Energy storage load multi-stage frequency modulation method and system based on soc dynamic partitioning

Abstract

The application belongs to the technical field of frequency modulation management, and specifically discloses a multi-stage frequency modulation method and system for energy storage load based on SOC dynamic partitioning. The method divides historical power grid frequency data according to hourly granularity, merges time periods according to the similarity of statistical characteristics of adjacent hours, and differentially configures the SOC sub-domain boundary for different time periods, thereby realizing dynamic matching of energy storage control and power grid frequency modulation demand, enabling the SOC partition to dynamically perceive the fluctuation intensity of the power grid frequency, solving the adaptability problem of insufficient adjustment capacity or excessive response of the fixed partition strategy. At the same time, the current battery state of charge value is obtained in real time, and the stability of the change trajectory thereof within a unit time is monitored. When the state of charge is unstable, the current state of charge value is filtered, and the high-confidence state of charge after filtering is used to determine the sub-domain, thereby ensuring the robustness of the sub-domain determination logic and making the issued frequency modulation instruction smoother and more continuous.

Description

Technical Field

[0001] This invention belongs to the field of frequency regulation management technology, and specifically discloses a multi-level frequency regulation method and system for energy storage load based on SOC dynamic zoning. Background Technology

[0002] As renewable energy penetration in the power system continues to increase, the pressure on power grid frequency regulation is growing. With the gradual phasing out of traditional thermal power units, the available frequency regulation capacity in the system is decreasing, thus requiring new power equipment, including energy storage, to have the capability to participate in power grid frequency regulation.

[0003] In the control of energy storage participating in grid frequency regulation, the main focus is on zone control based on state of charge (SOC) and mode switching control based on frequency deviation. By setting different SOC ranges, the charging and discharging behavior of energy storage under different states is constrained, which avoids overcharging and over-discharging of the battery while ensuring the availability of frequency regulation function.

[0004] For example, Chinese invention patent CN110492512A discloses a control method for frequency regulation or peak shaving modes in a photovoltaic-storage integrated system. The method sets the peak shaving charging and discharging time period according to the local load peak and valley periods and photovoltaic output characteristics, and configures fixed SOC boundaries for frequency regulation and peak shaving modes respectively (such as upper and lower limits of SOC in frequency regulation mode and charging / discharging limits in peak shaving mode), and gives higher priority to frequency regulation mode. When the grid frequency exceeds the frequency regulation dead zone, the energy storage is put into frequency regulation at rated power; otherwise, it participates in peak shaving as planned.

[0005] However, the above scheme still has the following shortcomings: First, the SOC domain boundary adopted by the above scheme is a pre-set fixed value. Since the fundamental purpose of SOC zoning is to respond to the frequency regulation demand driven by the grid frequency deviation, and the actual frequency fluctuations are different at different times, this fixed zoning strategy is difficult to match the time-varying characteristics of the actual frequency regulation demand, resulting in a mismatch between the available energy storage capacity and the regulation demand. During periods of strong frequency regulation demand, energy storage may be forced to exit the response due to prematurely reaching the preset SOC boundary; while during periods of weak frequency regulation demand, a large amount of available capacity is left idle due to the overly conservative boundary setting.

[0006] Secondly, after the SOC partition is defined, the above scheme usually makes the domain determination directly based on the instantaneous SOC value. However, due to factors such as measurement noise or control delay, the SOC may oscillate at high frequency and small amplitude near the partition boundary. Since the frequency modulation control switches the frequency modulation command directly based on the domain type, when the SOC frequently crosses the partition boundary, it may cause the domain result to switch frequently, which in turn causes the frequency modulation control to jump repeatedly, affecting the frequency modulation effect and battery life. Summary of the Invention

[0007] Therefore, one objective of this application is to provide a method and system for multi-level frequency regulation of energy storage load based on dynamic SOC zoning. By dynamically adjusting the SOC zoning boundary according to the grid frequency fluctuation characteristics and performing stability monitoring and processing after SOC zoning, the energy storage frequency regulation control effectively solves the problems mentioned in the background art.

[0008] The objective of this invention can be achieved through the following technical solution: The first aspect of this invention proposes a multi-level frequency regulation method for energy storage load based on SOC dynamic zoning, including: extracting the mean and standard deviation of frequency deviation for each time period of each day as statistical features based on historical power grid frequency data, and dividing the time period according to the statistical features of frequency deviation.

[0009] Configure corresponding SOC domain boundaries for each time period. The SOC domain includes a safe zone, an early warning zone, and a protection zone.

[0010] The frequency modulation level and direction are determined based on the current frequency deviation.

[0011] The system acquires the current state of charge (SOC) value in real time and monitors the stability of its change trajectory within a unit of time. When the SOC is unstable, the system filters the current SOC value and then determines the domain type of the current SOC based on the SOC domain boundary configured for the current time period.

[0012] The corresponding frequency modulation participation ratio is determined based on the current state of charge domain type and the frequency modulation level triggered by the frequency deviation.

[0013] The theoretical frequency modulation power requirement is calculated based on the current frequency deviation and frequency change rate, and multiplied by the frequency modulation participation ratio to obtain the preliminary frequency modulation power command. After amplitude verification, the final frequency modulation power command is generated.

[0014] The second aspect of the present invention proposes a multi-level frequency regulation system for energy storage load based on dynamic SOC zoning, comprising: a domain configuration module, which extracts the mean and standard deviation of frequency deviation for each time period of the day as statistical features based on historical power grid frequency data, divides the time period according to the frequency deviation statistical features, and configures corresponding SOC domain boundaries for each time period.

[0015] The frequency modulation detection module determines the frequency modulation level and direction based on the current frequency deviation.

[0016] The status monitoring module acquires the current battery state of charge value in real time and monitors the stability of its change trajectory within a unit of time. When the state of charge is unstable, it filters the current state of charge value and then determines the domain type of the current state of charge based on the SOC domain boundary configured for the current time period.

[0017] The participation ratio determination module determines the corresponding frequency modulation participation ratio based on the current state of charge domain type and the frequency modulation level triggered by the frequency deviation.

[0018] The frequency modulation command generation module calculates the theoretical frequency modulation power requirement based on the current frequency deviation and frequency change rate, and multiplies it by the frequency modulation participation ratio to obtain the preliminary frequency modulation power command. After amplitude verification, the final frequency modulation power command is generated.

[0019] Combining all the above technical solutions, the positive effects of this invention are as follows: 1. This invention divides historical power grid frequency data into hourly granularities, merges time periods based on the statistical similarity of adjacent hours, and configures SOC domain boundaries differently for different time periods, thereby achieving dynamic matching between energy storage control and power grid frequency regulation requirements. This enables SOC partitions to dynamically sense the intensity of power grid frequency fluctuations, automatically widening boundaries during periods of severe fluctuations and shrinking boundaries during periods of stability, thus solving the adaptability problem of insufficient regulation capacity or over-response in fixed partition strategies.

[0020] 2. This invention acquires the current state of charge (SOC) value of the battery in real time and monitors the stability of its change trajectory within a unit of time. When the SOC value is unstable, it filters the current SOC value. The filtered high-confidence SOC value determines the domain in which it is located, effectively blocking the transmission of noise signals, ensuring the stability of the domain determination logic, and making the issued frequency modulation commands smoother and more continuous. Attached Figure Description

[0021] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0022] Figure 1 This diagram illustrates the implementation steps of the multi-level frequency regulation method for energy storage load based on SOC dynamic zoning in this invention.

[0023] Figure 2 This is a schematic diagram illustrating how the frequency modulation participation ratio is determined based on the current state of charge domain type and the frequency modulation level triggered by the frequency deviation in this invention.

[0024] Figure 3 This is a module connection diagram of the energy storage load multi-level frequency regulation system based on SOC dynamic zoning in this invention. Detailed Implementation

[0025] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Example 1.

[0027] This invention proposes a multi-level frequency regulation method for energy storage loads based on SOC dynamic zoning.

[0028] When grid frequency deviations require frequency regulation by energy storage batteries, the battery's continuous frequency regulation capability is limited by its current state of charge (SOC), as the battery is essentially a finite energy device. Without SOC management, the battery is highly susceptible to overcharging or over-discharging during frequent adjustments, forcing it to withdraw from frequency regulation.

[0029] To achieve precise guidance on battery frequency regulation behavior, this invention divides the full charge range of the battery into a safe range, a warning range, and a protection range.

[0030] The safe range refers to the state of charge (SOC) being within the ideal center range. Within this range, the battery has ample bidirectional adjustment margin, allowing the battery to fully support the restoration of the grid frequency.

[0031] The warning range refers to the transition area where the SOC deviates from the center value and extends to both ends. In this range, the battery's unidirectional adjustment capability begins to be limited.

[0032] The protection zone refers to the dangerous area where the State of Charge (SOC) is close to its physical limit. In this zone, the battery faces a high risk of being damaged by overcharging or being cut off by over-discharging.

[0033] The aforementioned partitioning allows for dynamic matching of the battery's available capacity with the grid's real-time demand, thus avoiding frequency regulation interruptions caused by battery depletion and eliminating invalid mode switching.

[0034] However, existing technologies typically use static, fixed thresholds for zoning. For example, the safety zone is set at 40%–60% state of charge (SOC), the warning zone corresponds to 20%–40% or 60%–80% SOC, and the protection zone corresponds to <20% or >80% SOC. Considering that the fundamental purpose of SOC zoning is to respond to frequency regulation demands driven by grid frequency deviations, and that actual frequency fluctuations vary across different time periods, this fixed zoning strategy struggles to match the time-varying characteristics of actual frequency regulation demands, leading to a mismatch between available energy storage capacity and regulation requirements.

[0035] To address the aforementioned issues, this invention constructs several frequency fluctuation scenario periods based on the time-varying characteristics of power grid frequency fluctuations, and then configures corresponding SOC domain boundaries for each scenario period, ensuring that the battery can match the power grid frequency regulation requirements in the optimal state at any time.

[0036] See details Figure 1 As shown, the implementation of the present invention includes the following steps: S1, extracting the mean and standard deviation of frequency deviation for each time period of the day as statistical features based on historical power grid frequency data, and dividing the time period according to the statistical features of frequency deviation.

[0037] Given that power grid frequency fluctuations are not entirely random noise processes, the periodic rhythms of residential life and industrial production cause the active power deficit of the power grid to exhibit predictable time-varying fluctuations, resulting in similar characteristics of power grid frequency deviations within the same time period. This provides a theoretical basis for constructing several frequency fluctuation scenario periods based on the time-varying characteristics of power grid frequency fluctuations.

[0038] In one specific embodiment, the time period is divided according to the statistical characteristics of frequency deviation, as described in the following process: S11, Frequency deviation sequence construction: Based on historical power grid frequency monitoring records, the measured frequency of each record is subtracted from the nominal frequency of the power grid (e.g., 50Hz) to obtain a frequency deviation sequence, which characterizes the magnitude and direction of the power grid frequency deviation from the nominal value.

[0039] S12. Statistical Feature Extraction: Given that the hour is the basic time granularity, the frequency deviation sequence is grouped by hour. The mean of all frequency deviation samples in each hour period is calculated as the mean of frequency deviation in that period, which represents the average offset level of frequency deviation in that period. At the same time, the standard deviation of all frequency deviation samples in each hour period is calculated as the standard deviation of frequency deviation in that period, which represents the severity of frequency disturbance in that period. Here, the mean and standard deviation are used as two-dimensional statistical features, which can comprehensively reflect the overall characteristics of power grid frequency fluctuations in that period.

[0040] S13. Merging adjacent time periods: Since the changes in power grid load within adjacent time periods often have continuity and inertia, the mean frequency deviation and standard deviation of the frequency deviation of adjacent hours are clustered separately. When adjacent hours are divided into the same cluster because the mean frequency deviation and standard deviation of the frequency deviation are similar, it indicates that these hour segments have homogeneous frequency fluctuations. The adjacent hours are merged into the same time period to ensure that the divided time period represents a stable operating condition in a physical sense, rather than a mechanical time slice.

[0041] S14. Time period boundary definition: When adjacent hours are divided into different clusters due to differences in the mean or standard deviation of frequency deviation, it indicates that the power grid frequency fluctuation has changed. These adjacent hours are used as the dividing point of the time period.

[0042] The above clustering divides the day into several time periods, each corresponding to a frequency fluctuation scenario.

[0043] S2. Configure corresponding SOC domain boundaries for each time period. The SOC domain includes a safety zone, an early warning zone, and a protection zone.

[0044] As a preferred embodiment of the present invention, the specific configuration process of the SOC domain boundary is as follows: S21, for the period when the mean frequency deviation is close to zero and the standard deviation of frequency deviation is close to zero, the power grid is in steady state during these periods. At this time, the energy storage battery does not need to perform frequent high-power throughput to support the frequency. At this time, the safety range, warning range and protection range are kept unchanged by using the preset basic range, which represents the general control under normal operating conditions.

[0045] S22. For periods when the mean frequency deviation is positive and the standard deviation of the frequency deviation is greater than zero, the power grid has a continuous high-frequency deviation and violent fluctuations during these periods. Under such conditions, the power grid frequency is higher than the nominal value, and the energy storage battery needs to discharge to suppress the frequency rise. Since the discharge process will continuously reduce the battery's state of charge, in order to ensure sufficient discharge capacity during periods of high demand, it is necessary to shift the SOC domain towards the low-charge direction.

[0046] S221. Considering that the larger the average frequency deviation, the more serious the high-frequency offset of the power grid, and the higher the corresponding discharge frequency regulation demand, in order to match the frequency regulation capability with the demand intensity, the above time periods are arranged in descending order of the average frequency deviation to form a discharge preference time period sequence, which reflects the relative urgency of the discharge frequency regulation demand in each time period.

[0047] S222. Based on the above sequence, the SOC domain boundary is adjusted downward in stages. Specifically, the range from the minimum allowable state of charge of the battery (i.e., the hard lower limit set by the battery management system to prevent over-discharge) to the basic lower limit of the safe range is defined as the downward shifted usable capacity.

[0048] S223. Since the earlier time periods in the ranking represent the more severe the grid frequency deviation and the stronger the demand for discharge frequency regulation, they should be given a larger downward offset. Therefore, the average frequency deviation of each time period is compared with the cumulative average frequency deviation of all time periods in the sequence to obtain the allocation weight of each time period. The allocation weight of each time period is multiplied by the downward shiftable available capacity to obtain the downward shift of the safety range for each time period. Then, the lower limit of the safety range for each time period is adjusted according to the downward shift, while the upper limit of the safety range remains unchanged at the upper limit of the basic range.

[0049] S224. Given that the lower boundary of the protection zone was originally close to the minimum allowable state of charge of the battery, in order to avoid the overlap or gap between the warning zone and the protection zone due to the safety zone being shifted downward, the maximum downward shift amount in all discharge preference periods is taken as the unified downward shift amount of the lower boundary of the protection zone. The lower boundary of the protection zone is shifted downward by the unified downward shift amount, and after the shift, it is not lower than the minimum allowable state of charge of the battery. The lower boundary of the warning zone in the discharge preference period is reset to the upper boundary of the protection zone after the shift, and the upper boundary of the warning zone is reset to the lower limit of the safety zone after the current period.

[0050] S23. For periods where the mean frequency deviation is negative and the standard deviation of the frequency deviation is greater than zero, the power grid experiences persistent low-frequency deviation and severe fluctuations during these periods. Under such conditions, the grid frequency is lower than the nominal value, and the energy storage battery needs to be charged to suppress further frequency decline. Since the charging process continuously increases the battery's state of charge, it is necessary to shift the overall SOC domain towards higher charge to ensure sufficient charging capacity margin during periods of high demand.

[0051] S231. Considering that under negative conditions, the smaller the mean frequency deviation and the larger the absolute value, the more severe the low-frequency shift and the higher the corresponding charging frequency regulation demand intensity, the above time periods are arranged in ascending order of the mean frequency deviation to form a charging preference time period sequence.

[0052] S232. Based on the above sequence, the range between the upper limit of the basic range of the safety interval and the maximum allowable state of charge of the battery (the hard upper limit set by the battery management system to prevent overcharging) is determined as the upward shiftable capacity.

[0053] S233. Compare the absolute value of the mean frequency deviation of each time period with the cumulative value of the absolute value of the mean frequency deviation of all time periods in the sequence to obtain the allocation weight of each time period. Multiply the allocation weight of each time period by the available capacity for upward shift to obtain the upward shift amount of the safe interval for each time period. Adjust the upper limit of the safe interval for each time period according to the upward shift amount, while keeping the lower limit of the safe interval unchanged from the lower limit of the basic range.

[0054] S234. Take the maximum upward shift amount in all charging preference periods as the unified upward shift amount of the upper boundary of the protection interval, shift the upper boundary of the protection interval upward by the unified upward shift amount, and ensure that the shifted value is not higher than the maximum allowable state of charge of the battery. At the same time, reset the upper boundary of the warning interval in the charging preference period to the lower boundary of the protection interval after the shift, and reset the lower boundary of the warning interval to the upper limit of the safety interval after the current period.

[0055] For the example of applying the above operation, assume the minimum permissible SOC of the battery is 5%, the maximum permissible SOC of the battery is 95%, the basic range of the safety range is 40% to 60%, the warning range is 20% to 40% or 60% to 80%, and the protection range is <20% or >80%.

[0056] Suppose that through historical frequency data analysis, the day is divided into 5 time periods. The mean frequency deviation between time period 1 and time period 2 is positive, at 0.15 and 0.10 respectively. The mean frequency deviation between time period 3 and time period 4 is negative, at -0.12 and -0.08 respectively. The mean frequency deviation of time period 5 is close to zero and the dispersion value is close to zero.

[0057] During the aforementioned mid-period 5, the basic ranges of the safe zone, warning zone, and protection zone are maintained.

[0058] Time periods 1 and 2 are arranged in descending order of the mean frequency deviation to form a discharge preference time period sequence: [Time period 1, Time period 2].

[0059] Downward shift of available capacity = lower limit of safe range (40%) - minimum allowable SOC of battery (5%) = 35%.

[0060] The weighting for period 1 is 0.15 / 0.25 = 0.6, and the downward shift is 0.6 × 35% = 21%.

[0061] The weighting ratio for period 2 is 0.10 / 0.25 = 0.4, and the downward shift is 0.4 × 35% = 14%.

[0062] Period 1: Lower limit of the safe range = 40% - 21% = 19%, upper limit remains at 60%, therefore the safe range is adjusted to 19% to 60%.

[0063] Period 2: Lower limit of the safe range = 40% - 14% = 26%, upper limit remains at 60%, therefore the safe range is adjusted to 26% to 60%.

[0064] Since the maximum downward shift is 21%, the original lower boundary of the protection range was 20%, which becomes -1% after shifting downward by 21%, which is lower than the minimum allowable SOC of the battery by 5%. Therefore, the physical lower limit of 5% is taken as the adjusted lower boundary of the protection range, that is, the protection range is adjusted to <5%.

[0065] The lower boundary of the warning interval is uniformly set as 5% of the upper boundary of the adjusted protection interval, and the upper boundary of the warning interval is set as the lower limit of the safe interval for the current time period.

[0066] Warning range for period 1: 5% to 19%.

[0067] Warning range for period 2: 5% to 26%.

[0068] Time periods 3 and 4 are arranged in ascending order of the mean frequency deviation to form a charging preference time period sequence: [Time period 3, Time period 4].

[0069] Upward shift of usable capacity = maximum allowable SOC of battery (95%) - basic upper limit of safe range (60%) = 35%.

[0070] The weighting for period 3 is 0.12 / 0.20 = 0.6, and the upward shift is 0.6 × 35% = 21%.

[0071] The weighting for period 4 is 0.08 / 0.20 = 0.4, and the upward shift is 0.4 × 35% = 14%.

[0072] Period 3: The upper limit of the safe range = 60% + 21% = 81%, and the lower limit remains at 40%. Therefore, the safe range is adjusted to 40% to 81%.

[0073] Period 4: The upper limit of the safe range = 60% + 14% = 74%, and the lower limit remains at 40%. Therefore, the safe range is adjusted to 40% to 74%.

[0074] Since the maximum upward shift is 21%, the original upper boundary of the protection zone was 80%, and after shifting upward by 21%, it becomes 101%, which is higher than the battery's maximum allowable SOC of 95%. Therefore, the physical upper limit of 95% is taken as the adjusted upper boundary of the protection zone, that is, the protection zone is adjusted to >95%.

[0075] The upper boundary of the warning interval is uniformly set to 95% of the lower boundary of the adjusted protection interval, and the lower boundary of the warning interval is set to the upper limit of the safe interval for the current time period.

[0076] Warning range for period 3: 81% to 95%.

[0077] Warning range for period 4: 74% to 95%.

[0078] S3. Determine the frequency modulation level and frequency modulation direction based on the current frequency deviation.

[0079] In frequency regulation control of energy storage batteries, determining the extent to which energy storage batteries participate in frequency regulation requires not only clarifying the current domain state of the battery, but also determining the degree of frequency deviation. This is because the former determines the feasibility of participating in frequency regulation, while the latter determines the necessity of frequency regulation. Combining the degree of frequency deviation with the SOC domain state essentially matches the frequency regulation demand level of the power grid with the frequency regulation capability level of the battery, thereby achieving multi-level coordinated frequency regulation control.

[0080] As one possible way to implement the above scheme, determining the frequency modulation level and frequency modulation direction includes the following: real-time acquisition of the frequency deviation between the current frequency and the nominal frequency of the power grid; if the current frequency deviation is positive, the frequency modulation direction is discharging; if the current frequency deviation is negative, the frequency modulation direction is charging.

[0081] If the absolute value of the current frequency deviation is between the first frequency deviation limit and the second frequency deviation limit, it is classified as the first frequency modulation level, where the first frequency deviation limit is less than the second frequency deviation limit.

[0082] If the absolute value of the current frequency deviation is greater than the second frequency deviation limit, it is classified as the second frequency modulation level.

[0083] In the above implementation, the first frequency deviation limit represents the upper limit of a slight deviation of the power grid frequency, and the second frequency deviation limit represents the critical value of a significant abnormality in the power grid frequency. Together, they constitute the dividing point of multi-level frequency regulation. The specific setting can be referred to the national standard. For example, under normal circumstances, the power grid frequency deviation shall not exceed ±0.2Hz; under accident circumstances, ±0.5Hz is allowed for a short period of time. Therefore, the first frequency deviation limit can be set to 0.2Hz, and the second frequency deviation limit can be set to 0.5Hz.

[0084] S4. Real-time acquisition of the current battery state of charge value and stability monitoring of its change trajectory within a unit of time. When the state of charge is unstable, the current state of charge value is filtered, and then the domain type of the current state of charge is determined according to the SOC domain boundary configured for the current time period.

[0085] After configuring the SOC domain boundaries, when the frequency deviation exceeds the first frequency deviation limit, it is necessary to determine the domain type based on the current SOC value to match the corresponding frequency modulation strategy. However, in actual operation, due to sensor measurement noise and other factors, the SOC often exhibits high-frequency, small-amplitude oscillations near the domain boundaries. If the domain determination is made directly based on the instantaneous SOC value, it will lead to frequent switching of the domain results, thereby causing repeated jumps in frequency modulation control.

[0086] To address the aforementioned issues, this invention introduces stability monitoring of the SOC change trajectory, aiming to suppress domain misjudgment caused by instantaneous SOC disturbances and improve the continuity of frequency modulation control.

[0087] Specifically, stability monitoring is implemented as follows: record the battery state of charge values ​​at multiple consecutive sampling times in real time to form a sequence of state of charge change trajectories.

[0088] The trajectory sequence of the state of charge change is tracked to cross the boundary of the SOC domain. If the state of charge repeatedly crosses the same domain boundary in a short period of time (e.g., 5 sampling times), the state of charge is determined to be unstable.

[0089] When the domain state is determined to be unstable, the current charge state value is smoothed by using a moving average filter or median filter, and the filtered charge state value is used to replace the original value for domain type determination.

[0090] Furthermore, the real-time acquired current battery state of charge value or the filtered state of charge value is matched interval by interval with the SOC domain boundary configured for the current time period, and the successfully matched domain type is extracted as the domain type of the current state of charge.

[0091] S5. Determine the corresponding frequency modulation participation ratio based on the current state of charge domain type and the frequency modulation level triggered by the frequency deviation.

[0092] See Figure 2 As shown, in a preferred embodiment, the frequency regulation participation ratio is determined as follows: when the current domain type is a safe range, this range is far from the overcharge and over-discharge boundaries, and the battery has sufficient bidirectional adjustment margin. In this state, regardless of the frequency regulation level of the grid frequency deviation, the frequency regulation participation ratio is set to full participation, that is, the battery's available frequency regulation capacity can be fully utilized.

[0093] When the current domain type is a warning zone, this zone is close to the upper or lower limit of the safe zone, and the battery no longer has the full capacity available for frequency regulation. If it still participates in frequency regulation at a fixed ratio in this state, it may cause the frequency regulation response to be mismatched with the actual carrying capacity of the battery. Therefore, the frequency regulation participation ratio should be set according to the frequency regulation level and configured using a graded incremental strategy: the first frequency regulation level (with small frequency deviation) is set to a low participation ratio (such as 30%), which only provides fine-tuning support and prioritizes the protection of battery life.

[0094] The second frequency modulation level (with larger frequency deviations) allows for a high participation rate (e.g., 60%) to cope with emergency frequency instability.

[0095] When the current domain type is the protection zone, the zone has entered the danger zone of overcharging or over-discharging of the battery. Continuing to participate in frequency regulation is very likely to trigger cell damage. Therefore, the frequency regulation participation ratio is forcibly set to zero, and the energy storage battery actively exits the frequency regulation mode.

[0096] It should be noted that the frequency modulation participation ratio mentioned above is used to constrain the theoretical frequency modulation power demand. The actual frequency modulation power is determined by multiplying the theoretical demand value calculated by the droop control based on the current frequency deviation by this ratio.

[0097] S6. Calculate the theoretical frequency modulation power requirement based on the current frequency deviation and frequency change rate, and multiply it by the frequency modulation participation ratio to obtain the preliminary frequency modulation power command. After amplitude verification, generate the final frequency modulation power command.

[0098] In one optional implementation, the process of obtaining the initial frequency regulation power command is as follows: First, the current grid frequency deviation is input to the droop control system. The droop control system refers to a control strategy that simulates the primary frequency regulation characteristics of a synchronous generator. It mainly maps the frequency deviation linearly into an active power regulation quantity through the droop coefficient. The result is called the droop response component, which reflects the system's compensation requirement for steady-state frequency offset.

[0099] Secondly, by performing differential calculations on the grid frequency, the current frequency change rate is obtained, and multiplied by the virtual inertia response coefficient to obtain the inertia response component, which is used to provide instantaneous power support in the early stages of rapid frequency changes.

[0100] Subsequently, the droop response component and the inertia response component are algebraically superimposed to form the theoretical frequency modulation power requirement.

[0101] Finally, the theoretical frequency regulation power demand is multiplied by the frequency regulation participation ratio to obtain the preliminary frequency regulation power command. This reflects the control logic that couples and matches the objective frequency regulation demand of the power grid with the current subjective adjustment capability of the battery, so that the obtained preliminary frequency regulation power command is a power request adapted to the battery state.

[0102] Given that the initial frequency modulation power command may exceed the power range that the battery can currently safely execute, directly issuing such an over-limit command could easily lead to power response distortion and, in severe cases, accelerate battery aging. Therefore, it is necessary to perform amplitude verification on the initial frequency modulation power command. The specific verification process is as follows: real-time acquisition of the maximum allowable charging power value and the maximum allowable discharging power value corresponding to the current battery state of charge.

[0103] Determine the direction of the initial frequency modulation power command. When the initial frequency modulation power command is in the charging direction, compare the absolute value of the initial frequency modulation power command with the maximum allowable charging power value, and take the smaller value of the two as the final frequency modulation power command.

[0104] When the initial frequency modulation power command is in the discharge direction, the absolute value of the initial frequency modulation power command is compared with the maximum allowable discharge power value, and the smaller of the two values ​​is taken as the final frequency modulation power command.

[0105] After the above verification and necessary trimming, the final frequency regulation power command is generated, which satisfies the grid frequency regulation requirements and follows the battery safety operation boundary.

[0106] Example 2.

[0107] See Figure 3As shown, the present invention proposes a multi-level frequency regulation system for energy storage load based on SOC dynamic zoning, including: a domain configuration module, which extracts the mean and standard deviation of frequency deviation for each time period of the day as statistical features based on historical power grid frequency data, divides the time period according to the frequency deviation statistical features, and configures corresponding SOC domain boundaries for each time period.

[0108] The frequency modulation detection module determines the frequency modulation level and direction based on the current frequency deviation.

[0109] The state monitoring module is connected to the domain configuration module to obtain the current battery state of charge value in real time and monitor the stability of its change trajectory within a unit of time. When the state of charge is unstable, the current state of charge value is filtered and then the domain type of the current state of charge is determined according to the SOC domain boundary configured for the current time period.

[0110] The participation ratio determination module is connected to the state monitoring module and the frequency modulation detection module respectively. It determines the corresponding frequency modulation participation ratio based on the current state of charge domain type and the frequency modulation level triggered by the frequency deviation.

[0111] The frequency modulation command generation module is connected to the participation ratio determination module. It calculates the theoretical frequency modulation power requirement based on the current frequency deviation and frequency change rate, and multiplies it by the frequency modulation participation ratio to obtain the preliminary frequency modulation power command. After amplitude verification, the final frequency modulation power command is generated.

[0112] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0113] Those skilled in the art will recognize that the algorithmic steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0114] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0115] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0116] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-level frequency regulation method for energy storage load based on SOC dynamic zoning, characterized in that, Includes the following steps: The mean and standard deviation of frequency deviation for each time period of the day are extracted from historical power grid frequency data as statistical features, and time periods are divided based on the statistical features of frequency deviation. Configure corresponding SOC domain boundaries for each time period, wherein the SOC domain includes a safety zone, an early warning zone, and a protection zone; Determine the frequency modulation level and direction based on the current frequency deviation; The system acquires the current state of charge (SOC) value in real time and monitors the stability of its change trajectory within a unit of time. When the SOC value is unstable, it filters the current SOC value and then determines the domain type of the current SOC based on the SOC domain boundary configured for the current time period. Determine the corresponding frequency modulation participation ratio based on the current state of charge domain type and the frequency modulation level triggered by the frequency deviation; The theoretical frequency modulation power requirement is calculated based on the current frequency deviation and frequency change rate, and multiplied by the frequency modulation participation ratio to obtain the preliminary frequency modulation power command. After amplitude verification, the final frequency modulation power command is generated. The specific details for configuring corresponding SOC domain boundaries for each time period are as follows: For periods when the mean frequency deviation approaches zero and the standard deviation of frequency deviation approaches zero, the basic ranges of the safety interval, warning interval, and protection interval will remain unchanged. The time periods with a positive mean frequency deviation and a standard deviation greater than zero are arranged in descending order of the mean frequency deviation to form a discharge preference time period sequence. The range between the minimum permissible state of charge of the battery and the lower limit of the basic range of the safety interval is defined as the downward shifted usable capacity. Based on the proportion of the average frequency deviation of each time period to the total sum of the sequences, the available capacity for downward shift is allocated to obtain the downward shift amount of the lower limit of the safe interval for each time period, and the lower limit of the safe interval is adjusted accordingly, while the upper limit of the safe interval remains unchanged at the upper limit of the basic range. The maximum downward shift during all discharge preference periods is taken as the uniform downward shift of the lower boundary of the protection interval. The lower boundary of the protection interval is shifted downward by this amount, and the shifted value is not lower than the minimum allowable state of charge of the battery. The lower boundary of the warning interval during the discharge preference period is reset to the upper boundary of the shifted protection interval, and the upper boundary of the warning interval is reset to the lower limit of the safety interval after the current period is adjusted. The process of determining the frequency modulation level and frequency modulation direction based on the current frequency deviation includes: The deviation between the current frequency and the nominal frequency of the power grid is obtained in real time. If the frequency deviation is greater than zero, the frequency adjustment direction is the discharge direction; if the frequency deviation is less than zero, the frequency adjustment direction is the charging direction. If the absolute value of the current frequency deviation is greater than the first frequency deviation limit but less than the second frequency deviation limit, it is classified as the first frequency modulation level, where the first frequency deviation limit is less than the second frequency deviation limit. If the absolute value of the current frequency deviation is greater than the second frequency deviation limit, it is classified as the second frequency modulation level.

2. The multi-level frequency regulation method for energy storage load based on SOC dynamic zoning as described in claim 1, characterized in that: The process for dividing time periods based on frequency deviation statistical characteristics is as follows: Based on historical power grid frequency monitoring records, the measured frequency of each record is subtracted from the nominal frequency of the power grid to obtain a frequency deviation sequence; The frequency deviation sequence is grouped by hour, and the mean of all frequency deviation samples in each hour period is calculated as the mean of frequency deviation in that hour period. At the same time, the standard deviation of all frequency deviation samples in each hour period is calculated as the standard deviation of frequency deviation in that hour period. Cluster the mean frequency deviation and standard deviation of adjacent hours separately. When adjacent hours are classified into the same cluster because their mean frequency deviation and standard deviation are similar, then the adjacent hours are merged into the same time period. The hour that is divided into different clusters due to differences in the mean or standard deviation of frequency deviation between adjacent hours is used as the dividing point of the time period; The above clustering method divides the day into several time periods.

3. The multi-level frequency regulation method for energy storage load based on SOC dynamic zoning as described in claim 1, characterized in that: The configuration of corresponding SOC domain boundaries for each time period also includes the following: The time periods with a negative mean frequency deviation and a standard deviation of frequency deviation greater than zero are sorted in ascending order of the mean frequency deviation to form a charging preference time period sequence. The range between the upper limit of the basic range of the safety interval and the maximum allowable state of charge of the battery is defined as the upward shiftable available capacity. Based on the proportion of the absolute value of the frequency deviation in each time period to the total sum of the sequences, the available capacity for upward shift is allocated, and the upward shift amount of the upper limit of the safe interval in each time period is obtained. The upper limit of the safe interval is adjusted accordingly, while the lower limit of the safe interval remains unchanged from the lower limit of the basic range. The maximum upward shift during all charging preference periods is taken as the uniform upward shift of the upper boundary of the protection interval. The upper boundary of the protection interval is shifted upward by this amount, and the shifted value is not higher than the maximum allowable state of charge of the battery. The upper boundary of the warning interval during the charging preference period will be reset to the lower boundary of the shifted protection interval, and the lower boundary of the warning interval will be reset to the upper limit of the safety interval after the current period.

4. The multi-level frequency regulation method for energy storage load based on SOC dynamic zoning as described in claim 1, characterized in that: The stability monitoring process is as follows: The battery state of charge (SOC) values ​​are recorded in real time at multiple consecutive sampling moments to form a sequence of SOC change trajectories. By tracking the behavior of the state of charge change trajectory sequence near the SOC domain boundary, if the state of charge repeatedly crosses the same domain boundary in a short period of time, the state of charge is determined to be unstable.

5. The multi-level frequency regulation method for energy storage load based on SOC dynamic zoning as described in claim 1, characterized in that: Determining the domain type of the current state of charge includes the following: The current battery state of charge value, or the filtered state of charge value, is obtained in real time and matched against the SOC domain boundary configured for the current time period. The domain type of the successfully matched domain is extracted as the domain type of the current state of charge.

6. The multi-level frequency regulation method for energy storage load based on SOC dynamic zoning as described in claim 1, characterized in that: The determination of the frequency modulation participation ratio is as follows: When the current domain type is a safe zone, the frequency modulation participation ratio is set to full participation. When the current domain type is the warning interval, the frequency modulation participation ratio increases step by step as the frequency modulation level increases, with the first frequency modulation level set to a low participation ratio and the second frequency modulation level set to a high participation ratio. When the current domain type is a protection zone, the frequency modulation participation ratio is set to zero, and the frequency modulation mode is exited.

7. The multi-level frequency regulation method for energy storage load based on SOC dynamic zoning as described in claim 1, characterized in that: The process of generating the final frequency modulation power command after amplitude verification is as follows: Real-time acquisition of the maximum allowable charging power and maximum allowable discharging power corresponding to the current battery state of charge; When the initial frequency modulation power command is in the charging direction, the absolute value of the initial frequency modulation power command is compared with the maximum allowable charging power value, and the smaller of the two values ​​is taken as the final frequency modulation power command. When the initial frequency modulation power command is in the discharge direction, the absolute value of the initial frequency modulation power command is compared with the maximum allowable discharge power value, and the smaller of the two values ​​is taken as the final frequency modulation power command.

8. A multi-level frequency regulation system for energy storage load based on SOC dynamic zoning, characterized in that, The system is applied to the multi-level frequency regulation method for energy storage load based on SOC dynamic zoning as described in any one of claims 1-7, specifically including: The domain configuration module extracts the mean and standard deviation of frequency deviation for each time period of the day as statistical features based on historical power grid frequency data. It divides the time period according to the frequency deviation statistical features and configures the corresponding SOC domain boundary for each time period. The frequency modulation detection module determines the frequency modulation level and direction based on the current frequency deviation. The status monitoring module acquires the current battery state of charge value in real time and monitors the stability of its change trajectory within a unit of time. When the state of charge is unstable, it filters the current state of charge value and then determines the domain type of the current state of charge based on the SOC domain boundary configured for the current time period. The participation ratio determination module determines the corresponding frequency modulation participation ratio based on the current state of charge domain type and the frequency modulation level triggered by the frequency deviation. The frequency modulation command generation module calculates the theoretical frequency modulation power requirement based on the current frequency deviation and frequency change rate, and multiplies it by the frequency modulation participation ratio to obtain the preliminary frequency modulation power command. After amplitude verification, the final frequency modulation power command is generated.

Images