one of claims 1-7, characterized in that, include: The frequency feature extraction module is used to acquire real-time frequency signals of power, extract frequency features, and generate frequency feature vectors. The multi-timescale task decomposition module analyzes and derives sub-tasks at each time scale based on the frequency feature vector, thereby obtaining the time constant and power requirement at each time scale. The energy storage matching degree calculation module calculates the matching degree matrix between each energy storage device and the time scale based on the real-time status parameters of each energy storage device and the power demand at each time scale. The power optimization and allocation module analyzes and derives the output power of each energy storage device at each time scale based on the matching degree matrix and the power demand at each time scale. The control command generation and distribution module generates control command sets for each energy storage device based on the optimized allocation results, and sends them to the corresponding energy storage controllers to perform frequency support operations. The effect evaluation and adaptive adjustment module evaluates the frequency support effect in real time and makes adaptive adjustments based on the evaluation results.

Images