A flexible control based photovoltaic power storage system

Abstract

The present application relates to the technical field of photovoltaic electric energy, and particularly relates to a photovoltaic electric energy storage system based on flexible control, which comprises a data acquisition module, a mode determination module, a strategy generation module, a strategy execution module and a strategy adjustment module. The present application couples the change gradient of the partition output power and the change rate of the sudden cloud cluster blocking coefficient to construct a photovoltaic jump index, simultaneously fuses the bus frequency deviation and the DC bus voltage fluctuation amplitude to construct a bus stability index, and combines the battery cluster temperature difference gradient to construct an energy storage bearing coefficient. After the strategy execution, the correction evaluation index is constructed by the bus frequency backfall amplitude and the battery cluster temperature difference change amount, which effectively solves the problems of the prediction error accumulation in the isolated network scene without communication support, the secondary fluctuation of the system frequency and the lag of the support response caused by the open-loop feedforward scheduling mode, the lack of real-time evaluation of the actual adjustment effect and the threshold dynamic optimization.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power technology, and in particular to a photovoltaic power storage system based on flexible control. Background Technology

[0002] As the penetration rate of clean energy in power supply systems in remote areas continues to increase, isolated grid photovoltaic energy storage systems in mountainous environments are becoming a key infrastructure for ensuring continuous power supply to important loads such as communication base stations and monitoring sites. However, due to the influence of complex terrain and local microclimates, the rapid generation and movement of cloud clusters cause short-term drastic power fluctuations, posing a severe challenge to the frequency stability and voltage support capabilities of the system. At the same time, during the high-frequency and large-amplitude charging and discharging regulation process, the internal temperature gradient and rate stress of the energy storage battery clusters continuously accumulate, and the thermal safety risks increase accordingly. How to achieve synergistic optimization of stable support and equipment life protection under high dynamic disturbances has become a core challenge facing current technological development.

[0003] Chinese Patent Publication No. CN120497977A discloses a method and device for intelligent scheduling of photovoltaic power generation based on dynamic supply and demand forecasting. The method includes: acquiring the current location information of a cloud cluster; wherein the cloud cluster location information includes the cloud cluster centroid location, cloud cluster area, cloud cluster velocity state vector, and cloud layer thickness; combining the current cloud cluster location information with wind speed data, and using Kalman filtering for gridded prediction to obtain the cloud cluster coverage state within a preset time period; and using an shading probability function to calculate the shading of the cloud cluster coverage state within the preset time period to obtain the cloud coverage state within the preset time period. The probability of cloud cover obstruction is determined. A light intensity attenuation model is used to predict the cloud cover status and probability of cloud cover obstruction within a preset time period, thereby obtaining the predicted light intensity value within the preset time period. A photovoltaic power linear model is used to convert the predicted light intensity value within the preset time period, thereby obtaining the predicted photovoltaic power generation value within the preset time period. The difference between the predicted electricity load curve within a preset time period and the corresponding predicted photovoltaic power generation value is calculated to obtain the predicted supply and demand power within the preset time period. The energy storage system is scheduled and allocated according to the predicted supply and demand power within the preset time period to complete the predictive power supply scheduling for electricity consumption.

[0004] Therefore, the existing technology has the following problems: it relies on complex multi-level prediction models and meteorological data sources, resulting in high parameter acquisition costs and difficulty in accurately modeling local microclimates in mountainous areas. This can easily lead to significant deviations between the predicted cloud cover probability and photovoltaic power, thus affecting the accuracy of scheduling decisions. It relies on preset electricity load prediction curves and static scheduling strategies of energy storage systems, failing to consider the balance of charge state and thermal stress accumulation of energy storage battery clusters during dynamic adjustment in real time. This can easily lead to safety risks such as overcharging or over-discharging of some battery clusters or excessive temperature differences, reducing the overall lifespan of the energy storage system. It relies on an open-loop feedforward scheduling mode, which only allocates power based on predicted supply and demand in a one-time manner. It lacks a real-time feedback correction mechanism for the actual bus frequency and voltage recovery effect after scheduling. This can easily cause secondary fluctuations in system frequency or delayed support response due to the accumulation of prediction errors, making it difficult to achieve stable and reliable operation in isolated network scenarios without communication support. Summary of the Invention

[0005] To address this, the present invention provides a photovoltaic energy storage system based on flexible control. This system overcomes the problems in existing technologies, such as the accumulation of prediction errors, easy secondary fluctuations in system frequency, or delayed support response, caused by the reliance on open-loop feedforward scheduling mode, lack of real-time evaluation of actual adjustment effects and dynamic optimization of thresholds, which lead to these issues in isolated grid scenarios without communication support. This is achieved by sensing the dynamics of cloud shading and the stable state of the bus in real time and performing adaptive feedback correction based on the execution effect.

[0006] To achieve the above objectives, the present invention provides a photovoltaic energy storage system based on flexible control, comprising: The data acquisition module is used to collect in real time the zoned output power, bus frequency deviation, DC bus voltage fluctuation amplitude, battery cluster temperature gradient and charge / discharge rate of several energy storage battery clusters, as well as the sudden cloud shading coefficient of the mountain photovoltaic array under isolated grid operation. The mode determination module is used to determine whether to enter the jump absorption mode based on the photovoltaic jump index and the preset jump trigger threshold. The photovoltaic jump index is constructed based on the rate of change of the sudden cloud shading coefficient and the gradient of the change of the partition output power. The strategy generation module is used to determine the preset participation ratio of each energy storage cluster and the upper limit of the charge / discharge rate corresponding to the energy storage capacity coefficient and the bus stability index based on the determination result of entering the jump absorption mode, so as to form a partitioned flexible storage strategy. The energy storage capacity coefficient is determined based on the bus stability index and the temperature difference gradient of the battery cluster, and the bus stability index is constructed based on the bus frequency deviation and the DC bus voltage fluctuation amplitude. The strategy execution module is used to execute power-limited storage control based on the partitioned flexible storage strategy; The strategy adjustment module is used to adjust the preset jump trigger threshold or the upper limit of the multiplier based on the correction evaluation index and the preset stability conditions. The correction evaluation index is constructed based on the magnitude of the decline in the bus frequency deviation and the change in the battery cluster temperature gradient within a preset response time after the execution of the partitioned flexible storage strategy.

[0007] Furthermore, the mode determination module includes: The occlusion rate calculation unit is used to calculate the rate of change of the occlusion coefficient of the sudden cloud cluster within a preset jump construction time. A power gradient calculation unit is used to calculate the gradient of the change in the output power of the partition within the preset jump construction time. A jump index construction unit is used to construct the photovoltaic jump index based on the rate of change and the gradient of change; The mode determination unit is used to determine whether to enter the jump absorption mode when the photovoltaic jump index is greater than the preset jump trigger threshold.

[0008] Furthermore, the strategy generation module includes: The oscillation feature extraction unit is used to construct a bus stability index curve based on the bus frequency deviation and the DC bus voltage fluctuation amplitude, and extract the oscillation amplitude features within a preset analysis time. A thermal stress construction unit is used to construct energy storage thermal stress indicators based on the temperature difference gradient of the battery cluster and the charge / discharge rate. An adjustment coefficient generation unit is used to generate a flexible adjustment coefficient based on the degree of difference between the oscillation amplitude characteristics and the energy storage thermal stress index. The strategy scaling unit is used to scale the preset participation ratio and the multiplier limit according to the flexibility adjustment coefficient to generate the partitioned flexible storage strategy.

[0009] Furthermore, the adjustment coefficient generation unit includes: The difference calculation subunit is used to calculate the deviation value after normalizing the oscillation amplitude characteristics and the energy storage thermal stress index respectively. A proportional mapping subunit is used to generate a proportional mapping factor based on the deviation value and a preset adjustment mapping range; The coefficient output subunit is used to output the flexible adjustment coefficient according to the proportional mapping factor.

[0010] Furthermore, the proportional mapping subunit is used to generate a corresponding nonlinear scaling factor based on the segmented adjustment interval where the deviation value is located, and to perform a smooth transition processing on the nonlinear scaling factor between adjacent segmented adjustment intervals to generate the proportional mapping factor, wherein the segmented adjustment interval is obtained by dividing the preset difference adjustment interval according to a preset segmented threshold.

[0011] Furthermore, the oscillation feature extraction unit includes: A peak detection subunit is used to identify local peaks of the bus stability index curve within the preset analysis time. The amplitude calculation subunit is used to calculate the oscillation amplitude between adjacent local peaks; The frequency statistics subunit is used to count the oscillation frequency per unit time. The feature output subunit is used to output the oscillation amplitude feature based on the oscillation amplitude and the oscillation frequency.

[0012] Furthermore, the thermal stress building unit includes: The temperature difference change rate calculation subunit is used to calculate the rate of change of the temperature difference gradient of the battery cluster within a preset thermal evaluation time. A rate offset calculation subunit is used to calculate the offset of the charge / discharge rate relative to the nominal rate; A stress fusion subunit is used to construct the energy storage thermal stress index based on the rate of change and the offset.

[0013] Furthermore, the policy execution module includes: A power allocation unit is used to allocate target charge and discharge power to several energy storage battery clusters according to the partitioned flexible storage strategy; A rate limiting unit is used to limit the charge / discharge rate of each energy storage battery cluster from exceeding the upper limit of the rate. A real-time feedback unit is used to provide real-time feedback on the bus frequency deviation and the battery cluster temperature gradient during the execution of the power-limited storage control.

[0014] Furthermore, the strategy adjustment module includes: The fall-off amplitude calculation unit is used to calculate the fall-off amplitude of the bus frequency deviation within the preset response time. Temperature difference change calculation unit, which is used to calculate the change in temperature difference gradient of the battery cluster; A correction index construction unit is used to construct the correction evaluation index based on the decline magnitude and the change amount; A threshold update unit is used to update the preset jump trigger threshold or the multiplier limit when the modified evaluation index does not meet the preset stability condition.

[0015] Furthermore, the threshold update unit includes: The trend analysis subunit is used to analyze the changing trend of the modified evaluation index within a continuous response period of a preset update quantity; A dynamic correction subunit is used to incrementally or incrementally correct the preset jump trigger threshold according to the changing trend. A multiplier compensation subunit is used to simultaneously adjust the upper limit of the multiplier while the preset jump trigger threshold is being corrected.

[0016] Compared with existing technologies, the advantages of this invention lie in the following: by coupling the gradient of the change in zoned output power with the rate of change of the sudden cloud shading coefficient to construct a photovoltaic jump index, external irradiation disturbances can be mapped to the system control layer in a dynamic rate form, thereby triggering the jump absorption mode in advance before a sudden drop or rise in power; simultaneously, by fusing the bus frequency deviation and the DC bus voltage fluctuation amplitude to construct a bus stability index, the AC side frequency imbalance and DC side voltage disturbance are given a unified quantitative characterization; and by combining the battery cluster temperature difference gradient to construct the energy storage carrying capacity coefficient, the energy storage participation ratio is constrained by both the system electrical stability requirements and the battery thermal state. Constraints are imposed to achieve a dynamic balance between power support capability and thermal safety boundary. After the strategy is executed, a correction evaluation index is constructed by the bus frequency drop amplitude and the battery cluster temperature difference change, so that the system response effect and the energy storage heat load change form a closed loop feedback, thereby adaptively correcting the preset jump trigger threshold and rate limit, avoiding secondary disturbances caused by overcharging or under-response. This effectively solves the problems of prediction error accumulation, system frequency easy secondary fluctuation or support response lag caused by the reliance on open-loop feedforward scheduling mode, lack of real-time evaluation of actual adjustment effect and dynamic threshold optimization.

[0017] Furthermore, by extracting the rate of change of cloud shading coefficient and the gradient of regional output power within the same jump construction time, and coupling the two to construct a photovoltaic jump index, the external driving factors of meteorological disturbances and the power response characteristics on the electrical side are kept consistent on the time scale. Rapid cloud movement causes instantaneous changes in irradiance, and the rate of change directly determines the trend of incident energy change per unit time. The regional output power gradient reflects the actual response amplitude of photovoltaic module current and voltage characteristics under sudden irradiance changes. When both rise synchronously, it indicates that the power jump is driven by a real sudden irradiance change rather than measurement noise or local shading interference. By comparing the jump index with a preset jump trigger threshold, potential instability trends can be identified in advance before the power disturbance is fully transmitted to the bus frequency and voltage levels. This enables rapid judgment of sudden power drops or rises, reduces the probability of false triggering, and improves the transient stability of islanded grid operation.

[0018] Furthermore, by coupling the bus frequency deviation with the DC bus voltage fluctuation amplitude to construct a bus stability index curve, and extracting oscillation amplitude characteristics within a unified analysis period, the intensity of electrical-side disturbances is quantitatively characterized. Simultaneously, the battery cluster temperature gradient and charge / discharge rate are used to construct an energy storage thermal stress index, establishing a correlation between the energy storage side's heat load level and power usage intensity. A larger bus oscillation amplitude indicates a higher degree of instantaneous power imbalance and a stronger demand for rapid energy storage output. Conversely, an increase in temperature gradient and charge / discharge rate signifies increased internal battery heat generation and incomplete heat diffusion; continued high-rate operation will further amplify the temperature difference. By comparing the differences between oscillation amplitude characteristics and thermal stress indices and generating flexible adjustment coefficients, the system can increase energy storage participation when the system stability margin is sufficient, while automatically suppressing rate increases when thermal stress approaches its upper limit. This establishes a dynamic balance between transient stability requirements and battery thermal safety constraints, ultimately achieving a zoned flexible storage strategy that both suppresses bus oscillations and controls heat accumulation.

[0019] Furthermore, by normalizing the oscillation amplitude characteristics and the energy storage thermal stress index respectively and then comparing the difference, the system's immediate stability requirements and the energy storage's thermal carrying capacity are dynamically balanced under the same dimension. When the oscillation is enhanced and the thermal stress is low, the deviation value increases, and the flexible adjustment coefficient is increased after proportional mapping, thereby expanding the participation ratio and the upper limit of the multiplier. When the thermal stress increases and the oscillation demand decreases, the deviation value decreases, and the adjustment coefficient converges accordingly, so that the output intensity decreases according to the degree of heat accumulation. This achieves the linkage adjustment between power support strength and temperature rise rate, avoiding heat accumulation runaway caused by continuous high-multiplier operation, while ensuring sufficient response capability when the bus is disturbed, thereby achieving a coordinated unity of stable support and life protection.

[0020] Furthermore, by comparing and normalizing the decline in bus frequency deviation with the preset frequency recovery reference amplitude, the actual support strength provided by the energy storage system for the stable recovery of the bus after its participation in regulation can be quantitatively reflected. A larger decline amplitude indicates more sufficient compensation of the system's power imbalance by the energy storage output power, and a stronger dynamic stability recovery capability. Simultaneously, by normalizing the change in battery cluster temperature gradient with the preset temperature difference safety reference change, the degree of internal heat accumulation and heat load change level of the battery during energy storage participation in regulation can be quantified. A larger temperature difference change indicates greater internal losses and heat generation caused by power regulation. The more significant the frequency recovery contribution, the smaller the battery thermal safety margin. By fusing the difference between the two to construct a correction evaluation index, a direct correlation is established between the frequency recovery effect and the battery thermal load change. When the frequency recovery contribution is large and the temperature difference change is small, the correction evaluation index increases, indicating that the system has the conditions to improve its regulation capability. Conversely, when the frequency recovery contribution is insufficient or the temperature difference change is too large, the correction evaluation index decreases, indicating that the current regulation intensity does not match the system or battery load capacity. This provides a quantitative basis for the dynamic update of the preset jump trigger threshold or rate limit, ensuring that the energy storage regulation capability is always coordinated with the system stability requirements and the battery thermal safety status.

[0021] Furthermore, by using the changing trend of the correction evaluation index sequence as the basis for correcting the preset jump trigger threshold, the adjustment direction of the preset jump trigger threshold is directly determined by the combined result of the energy storage regulation effect and the change in thermal state. When the correction evaluation index continues to rise, it indicates that the bus frequency recovery capability under unit heat load conditions is enhanced. At this time, the preset jump trigger threshold is decreased while the upper limit of the rate is increased simultaneously, enabling the energy storage system to participate in regulation earlier and have a higher power output capability under the same disturbance conditions, thereby accelerating the frequency deviation suppression speed. When the correction evaluation index continues to fall, it indicates that the frequency recovery benefit is weakened relative to the heat load cost. At this time, the rate is increased. By setting a preset jump trigger threshold and simultaneously reducing the upper limit of the rate, the triggering conditions for energy storage to participate in regulation become stricter and the power output intensity decreases, thereby suppressing further accumulation of thermal stress. At the same time, the adjustment range of the preset jump trigger threshold is controlled by the product of the preset threshold correction step size and the average increment or average decrease of the correction evaluation index, so that the threshold change is consistent with the rate of change of the system's stable income. Furthermore, by setting a preset rate compensation ratio, the change direction of the upper limit of the rate and the preset jump trigger threshold are kept synchronized, so that the power output capability and the trigger sensitivity form a coordinated adjustment relationship, thereby achieving dynamic adaptive matching between the energy storage regulation response capability and the battery thermal safety margin. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the photovoltaic energy storage system based on flexible control in this embodiment; Figure 2 This is a logic diagram for determining whether to enter the jump absorption mode by the mode determination unit in this embodiment. Figure 3 The determination logic diagram for updating the preset jump trigger threshold or multiplier limit of the threshold update unit in this embodiment; Figure 4 This is a schematic diagram of the threshold update unit in this embodiment. Detailed Implementation

[0023] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0024] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0025] Please see Figure 1 As shown, this is a schematic diagram of a photovoltaic energy storage system based on flexible control according to this embodiment. This embodiment provides a photovoltaic energy storage system based on flexible control, including: The data acquisition module is used to collect in real time the zoned output power, bus frequency deviation, DC bus voltage fluctuation amplitude, battery cluster temperature gradient and charge / discharge rate of several energy storage battery clusters, as well as the sudden cloud shading coefficient of the mountain photovoltaic array under isolated grid operation. The mode determination module is connected to the data acquisition module and is used to determine whether to enter the jump absorption mode based on the photovoltaic jump index and the preset jump trigger threshold. The photovoltaic jump index is constructed based on the rate of change of the sudden cloud shading coefficient and the gradient of change of the partition output power. The strategy generation module, which is connected to the mode determination module and the data acquisition module respectively, is used to determine the preset participation ratio of each energy storage cluster and the upper limit of the charge and discharge rate corresponding to the energy storage capacity coefficient and the matching result of the energy storage capacity coefficient and the bus stability index based on the determination result of entering the jump absorption mode, so as to form a partitioned flexible storage strategy. The energy storage capacity coefficient is determined based on the bus stability index and the temperature difference gradient of the battery cluster, and the bus stability index is constructed based on the bus frequency deviation and the DC bus voltage fluctuation amplitude. A strategy execution module, which is connected to the strategy generation module, is used to execute power-limited storage control based on the partitioned flexible storage strategy; The strategy adjustment module is connected to the strategy execution module, the data acquisition module, the mode determination module, and the strategy generation module, respectively. It is used to adjust the preset jump trigger threshold or the upper limit of the multiplier based on the correction evaluation index and the preset stability conditions. The correction evaluation index is constructed based on the magnitude of the decline in the bus frequency deviation and the change in the temperature gradient of the battery cluster within a preset response time after the execution of the partitioned flexible storage strategy.

[0026] In this embodiment, the system is applied to an isolated photovoltaic energy storage microgrid deployed in a remote mountainous area. It does not operate in parallel with the main power grid for extended periods and primarily provides continuous power to communication base stations, monitoring stations, and small loads in the mountainous region. The photovoltaic arrays are distributed across different slopes and altitudes, significantly affected by rapidly moving cloud formations in the mountains. The sunlight intensity fluctuates frequently within a short period, causing sudden jumps in output power. Multiple energy storage battery clusters are configured as the main energy storage units to maintain stable DC bus voltage and system frequency during rapid fluctuations in photovoltaic power. During operation, a dynamic balance control is required between energy storage capacity and battery thermal safety.

[0027] In this embodiment, the data acquisition module is responsible for real-time acquisition and preprocessing of key quantities used for control: the zone output power is calculated by installing current and voltage sensors on the DC or AC side of the inverter in each photovoltaic zone and using a local power meter. The current sensors used are Hall effect or shunt-type current sensors, and the voltage is sampled by an isolated voltage transmitter and acquired by a 16-bit ADC; the bus frequency deviation is calculated in real-time on the AC side by the inverter or an independent frequency measuring device using a phase-locked loop (PLL) algorithm and output with 1Hz as the reference; the DC bus voltage fluctuation amplitude is directly sampled by the DC voltage transmitter and obtained by bandpass filtering and envelope analysis to obtain the short-time fluctuation amplitude; the battery cluster temperature gradient is calculated by deploying multiple temperature sensing points at the inlet, middle, and end of each battery cluster, and measuring the maximum and minimum temperature difference and gradient using an NTC or digital temperature sensor; the charge / discharge rate is the ratio of the battery cluster current to the cluster's rated capacity. The calculation involves real-time acquisition of current by a Hall sensor, and the cluster rated capacity and SOC provided by the coulomb count and open-circuit voltage correction of the battery management system (BMS). The sudden cloud shading coefficient is obtained by measuring the global and regional instantaneous irradiance using a field flat-panel irradiance meter or a platinum sinusoidal response light sensor, combined with the temperature change rate of the array backplane temperature sensor, and then differentially processed and low-pass filtered to obtain the irradiance reduction rate index. Several indices are linearly combined according to preset weights to generate the shading coefficient. All sampling points are equipped with signal conditioning circuits for anti-aliasing filtering and outlier removal. The sampling frequency is recommended to be 1Hz, with a range of 0.5 to 2Hz. The acquired data is aggregated by the field RTU or BMS and then sent to the main control unit or SCADA system via RS485 Modbus or Ethernet ModbusTCP. The main control unit performs a moving average or median filter on the raw data and outputs normalized feature quantities for mode determination for use by subsequent modules.

[0028] The preset jump trigger threshold is a critical value of the photovoltaic jump index used to determine whether the system has entered the jump absorption mode. Its setting is based on the following technical factors: First, statistical analysis is performed on the output power data of each region for at least 30 consecutive days at a 1-second sampling period. The jump index sequence, composed of the power change gradient and the rate of change of the cloud shading coefficient, is calculated, and its mean, standard deviation, and 95th percentile are extracted. Second, the maximum deviation of the bus frequency and the voltage fluctuation amplitude of the isolated grid system under different power disturbance amplitudes are measured to establish a correlation curve between the jump index and the risk of bus instability. Third, the jump boundary that the system can safely absorb is determined by combining the inverter's maximum permissible transient overload capacity and the maximum instantaneous charge / discharge rate limit of the energy storage. Typically, the threshold is set within the range of the statistical jump index mean plus 1.5 times the standard deviation to the 95th percentile, generally between 0.6 and 1.2. In this embodiment, the mean value of the jump index obtained through field testing is 0.52, the standard deviation is 0.22, and the 95th percentile value is 0.91. Considering the constraint of the maximum allowable frequency deviation of 0.3 Hz for the bus, the preset jump trigger threshold is set to 0.85. This can trigger energy storage regulation in advance before the frequency deviation exceeds the allowable limit, while avoiding misjudgment of normal small irradiation fluctuations.

[0029] The preset participation ratio refers to the proportion of the rated capacity of the energy storage battery clusters participating in power regulation to the total available capacity when entering the jump absorption mode. Its setting is based on several factors: first, the dispersion of the current state of charge (SOC) distribution of each battery cluster, assessed by calculating the SOC standard deviation to evaluate capacity dispatchability balance; second, the coupling relationship between the battery cluster temperature gradient and the charge / discharge rate, as measured by a thermal model, where sustained high-rate operation significantly accelerates aging when the temperature gradient exceeds 6 degrees Celsius; and third, the functional relationship between the bus stability index and the required instantaneous regulation power, obtained through step disturbance tests to obtain the minimum required participation capacity ratio curve. Typically, the participation capacity ratio is calculated by inversely proposing the ratio of the required instantaneous regulation power to the photovoltaic rated power, and is set between 0.3 and 0.8. In this embodiment, the photovoltaic rated capacity is 500 kW, with a typical jump amplitude of 180 kW. Testing showed that when the participation capacity ratio is not less than 0.55, the frequency deviation can be controlled within 0.2 Hz. Considering thermal safety margin, the preset participation ratio is ultimately set to 0.6, which can control battery heat accumulation while meeting transient stability requirements.

[0030] The preset response time refers to the observation time window used to evaluate the bus frequency deviation drop and battery cluster temperature difference change after implementing the partitioned flexible storage strategy. Its setting is based on: first, the equivalent inertia constant and damping coefficient of the isolated grid are measured through system small-disturbance tests, and the frequency recovery time constant is calculated accordingly; second, the equilibrium time of the temperature difference between the cell surface and interior is measured through battery thermal diffusion experiments at 1x charge / discharge rate; and third, the inverter control bandwidth and sampling period are combined to ensure that at least three control adjustment cycles are covered within the evaluation window. Typically, the response time is set to 2 to 4 times the frequency time constant, and not less than 0.5 times the thermal diffusion time constant, generally between 10 and 60 seconds. In this embodiment, the measured frequency time constant is 8 seconds, and the battery thermal diffusion time constant is 40 seconds. After considering these values, the preset response time is set to 30 seconds, which can fully cover the frequency drop process after a typical power jump, while also reflecting the initial trend of battery temperature difference change, providing sufficient dynamic data support for correcting the evaluation index.

[0031] By coupling the gradient of output power variation in different zones with the rate of change of sudden cloud shading coefficients, a photovoltaic jump index is constructed. This allows external irradiance disturbances to be mapped to the system control layer at a dynamic rate, thus triggering the jump absorption mode in advance before power surges or drops. Simultaneously, a bus stability index is constructed by integrating bus frequency deviation and DC bus voltage fluctuation amplitude. This provides a unified quantitative representation of AC-side frequency imbalance and DC-side voltage disturbances. Furthermore, combining this with the battery cluster temperature gradient constructs an energy storage carrying capacity coefficient, ensuring that the energy storage participation ratio is constrained by both system electrical stability requirements and battery thermal state, thereby achieving power support. The dynamic balance between capacity and thermal safety boundary; after strategy execution, a correction evaluation index is constructed by the bus frequency drop amplitude and the battery cluster temperature difference change, so that the system response effect and the energy storage heat load change form a closed loop feedback, thereby adaptively correcting the preset jump trigger threshold and rate limit, avoiding secondary disturbances caused by overcharging or under-response, effectively solving the problems of prediction error accumulation, system frequency easy secondary fluctuation or support response lag caused by the reliance on open-loop feedforward scheduling mode, lack of real-time evaluation of actual adjustment effect and threshold dynamic optimization.

[0032] Please see Figure 2 As shown, this is the logic diagram for determining whether to enter the jump absorption mode by the mode determination unit in this embodiment. In this embodiment, the mode determination module includes: The occlusion rate calculation unit is used to calculate the rate of change of the occlusion coefficient of the sudden cloud cluster within a preset jump construction time. A power gradient calculation unit is used to calculate the gradient of the change in the output power of the partition within the preset jump construction time. A jump index construction unit is connected to the shading rate calculation unit and the power gradient calculation unit respectively, and is used to construct the photovoltaic jump index according to the rate of change and the gradient of change; The mode determination unit is used to determine whether to enter the jump absorption mode when the photovoltaic jump index is greater than the preset jump trigger threshold.

[0033] In this embodiment, the system operates on a fixed sampling period Δt. At the k-th sampling time tk, the sudden cloud occlusion coefficient is Ck, and the partitioned output power is Pk. The preset transition construction time is T1, and the corresponding number of sampling points is N = T1 / Δt. The occlusion speed calculation unit calculates the occlusion speed Vc(k) over N consecutive sampling points: Vc(k) = (Ck) / ( ... Ck N) / T1, where Ck N is the occlusion coefficient of the sudden cloud cluster corresponding to the initial sampling point; the power gradient calculation unit calculates the gradient of the output power of the partition within the preset jump construction time within the same time window, Gp(k), Gp(k) = (Pk) Pk N) / T1, where Pk N represents the output power of the partition corresponding to the initial sampling point.

[0034] The photovoltaic jump index construction unit is constructed according to the following formula: J(k) = |Vc(k)| / V0×|Gp(k)| / Gp0, where V0 is the preset shading speed threshold and Gp0 is the preset change gradient threshold.

[0035] The preset shading rate threshold is a reference value for the rate of change of the shading coefficient that can be considered as constituting a valid cloud cluster rapid shading event per unit time. It depends on the accuracy of the on-site irradiance sensor, the statistical distribution of the rate of change of the shading coefficient within the historical operating cycle, and the minimum identifiable disturbance amplitude allowed by the system. It is usually determined by extracting the 90% to 99% percentile value of the shading rate from sampling data for no less than 30 consecutive days. It is generally set between 0.02 seconds and 0.15 seconds. In this embodiment, it is set to 0.08 seconds, which can filter out slow illumination changes and measurement noise, and only normalize the benchmark calibration for significant irradiance changes caused by rapid cloud cluster shading.

[0036] The preset gradient threshold is a reference value for the rate of change of output power that constitutes an effective power jump per unit time. It depends on the rated capacity of the photovoltaic zone, the dynamic response capability of the inverter, and the upper limit of the allowed frequency deviation of the islanded grid. It is usually set according to 80% to 100% of the maximum safe ramp rate allowed under rated power or the upper limit of historical power gradient statistics. It is generally set between 0.01 times the rated power per second and 0.10 times the rated power per second. In this embodiment, it is set to 0.05 times the rated power per second, which can make the jump index sensitive to the actual power change and avoid misjudgment caused by normal power fine-tuning.

[0037] By extracting the rate of change of cloud shading coefficient and the gradient of regional output power within the same jump construction time, and coupling the two to construct a photovoltaic jump index, the external driving factors of meteorological disturbances and the power response characteristics on the electrical side are kept consistent on the time scale. Rapid cloud movement causes instantaneous changes in irradiance, and the rate of change directly determines the trend of incident energy change per unit time. The regional output power gradient reflects the actual response amplitude of photovoltaic module current and voltage characteristics under sudden irradiance changes. When both rise synchronously, it indicates that the power jump is driven by a real sudden irradiance change rather than measurement noise or local shading interference. By comparing the jump index with a preset jump trigger threshold, potential instability trends can be identified in advance before the power disturbance is fully transmitted to the bus frequency and voltage levels. This enables rapid judgment of sudden power drops or rises, reduces the probability of false triggers, and improves the transient stability of islanded grid operation.

[0038] Specifically, the strategy generation module includes: The oscillation feature extraction unit is used to construct a bus stability index curve based on the bus frequency deviation and the DC bus voltage fluctuation amplitude, and extract the oscillation amplitude features within a preset analysis time. A thermal stress construction unit is used to construct energy storage thermal stress indicators based on the temperature difference gradient of the battery cluster and the charge / discharge rate. An adjustment coefficient generation unit is connected to an oscillation feature extraction unit and a thermal stress construction unit, respectively, to generate a flexible adjustment coefficient based on the degree of difference between the oscillation amplitude feature and the energy storage thermal stress index. A strategy scaling unit, connected to an adjustment coefficient generation unit, is used to scale the preset participation ratio and the multiplier limit according to the flexible adjustment coefficient to generate the partitioned flexible storage strategy.

[0039] The preset analysis duration is the length of the time window used to extract the oscillation amplitude characteristics of the bus stability index. It depends on the equivalent inertial time constant of the islanded network system, the inverter control bandwidth, and the frequency decay period after a power disturbance. It is usually set to 2 to 4 times the measured frequency recovery time constant, and is generally set between 10 and 40 seconds. In this embodiment, it is set to 20 seconds, which can cover the complete oscillation decay process caused by a typical power jump, thereby ensuring that the extracted oscillation amplitude characteristics are representative and stable.

[0040] By coupling the bus frequency deviation with the DC bus voltage fluctuation amplitude to construct a bus stability index curve, and extracting oscillation amplitude characteristics within a unified analysis period, the intensity of electrical-side disturbances can be quantitatively characterized. Simultaneously, the battery cluster temperature gradient and charge / discharge rate are used to construct an energy storage thermal stress index, establishing a correlation between the energy storage side's heat load level and power usage intensity. A larger bus oscillation amplitude indicates a higher degree of instantaneous power imbalance and a stronger demand for rapid energy storage output. Conversely, an increase in temperature gradient and charge / discharge rate signifies increased internal battery heat generation and incomplete heat diffusion; continued high-rate operation will further amplify the temperature difference. By comparing the differences between oscillation amplitude characteristics and the thermal stress index and generating a flexible adjustment coefficient, the system can increase energy storage participation when the system stability margin is sufficient, while automatically suppressing rate increases when thermal stress approaches its upper limit. This establishes a dynamic balance between transient stability requirements and battery thermal safety constraints, ultimately achieving a zoned flexible storage strategy that both suppresses bus oscillations and controls heat accumulation.

[0041] Specifically, the oscillation feature extraction unit includes: A peak detection subunit is used to identify local peaks of the bus stability index curve within the preset analysis time. An amplitude calculation subunit, connected to a peak detection subunit, is used to calculate the oscillation amplitude between adjacent local peaks; The frequency statistics subunit is used to count the oscillation frequency per unit time. The feature output subunit is connected to the amplitude calculation subunit and the frequency statistics subunit respectively, and is used to output the oscillation amplitude feature according to the oscillation amplitude and the oscillation frequency.

[0042] In this embodiment, the bus stability index curve is denoted as S(k), which is obtained by normalizing the bus frequency deviation Δf(k) and the DC bus voltage fluctuation amplitude ΔU(k) ​​at the kth sampling time and then linearly combining them. It is expressed as S(k) = a × |Δf(k)| / Δf0 + b × |ΔU(k)| / ΔU0, where Δf0 is the preset maximum frequency deviation reference value, ΔU0 is the preset DC voltage fluctuation reference value, and a and b are preset stability weighting coefficients and a + b = 1.

[0043] Within the preset analysis duration T2, the corresponding number of sampling points is M = T2 / Δt. The peak detection subunit performs extreme value judgment on S(k) within M consecutive sampling points. When S(k) > S(k) 1) When S(k) > S(k+1), it is determined to be a local peak point, and its amplitude S(pi) is recorded, where pi is the sampling number of the i-th peak.

[0044] The amplitude calculation subunit calculates the oscillation amplitude Ai for two adjacent peaks S(pi) and S(pi+1), defined as Ai = |S(pi+1) If m peaks are detected within a preset analysis time, then m is obtained. One oscillation amplitude; The frequency statistics subunit calculates the oscillation frequency F based on the number of peak values, expressed as F=m / T2; The feature output subunit constructs the oscillation amplitude feature Aosc(k) based on the oscillation amplitude and oscillation frequency, defined as Aosc(k) = (1 / (m... 1))×ΣAi×F, when m is less than 2, Aosc(k) takes the value of 0.

[0045] The preset maximum frequency deviation reference value depends on the safe frequency operating range allowed by the islanded system, the primary frequency regulation response capability, and the load's frequency sensitivity. It is usually determined based on the frequency control standards of grid-connected or islanded systems and historical operating data statistics, and is generally set between 0.2Hz and 1.0Hz. In this embodiment, it is set to 0.5Hz, which can cover the typical disturbance range while avoiding excessive exponential amplification due to small frequency fluctuations.

[0046] The preset DC voltage fluctuation reference value depends on the rated voltage level of the DC bus, the allowable voltage deviation range of the inverter, and the safe operating range of the energy storage interface converter. It is usually determined based on the allowable fluctuation percentage of the rated DC bus voltage and the statistical results of voltage fluctuations in historical operation. It is generally set between 2% and 10% of the rated DC voltage. In this embodiment, the rated DC bus voltage is 750V, and the preset DC voltage fluctuation reference value is set to 30V, which can cover the voltage disturbance amplitude under typical power jumps, while avoiding the excessive amplification of the stability index due to small voltage ripples.

[0047] The preset stability weighting coefficient depends on the relative sensitivity of the system frequency stability margin and the DC bus voltage stability margin, the inverter control structure, and the statistical results of historical oscillation contribution. It is usually allocated proportionally according to the degree of influence of the two types of indicators on system stability. Generally, each is set between 0.3 and 0.7 and a+b=1. In this embodiment, a=0.6 and b=0.4 are set, which can make the bus stability index more sensitive to frequency disturbances, while taking into account the impact of DC side voltage fluctuations.

[0048] Specifically, the thermal stress building unit includes: The temperature difference change rate calculation subunit is used to calculate the rate of change of the temperature difference gradient of the battery cluster within a preset thermal evaluation time. A rate offset calculation subunit is used to calculate the offset of the charge / discharge rate relative to the nominal rate; The stress fusion subunit is connected to the temperature difference change rate calculation subunit and the multiplier offset calculation subunit, respectively, to construct the energy storage thermal stress index based on the change rate and the offset.

[0049] In this embodiment, within the preset thermal evaluation time T3, the corresponding number of sampling points is L=T3 / Δt. The difference between the highest and lowest temperatures of the battery cluster at the kth sampling time is denoted as ΔT(k), the charge / discharge rate is denoted as Cr(k), and the nominal rate is Cr0.

[0050] The temperature difference change rate calculation subunit calculates the temperature difference change rate Rt(k) over L consecutive sampling points within a preset thermal assessment time T3, defined as Rt(k) = (ΔT(k)). ΔT(k) L)) / T3 is used to characterize the rate of evolution of the degree of thermal imbalance inside the battery cluster per unit time; The magnification offset calculation subunit calculates the magnification offset Rc(k), defined as Rc(k) = |Cr(k)|. Cr0| / Cr0 is used to characterize the degree of deviation of the current charge and discharge conditions from the rated thermal design conditions; The stress fusion subunit constructs an energy storage thermal stress index H(k) based on the temperature difference change rate and the rate offset, defined as H(k) = |Rt(k)| / R0×Rc(k), where R0 is a preset temperature difference change rate reference value, used to normalize and calibrate the thermal gradient evolution rate, thereby forming a comprehensive thermal stress characterization quantity that simultaneously reflects the degree of uneven thermal diffusion and the rate loading intensity.

[0051] The preset temperature difference change rate reference value depends on the thermal diffusion time constant of the battery cell, the heat dissipation capacity of the battery cluster structure, and the statistical upper limit of the temperature difference evolution rate in long-term operating data. It is usually determined by extracting the 90% to 99th percentile value of the temperature difference change rate from continuous operating data of no less than 30 days, and then correcting it in combination with the maximum allowable temperature rise rate of the battery. It is generally set between 0.05 degrees Celsius per second and 0.50 degrees Celsius per second. In this embodiment, it is set to 0.20 degrees Celsius per second, which can ensure sensitivity to abnormal heat accumulation while avoiding the amplification of thermal stress index due to slow and uniform temperature rise.

[0052] The nominal rate refers to the reference value of the charge-discharge rate that a battery cluster can operate stably for a long time under rated design conditions. It depends on the rated rate parameters of the battery type, the continuous heat dissipation capacity of the battery thermal management system, and the battery life design target. It is also determined by combining the rated continuous operating rate provided by the battery manufacturer and the statistical results of the system's long-term steady-state operation conditions. It is usually set at 0.8 to 1.0 times the rated continuous operating rate, generally between 0.5C and 1.5C. In this embodiment, it is set to 1.0C, which can serve as a benchmark to measure the degree to which the current charge-discharge rate deviates from the standard thermal design conditions, thereby accurately reflecting the impact of rate changes on the trend of battery thermal stress accumulation.

[0053] Specifically, the adjustment coefficient generation unit includes: The difference calculation subunit is used to calculate the deviation value D(k) after normalizing the oscillation amplitude characteristics and the energy storage thermal stress index respectively. D(k) = Aosc(k) / A0-H(k) / H0, where A0 is a preset amplitude characteristic reference value and H0 is a preset thermal stress index reference value. A proportional mapping subunit, which is connected to the difference calculation subunit, is used to generate a proportional mapping factor based on the deviation value and a preset adjustment mapping range; The coefficient output subunit, connected to the proportional mapping subunit, is used to output the flexible adjustment coefficient Kf(k) according to the proportional mapping factor, where Kf(k) = Kmin + μ(k) × (Kmax) Kmin), where Kmin is the upper limit of the preset adjustment coefficient, Kmax is the lower limit of the preset adjustment coefficient, and μ(k) is the proportional mapping factor.

[0054] The upper limit of the preset regulation coefficient depends on the maximum allowable participation ratio of the energy storage system, the instantaneous overload capacity of the inverter, and the upper limit of the battery safety rate. It is also constrained by the maximum support strength requirements of the bus stability control. It is usually set according to 80% to 100% of the rated participation ratio, and is generally set between 0.8 and 1.0. In this embodiment, it is set to 1.0, which can release the full regulation capacity when the system oscillation demand is significant and the thermal stress is within a controllable range.

[0055] The lower limit of the preset adjustment coefficient depends on the minimum participation ratio required for the energy storage system to maintain basic steady-state support, the battery thermal protection threshold, and the life decay control requirements. It is usually set according to 10% to 40% of the rated participation ratio, and is generally set between 0.1 and 0.4. In this embodiment, it is set to 0.2, which can significantly reduce the participation intensity when thermal stress increases, while maintaining the necessary bus support capacity.

[0056] Specifically, the proportional mapping subunit is used to generate a corresponding nonlinear scaling factor based on the segmented adjustment interval where the deviation value is located, and to perform a smooth transition processing on the nonlinear scaling factor between adjacent segmented adjustment intervals to generate the proportional mapping factor, wherein the segmented adjustment interval is obtained by dividing the preset difference adjustment interval according to a preset segmented threshold.

[0057] In this embodiment, the proportional mapping subunit performs a linear mapping based on the deviation value D(k) and the preset adjustment mapping range Dmin to Dmax to obtain the proportional mapping factor μ(k) = (D(k)). Dmin) / (Dmax) Dmin), and μ(k) is clipped to be between 0 and 1.

[0058] The preset adjustment mapping interval [Dmin, Dmax] depends on the normalized difference distribution range of the oscillation amplitude characteristics and energy storage thermal stress index in historical operating data, the maximum allowable adjustment slope of the system, and the need to avoid secondary disturbances to the bus stability caused by abrupt changes in the adjustment coefficient. It is usually selected based on the difference statistical results, covering a symmetrical interval of 90% to 95% of the operating samples, and is generally set at […]. Between 1.0, 1.0], in this embodiment it is set to [ [0.8, 0.8] can retain the boundary buffer band while covering the main operating conditions, so that the scaling factor can be smoothly saturated under extreme differences, avoiding sudden changes in adjustment intensity.

[0059] By normalizing the oscillation amplitude characteristics and the energy storage thermal stress index respectively and then comparing the difference, the system's immediate stability requirements and the energy storage's thermal carrying capacity are dynamically balanced under the same dimension. When the oscillation is enhanced and the thermal stress is low, the deviation value increases, and the flexible adjustment coefficient is increased after proportional mapping, thereby expanding the participation ratio and the upper limit of the multiplier. When the thermal stress increases and the oscillation demand decreases, the deviation value decreases, and the adjustment coefficient converges accordingly, so that the output intensity decreases according to the degree of heat accumulation. This achieves the linkage adjustment between power support strength and temperature rise rate, avoids heat accumulation runaway caused by continuous high-multiplier operation, and ensures sufficient response capability when the bus is disturbed, thereby achieving a coordinated unity of stable support and life protection.

[0060] Specifically, the policy execution module includes: A power allocation unit is used to allocate target charge and discharge power to several energy storage battery clusters according to the partitioned flexible storage strategy; A rate limiting unit is used to limit the charge / discharge rate of each energy storage battery cluster from exceeding the upper limit of the rate. A real-time feedback unit is used to provide real-time feedback on the bus frequency deviation and the battery cluster temperature gradient during the execution of the power-limited storage control.

[0061] In this embodiment, the power allocation unit is used to allocate target charge and discharge power to several energy storage battery clusters according to the partitioned flexible storage strategy and the corresponding flexible adjustment coefficient, so that the target charge and discharge power Pi(k) of the i-th energy storage battery cluster satisfies Pi(k) = Kf(k) × Pref(k) × Wi(k), where Pref(k) is the target adjustment power corresponding to the partitioned flexible storage strategy, and Wi(k) is the available adjustment margin weight of the i-th energy storage battery cluster; the rate limiting unit is used to limit the target charge and discharge power according to the rated capacity Ci of each energy storage battery cluster and the preset rate limit Crmax. The charging and discharging power is constrained by a rate to ensure that the actual charging and discharging rate of each energy storage battery cluster, Cri(k) = |Pi(k)| / Ci, does not exceed the rate limit Crmax. When Cri(k) exceeds the rate limit Crmax, the target charging and discharging power Pi(k) is limited and adjusted. The real-time feedback unit is used to provide real-time feedback on the bus frequency deviation and battery cluster temperature gradient during the power-limited storage control process. The bus frequency deviation and battery cluster temperature gradient are sent to the oscillation feature extraction unit and the thermal stress construction unit, respectively, to update the oscillation amplitude characteristics and energy storage thermal stress indicators.

[0062] The target regulation power refers to the reference value of the total regulation power used for absorption or release determined by the partitioned flexible storage strategy at the current sampling time. It depends on the magnitude of the partitioned output power deviation caused by sudden cloud cover, the degree of bus frequency deviation, and the upper limit of the maximum allowable regulation capacity of the system. It is usually set between 0 times the rated energy storage power and 1.0 times the rated energy storage power. In this embodiment, it is set to the smaller value between the real-time power deviation and the upper limit of the rated energy storage power, which enables the energy storage system output power to dynamically match the actual power gap and avoid new bus fluctuations caused by insufficient or excessive regulation.

[0063] The available adjustment margin weight refers to a dimensionless weighting coefficient that reflects the relative proportion of the power regulation capability of the i-th energy storage battery cluster under the current operating state. It depends on the combined availability of the state of charge margin and thermal operation margin of the energy storage battery cluster. Specifically, firstly, the current state of charge (SOC) of the i-th energy storage battery cluster is obtained, and the SOC margin level is determined based on its distance from the upper and lower limits of the preset safe operating range. The further the SOC is from the limit boundary, the higher the corresponding SOC margin. Simultaneously, the current temperature gradient of the i-th energy storage battery cluster is obtained, and the thermal operating margin level is determined based on its proximity to the preset safe temperature difference threshold. The smaller the temperature gradient, the higher the corresponding thermal operating margin. Subsequently, a weighted comprehensive evaluation is performed based on the SOC margin level and the thermal operating margin level, and the evaluation result is converted into a weight value between 0 and 1 according to a preset normalization range. This weight is used as the available adjustment margin weight of the energy storage battery cluster. The more secure the operating state and the larger the adjustment margin of the energy storage battery cluster, the closer the available adjustment margin weight is to 1. The more secure the operating state and the larger the adjustment margin of the energy storage battery cluster, the closer the available adjustment margin weight is to 0. This allows the regulation power to be preferentially allocated to the energy storage battery cluster with stronger regulation capability and lower operating risk.

[0064] The upper limit of the charge / discharge rate refers to the reference value that limits the maximum charge / discharge rate allowed per unit capacity of each energy storage battery cluster. It is used to constrain the maximum power output intensity of the energy storage battery cluster during the adjustment process. It depends on the rated rate parameters of the battery type, battery thermal safety limits and cycle life protection requirements, and is usually set between 0.5C and 2.0C. In this embodiment, it is set to 1.0C, which can ensure that the energy storage system has a fast power response capability while preventing the battery temperature rise and lifespan degradation caused by excessively high charge / discharge rates.

[0065] Please see Figure 3 As shown, this is the logic diagram for determining the threshold update unit to update the preset jump trigger threshold or the upper limit of the multiplier in this embodiment. In this embodiment, the strategy adjustment module includes: The fall-off amplitude calculation unit is used to calculate the fall-off amplitude of the bus frequency deviation within the preset response time. Temperature difference change calculation unit, which is used to calculate the change in temperature difference gradient of the battery cluster; A correction index construction unit is connected to the fall amplitude calculation unit and the temperature difference change calculation unit, respectively, to construct the correction evaluation index based on the fall amplitude and the change amount; A threshold update unit, which is connected to the correction index construction unit, is used to update the preset jump trigger threshold or the multiplier limit when the correction evaluation index does not meet the preset stability condition.

[0066] In this embodiment, the fallback amplitude calculation unit is used to obtain the bus frequency deviation sequence within a preset response time, and identify the maximum absolute value of the bus frequency deviation within the time window and the absolute value of the bus frequency deviation at the current sampling time. The difference between the two is determined as the fallback amplitude of the bus frequency deviation, so as to reflect the degree of bus frequency recovery stability after the energy storage system participates in the adjustment. The larger the fallback amplitude, the more obvious the bus frequency recovery effect.

[0067] The temperature difference change calculation unit is used to obtain the battery cluster temperature difference gradient sequence within a preset response time, and the difference between the battery cluster temperature difference gradient at the current sampling time and the battery cluster temperature difference gradient at the beginning of the time window is determined as the temperature difference change, so as to reflect the degree of change of battery thermal state during the energy storage system's participation in regulation. The larger the temperature difference change, the more significant the increase in heat load brought about by the energy storage participation in regulation.

[0068] The correction index construction unit is used to construct a correction evaluation index based on the relative relationship between the decline magnitude and the temperature difference change. Specifically, the decline magnitude is normalized to the preset frequency recovery reference magnitude, and the temperature difference change is normalized to the preset temperature difference safety reference change. The difference between the two is then calculated to construct the correction evaluation index. This allows the correction evaluation index to comprehensively reflect the balance between the contribution of energy storage regulation to the stable recovery of the bus and the impact on the thermal state of the battery.

[0069] The preset frequency recovery reference amplitude refers to the minimum frequency deviation drop value used to characterize the effective and stable recovery of the bus frequency after the energy storage system participates in power regulation. It depends on the rated operating frequency of the islanded grid system, the upper limit of the allowable frequency deviation range, and the frequency support control accuracy of the inverter. It is also determined by combining the statistical results of the typical drop amplitude corresponding to the stable recovery after frequency disturbance in historical operating data. It is usually set according to the deviation range of 0.02 to 0.10 times the rated frequency, and is generally set between 0.01 Hz and 0.05 Hz. In this embodiment, it is set to 0.02 Hz, which can serve as a normalized reference benchmark for determining whether the energy storage regulation has an effective and stable recovery effect on the bus frequency.

[0070] The preset temperature difference safety reference change refers to the safe reference range of the allowable change in the temperature difference gradient of the battery cluster during the power regulation process of the energy storage system. It is used to characterize the maximum temperature difference change that the battery thermal state is still within a controllable range. It depends on the maximum allowable temperature difference range of the battery type, the heat dissipation capacity of the battery thermal management system, and the long-term operating life protection requirements of the battery. It is also determined by combining the safe temperature difference change limit provided by the battery manufacturer and the statistical results of temperature difference change when no thermal imbalance risk occurred in historical operation. It is usually set according to 0.10 to 0.30 times the maximum allowable temperature difference, and is generally set between 0.5℃ and 2.0℃. In this embodiment, it is set to 1.0℃, which can serve as a normalized reference benchmark for judging whether the thermal safety margin is sufficient during the energy storage regulation process.

[0071] By comparing and normalizing the decline in bus frequency deviation with the preset frequency recovery reference amplitude, the actual support strength provided by the energy storage system to the stable recovery of the bus can be quantitatively reflected. A larger decline amplitude indicates more sufficient compensation of system power imbalance by the energy storage output power, and a stronger dynamic stability recovery capability of the system. Simultaneously, by normalizing the change in battery cluster temperature gradient with the preset temperature difference safety reference change, the degree of internal heat accumulation and heat load change level of the battery during energy storage participation in regulation can be quantified. A larger temperature difference change indicates more significant internal losses and heat generation caused by power regulation. The smaller the battery thermal safety margin, the better. By fusing the difference between the two to construct a correction evaluation index, a direct correlation is established between the frequency recovery effect and the battery thermal load change. When the frequency recovery contribution is large and the temperature difference change is small, the correction evaluation index increases, indicating that the system has the conditions to improve its regulation capability. When the frequency recovery contribution is insufficient or the temperature difference change is too large, the correction evaluation index decreases, indicating that the current regulation intensity does not match the system or battery load capacity. This provides a quantitative basis for the dynamic update of the preset jump trigger threshold or rate limit, ensuring that the energy storage regulation capability is always coordinated with the system stability requirements and the battery thermal safety status.

[0072] Please see Figure 4 As shown, this is a schematic diagram of the threshold update unit in this embodiment. In this embodiment, the threshold update unit includes: The trend analysis subunit is used to analyze the changing trend of the modified evaluation index within a continuous response period of a preset update quantity; A dynamic correction subunit, which is connected to the trend analysis subunit, is used to incrementally or incrementally correct the preset jump trigger threshold according to the changing trend. A rate compensation subunit, which is connected to a dynamic correction subunit, is used to simultaneously adjust the upper limit of the rate while the preset jump trigger threshold is being corrected.

[0073] In this embodiment, the trend analysis subunit is used to obtain the corresponding corrected evaluation index sequence Rj within a preset number of consecutive response cycles, where j is the response cycle number, and to determine the trend based on the direction and magnitude of change of the corrected evaluation index sequence; specifically, when the preset stability condition is met: Rj is greater than Rj An upward trend is defined as the percentage of occurrences of 1 being greater than a preset threshold for the percentage of increases. This is further defined as the condition that Rj is less than Rj. When the percentage of occurrences of 1 exceeds a preset threshold for the percentage of decrease, it is determined to be a downward trend; otherwise, it is determined to be a stable trend, in order to reflect the adaptation and change state between energy storage regulation capacity and system stability requirements.

[0074] The preset update quantity is the number of consecutive response cycles required to determine the trend of the modified evaluation index. Its setting depends on the average frequency of power disturbances in the isolated grid system under typical operating conditions. The more frequent the disturbances, the shorter the statistical window needs to be to maintain the sensitivity of the trend response. It also depends on the duration of a single response cycle. In this embodiment, a single response cycle is equal to the preset response duration. The longer the cycle, the longer the total duration of the statistical window. A trade-off needs to be made between the timeliness of trend identification and the reliability of statistics. Finally, it depends on the fluctuation characteristics of the modified evaluation index sequence itself, including the mean, standard deviation, and autocorrelation. The greater the fluctuation, the more samples need to be added to filter out random interference and ensure the confidence of the trend determination. Considering the above factors, the preset update quantity is usually set between 5 and 20. In this embodiment, the average interval of disturbances is about 40 seconds, and the single response cycle is 30 seconds. The measured short-term fluctuation standard deviation of the modified evaluation index is about 0.12. In order to complete a trend determination within 3 to 5 minutes and take into account statistical stability, the preset update quantity is set to 10, which can effectively filter out the interference of single random fluctuations and identify the continuous trend of energy storage regulation effect in a timely manner.

[0075] The dynamic correction subunit is used to decrease the preset jump trigger threshold when an upward trend is detected, so that the updated preset jump trigger threshold is the current preset jump trigger threshold minus the product of the preset threshold correction step size and the average increment of the correction evaluation index, thereby improving the system's response sensitivity to power jump events; when a downward trend is detected, the preset jump trigger threshold is increased, so that the updated preset jump trigger threshold is the current preset jump trigger threshold plus the product of the preset threshold correction step size and the average decrease of the correction evaluation index, thereby reducing the trigger frequency of the energy storage system's participation in regulation and thus avoiding excessive battery participation in regulation and heat load accumulation; when a stable trend is detected, the preset jump trigger threshold remains unchanged, so that the system maintains the stability of the current regulation strategy.

[0076] The rate compensation subunit is used to simultaneously adjust the upper limit of the rate according to the direction of change of the preset jump trigger threshold while completing the correction. When the preset jump trigger threshold decreases, the upper limit of the rate is increased according to the preset rate compensation ratio, so that the energy storage system has a higher instantaneous power support capability. When the preset jump trigger threshold increases, the upper limit of the rate is decreased according to the preset rate compensation ratio to limit the battery power output intensity and reduce the rate of thermal stress accumulation. This ensures that the preset jump trigger threshold and the upper limit of the rate change in synergy, thereby ensuring a dynamic balance between the energy storage system's adjustment sensitivity and the battery's safe operation capability.

[0077] The preset rise percentage threshold is a reference value used to determine whether the modified evaluation index forms a significant upward trend within a continuous response period. It depends on the statistical fluctuation characteristics of the modified evaluation index sequence, the number of sampling periods, and the minimum confidence level required for trend identification. It is usually set according to the statistical stability range of the percentage of rises within a continuous response period, and is generally set between 0.6 and 0.9. In this embodiment, it is set to 0.7, which can accurately identify the trend of continuous enhancement of energy storage regulation effect while eliminating random fluctuation interference, thereby timely reducing the preset jump trigger threshold to improve the system response capability.

[0078] The preset decline ratio threshold is a reference value used to determine whether the modified evaluation index forms a significant downward trend within a continuous response period. It depends on the short-cycle fluctuation amplitude of the modified evaluation index sequence, the sensitivity of the system thermal state changes, and the strength of the stability criterion required for trend determination. It is usually set according to the proportion of the decline times exceeding the upper limit of the statistical random fluctuation range, and is generally set between 0.6 and 0.9. In this embodiment, it is set to 0.7, which can identify the downward trend of regulation capacity in a timely manner when the energy storage heat load gradually accumulates, thereby increasing the preset jump trigger threshold to reduce the intensity of energy storage participation.

[0079] The preset threshold correction step size is a reference value for the correction ratio used to control the single adjustment range of the preset jump trigger threshold. It depends on the allowable adjustment range of the jump trigger threshold, the rate of change of the correction evaluation index, and the balance requirements between the system's adjustment sensitivity and stability. It is usually set according to 1% to 10% of the full adjustment range of the jump trigger threshold, and is generally set between 0.01 and 0.10. In this embodiment, it is set to 0.05, which can make the preset jump trigger threshold gradually and smoothly adjusted, avoiding excessive fluctuations in the system response due to excessive correction.

[0080] The preset rate compensation ratio is a reference value of the proportional coefficient used to synchronously adjust the upper limit of the rate when the preset jump trigger threshold changes. It depends on the allowable adjustment range of the upper limit of the rate, the thermal response characteristics of the battery, and the coupling relationship between power support capability and thermal safety margin. It is usually set according to 5% to 20% of the full adjustment range of the upper limit of the rate, and is generally set between 0.05 and 0.20. In this embodiment, it is set to 0.10, which enables the upper limit of the rate to be coordinated and adjusted with the change of the preset jump trigger threshold, thereby maintaining the dynamic adaptation relationship between energy storage regulation capability and battery thermal safety state.

[0081] By using the changing trend of the correction evaluation index sequence as the basis for adjusting the preset jump trigger threshold, the adjustment direction of the preset jump trigger threshold is directly determined by the combined result of the energy storage regulation effect and the thermal state change. When the correction evaluation index continues to rise, it indicates that the bus frequency recovery capability under unit heat load conditions is enhanced. At this time, the preset jump trigger threshold is decreased while the upper limit of the rate is increased simultaneously, enabling the energy storage system to participate in regulation earlier and have a higher power output capability under the same disturbance conditions, thereby accelerating the frequency deviation suppression speed. When the correction evaluation index continues to fall, it indicates that the frequency recovery benefit is weakened relative to the heat load cost. At this time, the preset jump trigger threshold is increased. The threshold for triggering a sudden change is lowered simultaneously with the upper limit of the rate of change, making the triggering conditions for energy storage to participate in regulation more stringent and reducing the power output intensity, thereby suppressing further accumulation of thermal stress. At the same time, the adjustment range of the preset threshold for triggering a sudden change is controlled by the product of the preset threshold correction step size and the average increment or average decrement of the correction evaluation index, so that the threshold change is consistent with the rate of change of the system's stable returns. Furthermore, the upper limit of the rate of change is kept synchronized with the direction of change of the preset threshold for triggering a sudden change through the preset rate compensation ratio, so that the power output capability and the trigger sensitivity form a coordinated adjustment relationship, thereby achieving dynamic adaptive matching between the energy storage regulation response capability and the battery thermal safety margin.

[0082] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A photovoltaic energy storage system based on flexible control, characterized in that, include: The data acquisition module is used to collect in real time the zoned output power, bus frequency deviation, DC bus voltage fluctuation amplitude, battery cluster temperature gradient and charge / discharge rate of several energy storage battery clusters, as well as the sudden cloud shading coefficient of the mountain photovoltaic array under isolated grid operation. The mode determination module is used to determine whether to enter the jump absorption mode based on the photovoltaic jump index and the preset jump trigger threshold. The photovoltaic jump index is constructed based on the rate of change of the sudden cloud shading coefficient and the gradient of the change of the partition output power. The strategy generation module is used to determine the preset participation ratio of each energy storage cluster and the upper limit of the charge / discharge rate corresponding to the energy storage capacity coefficient and the bus stability index based on the determination result of entering the jump absorption mode, so as to form a partitioned flexible storage strategy. The energy storage capacity coefficient is determined based on the bus stability index and the temperature difference gradient of the battery cluster, and the bus stability index is constructed based on the bus frequency deviation and the DC bus voltage fluctuation amplitude. The strategy execution module is used to execute power-limited storage control based on the partitioned flexible storage strategy; The strategy adjustment module is used to adjust the preset jump trigger threshold or the upper limit of the multiplier based on the correction evaluation index and the preset stability conditions. The correction evaluation index is constructed based on the magnitude of the decline in the bus frequency deviation and the change in the battery cluster temperature gradient within a preset response time after the execution of the partitioned flexible storage strategy.

2. The photovoltaic energy storage system based on flexible control according to claim 1, characterized in that, The mode determination module includes: The occlusion rate calculation unit is used to calculate the rate of change of the occlusion coefficient of the sudden cloud cluster within a preset jump construction time. A power gradient calculation unit is used to calculate the gradient of the change in the output power of the partition within the preset jump construction time. A jump index construction unit is used to construct the photovoltaic jump index based on the rate of change and the gradient of change; The mode determination unit is used to determine whether to enter the jump absorption mode when the photovoltaic jump index is greater than the preset jump trigger threshold.

3. The photovoltaic energy storage system based on flexible control according to claim 1, characterized in that, The strategy generation module includes: The oscillation feature extraction unit is used to construct a bus stability index curve based on the bus frequency deviation and the DC bus voltage fluctuation amplitude, and extract the oscillation amplitude features within a preset analysis time. A thermal stress construction unit is used to construct energy storage thermal stress indicators based on the temperature difference gradient of the battery cluster and the charge / discharge rate. An adjustment coefficient generation unit is used to generate a flexible adjustment coefficient based on the degree of difference between the oscillation amplitude characteristics and the energy storage thermal stress index. The strategy scaling unit is used to scale the preset participation ratio and the multiplier limit according to the flexibility adjustment coefficient to generate the partitioned flexible storage strategy.

4. The photovoltaic energy storage system based on flexible control according to claim 3, characterized in that, The adjustment coefficient generation unit includes: The difference calculation subunit is used to calculate the deviation value after normalizing the oscillation amplitude characteristics and the energy storage thermal stress index respectively. A proportional mapping subunit is used to generate a proportional mapping factor based on the deviation value and a preset adjustment mapping range; The coefficient output subunit is used to output the flexible adjustment coefficient according to the proportional mapping factor.

5. The photovoltaic energy storage system based on flexible control according to claim 4, characterized in that, The proportional mapping subunit is used to generate a corresponding nonlinear scaling factor based on the segmented adjustment interval where the deviation value is located, and to perform smooth transition processing on the nonlinear scaling factor between adjacent segmented adjustment intervals to generate the proportional mapping factor, wherein the segmented adjustment interval is obtained by dividing the preset difference adjustment interval according to a preset segmented threshold.

6. The photovoltaic energy storage system based on flexible control according to claim 3, characterized in that, The oscillation feature extraction unit includes: A peak detection subunit is used to identify local peaks of the bus stability index curve within the preset analysis time. The amplitude calculation subunit is used to calculate the oscillation amplitude between adjacent local peaks; The frequency statistics subunit is used to count the oscillation frequency per unit time. The feature output subunit is used to output the oscillation amplitude feature based on the oscillation amplitude and the oscillation frequency.

7. The photovoltaic energy storage system based on flexible control according to claim 3, characterized in that, The thermal stress building unit includes: The temperature difference change rate calculation subunit is used to calculate the rate of change of the temperature difference gradient of the battery cluster within a preset thermal evaluation time. A rate offset calculation subunit is used to calculate the offset of the charge / discharge rate relative to the nominal rate; A stress fusion subunit is used to construct the energy storage thermal stress index based on the rate of change and the offset.

8. The photovoltaic energy storage system based on flexible control according to claim 1, characterized in that, The strategy execution module includes: A power allocation unit is used to allocate target charge and discharge power to several energy storage battery clusters according to the partitioned flexible storage strategy; A rate limiting unit is used to limit the charge / discharge rate of each energy storage battery cluster from exceeding the upper limit of the rate. A real-time feedback unit is used to provide real-time feedback on the bus frequency deviation and the battery cluster temperature gradient during the execution of the power-limited storage control.

9. The photovoltaic energy storage system based on flexible control according to claim 1, characterized in that, The strategy adjustment module includes: The fall-off amplitude calculation unit is used to calculate the fall-off amplitude of the bus frequency deviation within the preset response time. Temperature difference change calculation unit, which is used to calculate the change in temperature difference gradient of the battery cluster; A correction index construction unit is used to construct the correction evaluation index based on the decline magnitude and the change amount; A threshold update unit is used to update the preset jump trigger threshold or the multiplier limit when the modified evaluation index does not meet the preset stability condition.

10. The photovoltaic energy storage system based on flexible control according to claim 9, characterized in that, The threshold update unit includes: The trend analysis subunit is used to analyze the changing trend of the modified evaluation index within a continuous response period of a preset update quantity; A dynamic correction subunit is used to incrementally or incrementally correct the preset jump trigger threshold according to the changing trend. The multiplier compensation subunit is used to simultaneously adjust the upper limit of the multiplier while the preset jump trigger threshold is being corrected.

Images