A black start power equalization control method and system for energy storage

Abstract

This invention relates to the field of power system energy storage technology, and particularly to a black-start power equalization control method and system for energy storage. The method includes identifying the internal resistance abrupt change inflection point of the battery energy storage module based on discharge depth segment distribution data, and determining the polarization steady-state window based on the internal resistance abrupt change inflection point; detecting the droop slope oscillation risk based on transient internal resistance sampling data and the polarization steady-state window, and determining the internal resistance sampling stability interval; after the flexible boost energy storage system enters the black-start operation mode, establishing a linkage mapping relationship between the real-time value of the droop slope and the real-time charge data based on the internal resistance sampling stability interval, dynamically compensating for the nonlinear drift of the internal resistance of each battery energy storage module, and obtaining the power output ratio of each battery energy storage module in the black-start operation mode. This invention improves the power distribution stability of the energy storage system under black-start and emergency conditions by dynamically identifying the internal resistance abrupt change inflection point and establishing a linkage mapping compensation between the droop slope and the charge.

Description

Technical Field

[0001] This invention relates to the field of power system energy storage technology, and in particular to an energy storage black start power balancing control method and system. Background Technology

[0002] In the field of energy storage and power systems, the stable operation and balanced power distribution of energy storage devices are crucial to ensuring the safety and reliability of the power grid. Especially during system startup or emergency conditions, energy storage devices need to respond quickly and provide stable power support to ensure the normal operation of the entire power grid. With the large-scale integration of new energy sources, energy storage technology has become a core support for smoothing power fluctuations and meeting emergency needs. However, existing methods still have significant shortcomings when dealing with complex operating environments. Most solutions ignore the dynamic characteristics of changes in the internal state of energy storage devices. In particular, during battery operation, its internal parameters will drift unpredictably with factors such as usage time and environmental conditions. Traditional control strategies are difficult to adapt to this time-varying characteristic, resulting in the system being unable to accurately allocate power at critical moments, thereby affecting overall efficiency and stability.

[0003] The internal impedance of a battery is affected by multiple factors, including the depth of discharge, temperature, and aging effects after long-term operation. This impedance drift directly interferes with the power distribution control logic. Especially in scenarios requiring rapid power output adjustments, instantaneous impedance fluctuations can easily cause control parameters to deviate from actual requirements, leading to uneven power distribution. For example, during the startup of an energy storage system or emergency power allocation, the rapid change in battery internal resistance often contradicts the adjustment process of control parameters. Attempting to match the real-time impedance by frequently adjusting parameters such as the droop coefficient often results in distorted impedance values ​​because the electrochemical reaction inside the battery has not yet reached a steady state. This causes repeated oscillations in control parameters, making it impossible to achieve balanced power distribution among the energy storage devices. This contradiction between control logic and battery physical characteristics is particularly prominent when multiple energy storage devices are operating in parallel. In scenarios where the energy storage system starts up or urgently allocates power, due to the lack of an effective identification and adaptive mechanism for internal impedance changes, the power output of one energy storage device may far exceed that of other energy storage devices, causing some devices to be overloaded and damaged, while the remaining devices are in an underloaded state for a long time. This serious imbalance in power distribution not only exacerbates the inconsistency of the battery pack, but also leads to the collapse of the entire system. Therefore, traditional control strategies are difficult to coordinate power sharing under dynamic operating conditions, which further exacerbates the power imbalance between energy storage modules during black start or emergency response, seriously threatening the safety and reliability of the system. Summary of the Invention

[0004] To address the above technical problems, this invention provides a black start power equalization control method and system for energy storage.

[0005] In a first aspect, the present invention provides a black-start power equalization control method for energy storage, applied to a flexible boost energy storage system composed of multiple battery energy storage modules connected in parallel, the energy storage black-start power equalization control method comprising the following steps: When the flexible boost energy storage system enters the black start preparation state, real-time charge data, terminal voltage drop rate and polarization relaxation response time of each battery energy storage module are collected from the flexible boost energy storage system to obtain discharge depth segment distribution data. Identify the internal resistance abrupt change inflection point of the battery energy storage module based on the discharge depth segment distribution data, and determine the polarization steady state window based on the discharge depth position corresponding to the internal resistance abrupt change inflection point. During the discharge process of each battery energy storage module, transient internal resistance sampling data is acquired, and the risk of droop slope oscillation is detected based on the transient internal resistance sampling data and the polarization steady-state window to determine the internal resistance sampling stability range. After the flexible boost energy storage system enters the black start operation mode, a linkage mapping relationship between the real-time value of the droop slope and the real-time charge data is established based on the internal resistance sampling stability range. Based on the aforementioned linkage mapping relationship, the nonlinear drift of the internal resistance of each battery energy storage module is dynamically compensated to obtain the power output ratio of each battery energy storage module in the black start operation mode.

[0006] In a further implementation, the step of acquiring real-time charge data, terminal voltage drop rate, and polarization relaxation response time of each battery energy storage module from the flexible boost energy storage system to obtain discharge depth segment distribution data includes: Collect real-time charge data and terminal voltage data of each battery energy storage module, and divide each battery energy storage module into several discharge depth segments based on the real-time charge data; The rate of voltage drop of each battery energy storage module is obtained by the ratio of the difference in terminal voltage data within adjacent sampling periods to the sampling period duration. The discharge state of each battery energy storage module is marked according to the terminal voltage drop rate, and based on the discharge state, the time taken for the terminal voltage to recover from the fluctuating state to the stable state during the discharge process is monitored to obtain the polarization relaxation response time. The real-time charge data, the terminal voltage drop rate, and the polarization relaxation response duration are time-aligned and matched according to a unified time base to obtain multi-dimensional battery module state parameters. The multi-dimensional battery module state parameters are mapped to the discharge depth segment of each battery energy storage module to obtain discharge depth segment distribution data.

[0007] In a further embodiment, the step of identifying the abrupt change in the internal resistance of the battery storage module based on the depth-of-discharge range distribution data includes: Based on the discharge depth segment distribution data, the terminal voltage drop rate of each battery energy storage module within the same discharge depth segment is compared cycle by cycle to obtain rate deviation data; Based on the rate deviation data and the preset deviation threshold, the inflection point of the internal resistance change of the battery energy storage module is identified.

[0008] In a further implementation, the step of determining the polarization steady-state window based on the discharge depth position corresponding to the internal resistance abrupt change inflection point includes: Extract the discharge depth location corresponding to the internal resistance abrupt change inflection point from the discharge depth segment distribution data to obtain the discharge depth inflection point. Based on the polarization relaxation response time corresponding to the discharge depth inflection point and the preset steady-state threshold, the range of continuous discharge depth segments is selected to obtain the initial time interval in which the polarization state tends to be stable. The initial time interval is time-constrained and corrected using the sampling timestamp of the internal resistance abrupt change in inflection point to obtain the polarization steady-state window.

[0009] In a further implementation, the step of acquiring transient internal resistance sampling data during the discharge process of each battery energy storage module includes: Based on the distribution data of the discharge depth segments, extract the real-time data change trend of the battery energy storage module during the discharge process in each discharge depth segment, and determine the gradient of the charge change. Based on the polarization relaxation response time, the characteristics of the polarization effect dissipation time of the battery energy storage module in each discharge depth segment are determined. The dynamic sampling time sequence is determined based on the charge change gradient and the polarization effect dissipation time characteristics, and the instantaneous terminal voltage and instantaneous discharge current values ​​are synchronously collected during the discharge process of each battery energy storage module according to the dynamic sampling time sequence. The difference between the instantaneous terminal voltage values ​​at adjacent sampling times is calculated to obtain the change in terminal voltage, and the difference between the instantaneous discharge current values ​​at adjacent sampling times is calculated to obtain the change in discharge current. The transient internal resistance value is obtained based on the ratio between the change in terminal voltage and the change in discharge current. The transient internal resistance value and its corresponding discharge depth segment are arranged in chronological order to form transient internal resistance sampling data.

[0010] In a further implementation, the step of detecting the risk of droop slope oscillation based on the transient internal resistance sampling data and the polarization steady-state window, and determining the internal resistance sampling stable range, includes: The transient internal resistance sampling data are sequentially mapped to the polarization steady-state window according to the sampling time to determine the fluctuation morphology characteristics at the edge of the polarization steady-state window; Based on the wave morphology characteristics, the droop slope oscillation risk energy density at each sampling point within the polarization steady-state window is calculated to obtain the continuous probability distribution of oscillation risk. Morphological clustering analysis was performed on the continuous probability distribution of the oscillation risk to obtain a stable candidate set of internal resistance sampling. Boundary envelope fitting is performed on the internal resistance sampling stable candidate set to obtain the internal resistance sampling stable interval.

[0011] In a further implementation, the step of establishing a linkage mapping relationship between the real-time value of the droop slope and the real-time charge data based on the internal resistance sampling stability interval includes: Based on the sampling time corresponding to the internal resistance sampling stability interval, the real-time charge data of each sampling time is extracted from the distribution data of the discharge depth segment to obtain the steady-state data point cloud; The steady-state data point cloud is traversed using a sliding window, and the local response coefficient sequence at the center of the sliding window with the state of charge is obtained by using a linear weighted least squares fitting method. The local response coefficient sequence is fitted with a nonlinear trend to obtain the droop slope response trend curve as the charge changes. Based on the polarization relaxation response time corresponding to the polarization steady-state window, the hysteresis effect characteristics of the internal electrochemical polarization of the battery during the black start process are determined. Based on the hysteresis effect characteristics, the droop slope response trend curve is phase-compensated and corrected to obtain a real-time estimate of the droop slope. Based on the relationship between the real-time estimated droop slope and the steady-state data point cloud, a linkage mapping relationship is established with real-time charge data as input and the real-time estimated droop slope as output.

[0012] In a further implementation, the step of dynamically compensating for the nonlinear drift of the internal resistance of each battery energy storage module according to the linkage mapping relationship to obtain the power output ratio of each battery energy storage module in the black start operation mode includes: The current charge data of each battery energy storage module is collected based on the real-time operating status of the flexible boost energy storage system after it enters the black start operation mode. By using the aforementioned linkage mapping relationship to query the real-time estimated value of the droop slope corresponding to the current charge data, the current droop slope value of each battery energy storage module can be obtained. The reference value of the internal resistance of each battery energy storage module is determined based on the internal resistance sampling stability range, and the real-time transient internal resistance value of each battery energy storage module is continuously collected during the black start operation. Based on the real-time transient internal resistance value and the internal resistance reference value, the nonlinear drift compensation amount of the internal resistance of each battery energy storage module during the black start operation process is obtained. Based on the internal resistance nonlinear drift compensation amount, the current droop slope value is reversed and corrected to obtain the droop slope control value after compensation and correction. Based on the rated power capacity of each battery energy storage module and the droop slope control value, the module equivalent droop coefficient of each battery energy storage module under the current operating state is obtained. Using the equivalent droop coefficient of the module as a weighting factor, the total load demand power of the flexible boost energy storage system in black start operation mode is weighted and allocated using the weighting factor to obtain the power output ratio of each battery energy storage module in black start operation mode.

[0013] In a further embodiment, the nonlinear drift compensation amount of the internal resistance is the difference between the real-time value of the transient internal resistance and the reference value of the internal resistance.

[0014] Secondly, the present invention provides an energy storage black start power balancing control system, applied to a flexible boost energy storage system composed of multiple battery energy storage modules connected in parallel, the energy storage black start power balancing control system comprising: The data acquisition module is used to collect real-time charge data, terminal voltage drop rate and polarization relaxation response time of each battery energy storage module from the flexible boost energy storage system when the flexible boost energy storage system enters the black start preparation state, so as to obtain the discharge depth segment distribution data. The steady-state analysis module is used to identify the internal resistance abrupt change inflection point of the battery energy storage module based on the discharge depth segment distribution data, and to determine the polarization steady-state window based on the discharge depth position corresponding to the internal resistance abrupt change inflection point. The oscillation detection module is used to acquire transient internal resistance sampling data during the discharge process of each battery energy storage module, and to perform droop slope oscillation risk detection based on the transient internal resistance sampling data and the polarization steady state window to determine the internal resistance sampling stable range. The linkage mapping module is used to establish a linkage mapping relationship between the real-time value of the droop slope and the real-time charge data based on the internal resistance sampling stability range after the flexible boost energy storage system enters the black start operation mode. The power balancing module is used to dynamically compensate for the nonlinear drift of the internal resistance of each battery energy storage module according to the linkage mapping relationship, so as to obtain the power output ratio of each battery energy storage module in the black start operation mode.

[0015] This invention provides a black-start power equalization control method and system for energy storage. The method includes: when the flexible boost energy storage system enters the black-start preparation state, collecting real-time charge data, terminal voltage drop rate, and polarization relaxation response time of each battery energy storage module in the flexible boost energy storage system to obtain discharge depth segment distribution data; identifying the internal resistance abrupt change inflection point of the battery energy storage module based on the discharge depth segment distribution data, and determining the polarization steady-state window based on the discharge depth position corresponding to the internal resistance abrupt change inflection point; acquiring transient internal resistance sampling data during the discharge process of each battery energy storage module, and performing droop slope oscillation risk detection based on the transient internal resistance sampling data and the polarization steady-state window to determine the internal resistance sampling stability interval; after the flexible boost energy storage system enters the black-start operation mode, establishing a linkage mapping relationship between the real-time value of the droop slope and the real-time charge data based on the internal resistance sampling stability interval; dynamically compensating for the nonlinear drift of the internal resistance of each battery energy storage module according to the linkage mapping relationship to obtain the power output ratio of each battery energy storage module in the black-start operation mode. Compared with existing technologies, this method establishes a real-time linkage mapping between droop slope and state of charge by dynamically identifying the characteristics of sudden changes in battery internal resistance and polarization steady-state window, and dynamically compensates and corrects nonlinear drift of internal resistance. This allows the droop slope control value to track the actual electrochemical characteristics of the battery in real time, thereby accurately calculating the power output ratio of each module in black-start operation mode. This effectively avoids the risk of power distribution imbalance and equipment overload caused by instantaneous impedance fluctuations, and improves the power balancing accuracy and operational stability of multi-module parallel systems under black-start and emergency conditions. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the energy storage black start power equalization control method provided in an embodiment of the present invention; Figure 2 This is a block diagram of the energy storage black start power balancing control system provided in an embodiment of the present invention.

[0017] Explanation of reference numerals in the attached diagram: 101, Data acquisition module; 102, Steady-state analysis module; 103, Oscillation detection module; 104, Linkage mapping module; 105, Power equalization module. Detailed Implementation

[0018] The embodiments of the present invention are described in detail below with reference to the accompanying drawings. The embodiments are given for illustrative purposes only and should not be construed as limiting the present invention. The accompanying drawings are for reference and illustration only and do not constitute a limitation on the scope of patent protection of the present invention, because many changes can be made to the present invention without departing from the spirit and scope of the present invention.

[0019] Figure 1This is a schematic flowchart of the energy storage black-start power balancing control method provided in an embodiment of the present invention. The present invention provides an energy storage black-start power balancing control method that can be applied to a flexible boost energy storage system composed of multiple battery energy storage modules connected in parallel, such as... Figure 1 As shown, the energy storage black-start power balancing control method includes the following steps: S1. When the flexible boost energy storage system enters the black start preparation state, real-time charge data, terminal voltage drop rate and polarization relaxation response time of each battery energy storage module are collected from the flexible boost energy storage system to obtain discharge depth segment distribution data.

[0020] In some embodiments, the step of acquiring real-time charge data, terminal voltage drop rate, and polarization relaxation response time of each battery energy storage module from the flexible boost energy storage system to obtain discharge depth segment distribution data includes: Collect real-time charge data and terminal voltage data of each battery energy storage module, and divide each battery energy storage module into several discharge depth segments based on the real-time charge data; The rate of voltage drop of each battery energy storage module is obtained by the ratio of the difference in terminal voltage data within adjacent sampling periods to the sampling period duration. The discharge state of each battery energy storage module is marked according to the terminal voltage drop rate, and based on the discharge state, the time taken for the terminal voltage to recover from the fluctuating state to the stable state during the discharge process is monitored to obtain the polarization relaxation response time. The real-time charge data, the terminal voltage drop rate, and the polarization relaxation response duration are time-aligned and matched according to a unified time base to obtain multi-dimensional battery module state parameters. The multi-dimensional battery module state parameters are mapped to the discharge depth segment of each battery energy storage module to obtain discharge depth segment distribution data.

[0021] Specifically, when the flexible boost energy storage system receives a black start command or detects a grid voltage loss and enters a standby state, this embodiment first obtains the real-time charge data of each parallel battery energy storage module through the battery management system. The real-time charge data represents the percentage of the battery's current remaining capacity relative to its rated total capacity. It is a core indicator reflecting the battery's remaining energy and a fundamental variable determining the battery's internal resistance characteristics. Since batteries exhibit different internal resistance change characteristics at different depths of discharge, this embodiment, based on the real-time charge data, determines the current discharge state of each battery energy storage module according to a preset charge threshold range. The discharge depth is quantified, and all battery energy storage modules are divided into multiple continuous discharge depth segments according to the numerical range of the discharge depth. For example, the charge capacity is divided into three segments from high to low: deep discharge segment, medium discharge segment, and shallow discharge segment. Each segment corresponds to a different range of changes in the internal electrochemical characteristics of the battery, thereby distinguishing the battery in different discharge stages. The discharge depth is defined as the ratio of the discharged capacity to the rated capacity. In the black start preparation state, the flexible boost energy storage system is usually connected to a preload or enters a standby discharge state. In this embodiment, data is continuously collected at preset fixed time intervals. The terminal voltage data of each battery energy storage module is collected, and the terminal voltage data of two adjacent sampling periods for each battery energy storage module are extracted. The difference between the terminal voltage data of the later sampling period and the terminal voltage data of the previous sampling period is calculated to obtain the terminal voltage data difference between two adjacent sampling periods (i.e., voltage drop value). In this embodiment, the instantaneous response characteristics of the battery internal resistance are characterized by the amount of voltage change per unit time. The faster the voltage drops, the greater the transient internal resistance of the battery or the stronger the polarization effect. Then, in this embodiment, the terminal voltage data difference between two adjacent sampling periods is divided by the sampling period duration to obtain the battery energy storage value. The terminal voltage drop rate of the module within the current sampling period reflects how quickly the terminal voltage decreases per unit time. A large terminal voltage drop rate indicates a significant internal polarization effect in the battery or severe fluctuations in the load current. A small terminal voltage drop rate indicates that the battery is in a relatively stable discharge state. In this embodiment, the terminal voltage data refers to the actual output voltage between the positive and negative terminals of the battery energy storage module. The sampling period is the time interval between two samplings. Since the terminal voltage generally decreases during the discharge process, the difference between the terminal voltage values ​​of two adjacent sampling periods is usually negative.

[0022] Since different discharge states correspond to different degrees of polarization effects within the battery, this embodiment marks the discharge state of each battery energy storage module based on the terminal voltage drop rate and several pre-set rate threshold intervals. For example, assuming the rate threshold intervals include three ranges: high, medium, and low, when the terminal voltage drop rate exceeds the high rate threshold interval, the battery energy storage module's discharge state is marked as a severe discharge state; when the terminal voltage drop rate is within the medium rate threshold interval, the battery energy storage module's discharge state is marked as a stable discharge state; and when the terminal voltage drop rate is below the low rate threshold interval, the battery energy storage module's discharge state is marked as a weak discharge state. For battery energy storage modules marked as severe or stable discharge states, this embodiment starts a high-precision timer during the discharge process to continuously monitor its terminal voltage fluctuations. When a step drop in terminal voltage due to load connection is detected, the system starts timing until... Timing stops when the terminal voltage no longer fluctuates drastically and the rate of voltage drop is within a moderate threshold range (i.e., the voltage enters a steady-state phase with a slow linear decrease). In this embodiment, the time span from the start of the voltage mutation to its recovery to the steady-state fluctuation range is taken as the polarization relaxation response duration. It should be noted that polarization relaxation refers to the process by which the terminal voltage gradually recovers from a fluctuating state to a relatively stable state after the battery's discharge current undergoes a sudden change or the discharge state changes. The system monitors the fluctuation of the terminal voltage in real time during the discharge process. When a significant voltage fluctuation is detected (e.g., due to load changes or control adjustments), timing begins until the amplitude of the terminal voltage fluctuation decays to within a preset stable threshold and remains below it for a certain period of time. Timing stops at this point, and the recorded time length is the polarization relaxation response duration. This polarization relaxation response duration reflects the dissipation rate of electrochemical polarization and concentration polarization inside the battery, and it characterizes the time required for the electrochemical reaction inside the battery to reach a steady state.

[0023] Because the three parameters—real-time charge data, terminal voltage drop rate, and polarization relaxation response duration—have different physical dimensions and different acquisition frequencies, this embodiment aligns all data according to a unified time base to ensure the accuracy of subsequent analysis. Specifically, using the terminal voltage sampling period as the base time axis, the charge data is mapped to each terminal voltage sampling time point using a linear interpolation method, so that each time point simultaneously has both charge value and terminal voltage drop rate value. For the polarization relaxation response duration, since it is itself a time interval length, the system associates it with the start or end time of the corresponding fluctuation event and assigns the polarization relaxation response duration at that time point. Through the above alignment... In this embodiment, a set of three-dimensional state vectors is established for each battery energy storage module at each key time point, namely the charge, terminal voltage drop rate, and polarization relaxation response duration at the same moment, to obtain multi-dimensional battery module state parameters. For each battery energy storage module, this embodiment assigns its multi-dimensional state parameters (i.e., the charge, terminal voltage drop rate, and polarization relaxation response duration of the battery energy storage module at the current moment) to the discharge depth segment in which it is located. Since the charge directly determines the discharge depth, each battery energy storage module corresponds to a unique discharge depth segment. This embodiment summarizes the multi-dimensional state parameters of all modules in the same discharge depth segment to form the discharge depth segment distribution data of that segment.

[0024] S2. Identify the internal resistance abrupt change inflection point of the battery energy storage module based on the discharge depth segment distribution data, and determine the polarization steady-state window based on the discharge depth position corresponding to the internal resistance abrupt change inflection point.

[0025] In some embodiments, the step of identifying the internal resistance abrupt change inflection point of the battery energy storage module based on the discharge depth segment distribution data, and determining the polarization steady-state window based on the discharge depth position corresponding to the internal resistance abrupt change inflection point, includes: Based on the discharge depth segment distribution data, the terminal voltage drop rate of each battery energy storage module within the same discharge depth segment is compared cycle by cycle to obtain rate deviation data; Based on the rate deviation data and the preset deviation threshold, the inflection point of the internal resistance change of the battery energy storage module is identified; Extract the discharge depth location corresponding to the internal resistance abrupt change inflection point from the discharge depth segment distribution data to obtain the discharge depth inflection point. Based on the polarization relaxation response time corresponding to the discharge depth inflection point and the preset steady-state threshold, the range of continuous discharge depth segments is selected to obtain the initial time interval in which the polarization state tends to be stable. The initial time interval is time-constrained and corrected using the sampling timestamp of the internal resistance abrupt change in inflection point to obtain the polarization steady-state window.

[0026] Specifically, since this embodiment assigns each battery energy storage module to different segments according to its current discharge depth, and multiple battery energy storage modules are contained within the same discharge depth segment, this embodiment reads the terminal voltage drop rate of all battery energy storage modules within the same discharge depth segment periodically, using a sampling period as the unit. In each sampling period, the average terminal voltage drop rate of all battery energy storage modules within that discharge depth segment is calculated to obtain the segment average rate. The terminal voltage drop rate of each battery energy storage module is then compared with the segment average rate to obtain the deviation of each battery energy storage module from the segment average rate. The rate deviation data of the battery energy storage module within the sampling period is obtained. The rate deviation data can be positive (above the average value) or negative (below the average value). The absolute value reflects the degree of rate deviation of the module relative to other modules in the same segment. In this embodiment, a rate deviation data is stored for each battery energy storage module in each period, forming a rate deviation data sequence that changes over time. This rate deviation data sequence can reflect the abnormal change trend of the battery energy storage module relative to the average level of the same segment during the discharge process. When the rate deviation data is small, it indicates that the characteristics of each battery energy storage module are consistent; when the rate deviation data suddenly increases, it indicates that the internal resistance of one or more battery energy storage modules has changed abnormally.

[0027] Meanwhile, this embodiment uses a pre-set deviation threshold to determine whether the deviation of the terminal voltage drop rate has reached a significant level rather than an abnormal fluctuation. When the rate deviation data of a battery energy storage module continuously exceeds the preset deviation threshold in multiple consecutive sampling periods, or when the absolute value of the deviation increases abruptly and exceeds the deviation threshold, the system determines that the battery energy storage module has experienced an internal resistance mutation near that moment. This embodiment uses the critical sampling moment when the rate deviation data jumps from below the deviation threshold to above the deviation threshold as the internal resistance mutation inflection point. This critical sampling moment corresponds to a drastic change in the internal polarization state of the battery (such as the sudden establishment of concentration polarization) and is the key boundary distinguishing the linear response region and the nonlinear mutation region of the battery. It should be noted that the internal resistance mutation inflection point specifically refers to the point before the critical sampling moment when the battery energy storage module... The rate of voltage drop is basically consistent with the average level of the same segment (deviation within the threshold); after the critical sampling moment, its rate deviates significantly from the average level (deviation exceeds the threshold), and this deviation is not caused by normal changes in the depth of discharge, but rather by nonlinear jumps in the internal impedance of the battery. The system records the sampling period number or specific timestamp corresponding to this inflection point as the internal resistance mutation inflection point of the battery energy storage module. In order to improve the reliability of identification, in another embodiment, this embodiment can take the current sampling period as the center and use a sliding window to take the rate deviation data of several periods before and after. The deviation change rate within the sliding window is calculated based on the rate deviation data of several periods before and after. When the deviation change rate exceeds the preset mutation slope threshold, it is determined that there is an internal resistance mutation, and the center point of the sliding window is marked as the internal resistance mutation inflection point.

[0028] After identifying the internal resistance abrupt change inflection point, this embodiment maps the internal resistance abrupt change inflection point to the corresponding discharge depth value. Since the discharge depth segment distribution data already contains the charge data of each module at each sampling time, this embodiment queries the charge state value corresponding to the internal resistance abrupt change inflection point from the state history record of the battery energy storage module through the timestamp index, and calculates the average charge state value of all modules in the discharge depth segment as the discharge depth position corresponding to the internal resistance abrupt change inflection point. This discharge depth position is marked as the discharge depth inflection point, which represents the critical charge position of the battery energy storage module from the stable output region to the internal resistance surge region. Different battery energy storage modules have different discharge depth inflection points. This embodiment uses the discharge depth inflection point as the key reference position for subsequently determining the polarization steady state window.

[0029] Since each battery energy storage module corresponds to a polarization relaxation response time at its discharge depth inflection point, and this polarization relaxation response time is obtained through the aforementioned monitoring of the terminal voltage fluctuation recovery process, the polarization relaxation response time reflects the speed at which the polarization effect inside the battery dissipates. The shorter the polarization relaxation response time, the faster the polarization dissipates, and the faster the system tends to a steady state; the longer the polarization relaxation response time, the greater the polarization inertia. In this embodiment, a preset steady-state threshold is used to define whether the polarization state has basically stabilized. Specifically, starting from the discharge depth inflection point, each subsequent discharge depth segment is checked sequentially along the direction of increasing discharge depth (i.e., the direction of deeper discharge). For each subsequent discharge depth segment, this embodiment obtains the average polarization relaxation response time of the battery energy storage module within that discharge depth segment and sets the average polarization... The relaxation response time is compared with a preset steady-state threshold. When the average polarization relaxation response time corresponding to multiple consecutive discharge depth segments is less than or equal to the steady-state threshold, the polarization state is considered to have entered a stable range. These segments mean that the electrochemical response time of the battery tends to be stable and no longer changes drastically with increasing depth. Therefore, in this embodiment, the range of continuous discharge depth segments from the discharge depth inflection point until the steady-state condition is met is defined as the initial time interval for the polarization state to tend to be stable. This initial time interval reflects the stable working period of the battery in a specific discharge depth region where the internal impedance changes slowly and the polarization effect is predictable. It should be noted that this initial time interval is arranged according to the discharge depth, but since the discharge process corresponds one-to-one with time, it can also be mapped to a time interval.

[0030] It should be noted that in actual systems, the rate of change in discharge depth is not constant (for example, current fluctuations can cause changes in discharge rate). Therefore, time intervals divided solely by discharge depth segments will deviate from the actual time progression. To improve accuracy, this embodiment uses the initial time interval obtained above as a benchmark and takes the sampling timestamp of the internal resistance abrupt change inflection point as a time constraint boundary to perform time constraint correction on the initial time interval. The principle of time constraint correction is that the start time of the polarization steady-state window must not be earlier than the internal resistance abrupt change in inflection point, and the duration of the polarization steady-state window must cover the entire process of polarization response duration stabilizing. This embodiment introduces timestamp constraints to eliminate invalid data caused by the lag in depth division, ultimately obtaining a polarization steady-state window that satisfies both the depth range requirement and the time synchronization requirement. Specifically, if the start time of the initial time interval is earlier than the sampling timestamp of the internal resistance abrupt change in inflection point, the start time is corrected to... The sampling timestamp ensures that the polarization steady-state window does not include the unstable period before the internal resistance mutation. If the end time of the initial time interval is later than the sampling timestamp of the next adjacent internal resistance mutation inflection point, the end time is corrected to the sampling timestamp of the next adjacent internal resistance mutation inflection point, ensuring that the polarization steady-state window ends before the next internal resistance mutation. The time period obtained after the above time constraint correction is the polarization steady-state window. This polarization steady-state window represents the optimal sampling period in which the internal polarization state of the battery energy storage module tends to be stable and the impedance sampling data is reliable and effective within a specific discharge depth range and a specific time interval. Within this polarization steady-state window, the polarization effect of the battery has basically subsided, and the terminal voltage and internal resistance response tend to be stable, ensuring that the subsequent power distribution control only takes effect after the battery has truly entered a steady state, thereby avoiding erroneous adjustments based on transient distortion data and laying a solid timing foundation for achieving accurate power balance during black start.

[0031] S3. Acquire transient internal resistance sampling data during the discharge process of each battery energy storage module, and perform droop slope oscillation risk detection based on the transient internal resistance sampling data and the polarization steady-state window to determine the internal resistance sampling stability range.

[0032] In some implementations, the step of acquiring transient internal resistance sampling data during the discharge process of each battery energy storage module, and determining the internal resistance sampling stability range based on the transient internal resistance sampling data and the polarization steady-state window to detect the risk of droop slope oscillations includes: Based on the distribution data of the discharge depth segments, extract the real-time data change trend of the battery energy storage module during the discharge process in each discharge depth segment, and determine the gradient of the charge change. Based on the polarization relaxation response time, the characteristics of the polarization effect dissipation time of the battery energy storage module in each discharge depth segment are determined. The dynamic sampling time sequence is determined based on the charge change gradient and the polarization effect dissipation time characteristics, and the instantaneous terminal voltage and instantaneous discharge current values ​​are synchronously collected during the discharge process of each battery energy storage module according to the dynamic sampling time sequence. The difference between the instantaneous terminal voltage values ​​at adjacent sampling times is calculated to obtain the change in terminal voltage, and the difference between the instantaneous discharge current values ​​at adjacent sampling times is calculated to obtain the change in discharge current. The transient internal resistance value is obtained based on the ratio between the change in terminal voltage and the change in discharge current. The transient internal resistance value and its corresponding discharge depth segment are arranged in chronological order to form transient internal resistance sampling data. The transient internal resistance sampling data are sequentially mapped to the polarization steady-state window according to the sampling time to determine the fluctuation morphology characteristics at the edge of the polarization steady-state window; Based on the wave morphology characteristics, the droop slope oscillation risk energy density at each sampling point within the polarization steady-state window is calculated to obtain the continuous probability distribution of oscillation risk. Morphological clustering analysis was performed on the continuous probability distribution of the oscillation risk to obtain a stable candidate set of internal resistance sampling. Boundary envelope fitting is performed on the internal resistance sampling stable candidate set to obtain the internal resistance sampling stable interval.

[0033] Specifically, to address the issue of fluctuating control parameters in traditional methods, this embodiment uses dynamic sampling and oscillation risk assessment to eliminate false internal resistance data in polarization transients or unstable states at the source, ensuring the accuracy of subsequent power distribution control. For each discharge depth segment, this embodiment extracts real-time charge data of each battery energy storage module during the discharge process from the discharge depth segment distribution data. The real-time charge data gradually decreases over time (during the discharge process). This embodiment observes the change in charge value sequentially according to the sampling period and calculates the charge reduction per unit time or per unit discharge depth. Specifically, this embodiment takes real-time charge data from two adjacent sampling periods within the same discharge depth segment, calculates the charge difference between the real-time charge data at the later moment and the previous moment, and divides it by the time interval between the two sampling periods to obtain the charge change gradient per unit time. This charge change gradient reflects the rate of discharge of the battery energy storage module within that discharge depth segment. The larger the absolute value of the charge change gradient, the larger the discharge current or the faster the battery capacity decays. Furthermore, different discharge... The time required for battery depolarization to dissipate varies with discharge depth. Therefore, for each discharge depth segment, this embodiment collects the polarization relaxation response times of all battery energy storage modules within that segment. Statistical analysis is used to obtain the polarization effect dissipation time characteristics of that segment. These characteristics can be the average, median, or distribution range of the polarization relaxation response times of all battery energy storage modules within that discharge depth segment. For example, if the polarization relaxation response times are relatively concentrated within the same discharge depth segment, the statistical center value of the polarization relaxation response times within that segment is taken as the value of that segment. The polarization effect dissipation time characteristic is determined by the following: If the distribution is relatively dispersed, a conservative estimate that covers most sampling times is selected as the polarization effect dissipation time characteristic, taking into account the fluctuation of the voltage drop rate within the segment. It should be noted that this polarization effect dissipation time characteristic represents the shortest time constraint required for the battery to transition from an unsteady state to a steady state. The larger the polarization effect dissipation time characteristic value, the stronger the polarization inertia of the battery within the discharge depth segment, and the longer the time required for the voltage to recover to stability. The smaller the polarization effect dissipation time characteristic value, the faster the polarization dissipates and the more sensitive the system response.

[0034] Traditional fixed-frequency sampling is prone to capturing voltage spikes or troughs, leading to errors in internal resistance calculation. To avoid unreliable data due to sampling during periods of unstable polarization or drastic charge changes, this embodiment constructs a dynamically adjusted sampling time sequence. The sampling interval of this dynamically adjusted sequence is not fixed but adjusted in real-time based on the charge change gradient and polarization dissipation time characteristics of the current discharge depth segment. When the absolute value of the charge change gradient is large (i.e., the discharge rate is fast), the battery state changes rapidly, and this embodiment requires a shorter sampling interval to capture transient characteristics. Conversely, when the absolute value of the charge change gradient is small, this embodiment can use a longer sampling interval. Furthermore, if the polarization dissipation time characteristic value is long, it indicates that the battery needs more time to recover from polarization disturbances; in this case, the sampling interval should be appropriately widened to avoid sampling when polarization is unstable. If the polarization dissipation time characteristic value is short, sampling can be more frequent. It should be noted that this embodiment automatically skips polarization effects after load jumps or operating condition changes. The sampling period should be determined based on the duration corresponding to the dissipation time characteristics before sampling begins. This embodiment combines the above two factors to dynamically calculate the sampling period to be used at each moment, thereby generating a non-uniform dynamic sampling time sequence. Then, according to the time points of this dynamic sampling time sequence, this embodiment synchronously collects the instantaneous values ​​of terminal voltage and discharge current during the discharge process of each battery energy storage module to ensure that the two electrical quantities are strictly aligned in time. The instantaneous values ​​of terminal voltage at two adjacent sampling moments are taken out in sequence, and the difference between the instantaneous values ​​of terminal voltage at two adjacent sampling moments is calculated to obtain the change in terminal voltage. Since the terminal voltage of the battery generally shows a downward trend during the discharge process, this change is negative, and its absolute value represents the magnitude of the voltage drop. Similarly, this embodiment performs a difference calculation on the instantaneous values ​​of discharge current at two adjacent sampling moments to obtain the change in discharge current. The discharge current increases or decreases due to load fluctuations, so the change in discharge current can be positive or negative. The change in terminal voltage and the change in discharge current must come from the same pair of adjacent sampling moments to ensure the physical consistency of subsequent calculations.

[0035] Next, this embodiment calculates the ratio of the change in terminal voltage to the change in discharge current to obtain the transient internal resistance value. This transient internal resistance value characterizes the amplitude of the change in terminal voltage caused by a unit change in current, reflecting the internal equivalent resistance of the battery in the current transient process. It should be noted that, since it is a small change, this transient internal resistance value is not the DC internal resistance or AC impedance, but a dynamic internal resistance calculated based on the instantaneous changes in voltage and current during the actual discharge process. It can more realistically reflect the impedance characteristics of the battery under operating conditions. Since this ratio loses its physical meaning when the change in discharge current approaches zero, the system will first determine whether the absolute value of the change in discharge current is greater than a preset minimum threshold before calculation. Only when the condition is met will a valid calculation be performed. In this embodiment, each calculated transient internal resistance value, its corresponding sampling time, and the discharge depth segment to which the sampling time belongs are arranged in chronological order to form continuous transient internal resistance sampling data. This transient internal resistance sampling data records the dynamic evolution trajectory of the internal resistance with the discharge process.

[0036] This embodiment filters out data points within the polarization steady-state window time range from the transient internal resistance sampling data and arranges them in chronological order. It analyzes the change patterns of these data points at the edges of the polarization steady-state window (i.e., near the beginning and end of the polarization steady-state window). Specifically, at the beginning of the polarization steady-state window (just past the inflection point of internal resistance abrupt change), the transient internal resistance value often exhibits large jumps or violent fluctuations; while at the end of the polarization steady-state window, the transient internal resistance value gradually converges to a relatively stable level. This embodiment extracts the fluctuation amplitude and fluctuation frequency at the beginning of the polarization steady-state window, as well as the convergence speed and steady-state residual fluctuations at the end of the polarization steady-state window, to obtain fluctuation pattern characteristics. These fluctuation pattern characteristics are used to quantify the specific behavioral pattern of the internal resistance transitioning from unstable to stable within the polarization steady-state window.

[0037] In droop control, fluctuations in internal resistance directly lead to oscillations in power distribution. The droop slope is a key parameter in the power balance control of energy storage systems. Adjusting the droop slope within a window where polarization is not yet stable can easily cause control oscillations. In this embodiment, a sliding window is used within the polarization steady-state window to calculate the difference between the transient internal resistance value of the current sampling point and the transient internal resistance value of the previous sampling point. The square of this difference is divided by the time interval between adjacent sampling points to obtain the droop slope oscillation risk energy density at that sampling point. This droop slope oscillation risk energy density describes the degree of risk of oscillation caused by internal resistance fluctuations in droop slope control at each time point or at each discharge depth position within the polarization steady-state window. The larger the fluctuation amplitude, the higher the fluctuation frequency, and the more severe the abrupt change at the beginning of the polarization steady-state window, the higher the oscillation risk energy density. Conversely, at the end of the polarization steady-state window, the fluctuation tends to be gentler, and the risk energy density gradually decreases. In this embodiment, all sampling points within the polarization steady-state window are traversed to obtain an oscillation risk energy density sequence covering the entire time period of the window. The sequence is normalized to convert the risk energy density of each sampling point into a relative proportion to the maximum risk value, forming a continuous probability distribution of oscillation risk. This continuous probability distribution of oscillation risk represents the probability or risk intensity of oscillation when the droop slope is adjusted at the corresponding position. The higher the oscillation risk energy density of the droop slope, the more unstable the internal resistance data at that moment, and the easier it is to cause control oscillation.

[0038] This embodiment employs a clustering algorithm to divide the continuous probability distribution of oscillation risk into multiple sub-intervals with similar morphological characteristics. During the clustering process, the mean and variance of the oscillation risk energy density of each sub-interval, as well as the similarity of the curve shape, are comprehensively considered. Sub-intervals with low oscillation risk energy density and relatively stable curve shapes are identified as potential stable regions. These identified potential stable regions are then merged, with adjacent sub-intervals possessing similar characteristics being combined into larger stable regions. Morphological clustering identifies the clustered regions with the lowest risk density and the most gradual changes. The transient internal resistance values ​​corresponding to the sampling points within these clustered regions are the most stable and least affected by polarization disturbances, serving as reliable internal resistance data for subsequent power equalization control. This embodiment aggregates these sampling points and their corresponding transient internal resistance values ​​to form the internal resistance... The internal resistance sampling stable candidate set contains multiple discrete sampling points, which may not be completely continuous on the time axis or discharge depth axis. In order to obtain a continuous and complete interval, this embodiment performs boundary envelope fitting on all sampling points in the internal resistance sampling stable candidate set. Specifically, the convex hull algorithm is used to enclose all points in the internal resistance sampling stable candidate set with a smooth curve or straight line. The smallest continuous time range covered by the envelope is the internal resistance sampling stable interval. This internal resistance sampling stable interval is completely located inside the polarization steady-state window, but it is usually narrower than the polarization steady-state window, eliminating the high-risk parts at the window edge. Within this internal resistance sampling stable interval, the transient internal resistance sampling data has been sufficiently stabilized, and the system can safely make dynamic adjustments to the droop slope based on these internal resistance values ​​without causing control oscillations.

[0039] S4. After the flexible boost energy storage system enters the black start operation mode, a linkage mapping relationship between the real-time value of the droop slope and the real-time charge data is established based on the internal resistance sampling stability range.

[0040] In some implementations, the step of establishing a linkage mapping relationship between the real-time value of the droop slope and the real-time charge data based on the internal resistance sampling stability interval includes: Based on the sampling time corresponding to the internal resistance sampling stability interval, the real-time charge data of each sampling time is extracted from the distribution data of the discharge depth segment to obtain the steady-state data point cloud; The steady-state data point cloud is traversed using a sliding window, and the local response coefficient sequence at the center of the sliding window with the state of charge is obtained by using a linear weighted least squares fitting method. The local response coefficient sequence is fitted with a nonlinear trend to obtain the droop slope response trend curve as the charge changes. Based on the polarization relaxation response time corresponding to the polarization steady-state window, the hysteresis effect characteristics of the internal electrochemical polarization of the battery during the black start process are determined. Based on the hysteresis effect characteristics, the droop slope response trend curve is phase-compensated and corrected to obtain a real-time estimate of the droop slope. Based on the relationship between the real-time estimated droop slope and the steady-state data point cloud, a linkage mapping relationship is established with real-time charge data as input and the real-time estimated droop slope as output.

[0041] Specifically, after the flexible boost energy storage system officially switches to black-start operation mode, since the transient internal resistance value within the internal resistance sampling stability range has stabilized, the corresponding charge data is also in a relatively stable change phase. For each sampling moment within the internal resistance sampling stability range, this embodiment queries the real-time charge data of each battery energy storage module at the same moment from the discharge depth segment distribution data. Simultaneously, since the droop slope in energy storage system droop control is proportional to the internal resistance or virtual impedance, this embodiment converts the transient internal resistance value into an initial droop slope using a preset proportionality coefficient. The initial reference value is used, and all the initial reference values ​​of the droop slope obtained after conversion and located within the stable range of internal resistance sampling, along with their corresponding charge values, are gathered together to form a steady-state data point cloud distributed on a two-dimensional plane. This steady-state data point cloud reflects the true electrochemical characteristic distribution of the battery after excluding polarization transient interference. Since the characteristics of the battery are nonlinear, a globally uniform slope cannot reflect local details. Therefore, this embodiment uses a sliding window to capture the local sensitivity when there are small changes in charge. Specifically, this embodiment arranges the points in the steady-state data point cloud in ascending order of charge (…). The data points are arranged naturally from largest to smallest (depending on the discharge direction), and a sliding window of fixed width is set. The width of the sliding window is adaptively set according to the density and distribution characteristics of the data point cloud to ensure that the window contains a sufficient number of data points for effective fitting. At the same time, the sliding window starts from the starting position of the steady-state data point cloud and moves backward step by step according to a preset step size. After each movement, linear weighted least squares fitting is performed on the data points in the sliding window. In the process of linear weighted least squares fitting, in order to ensure that the fitting result reflects the characteristics of the current moment rather than the characteristics of the past or future, this embodiment assigns different weight coefficients to the data points at different positions in the sliding window. The data points near the center of the sliding window are given higher weights, and the data points closer to the edge of the window are given lower weights. The allocation of weight coefficients can follow the Gaussian distribution or triangular distribution law to ensure that the fitting result is closer to the true response characteristics at the center position of the sliding window. The best straight line with the smallest sum of squared weighted vertical distances to all data points in the window is found through least squares fitting. The slope of this straight line is the local rate of change of the droop slope at the charge position of the center of the window with the change of charge, and it is used as the local response coefficient.

[0042] After the sliding window traverses the entire steady-state data point cloud, this embodiment obtains the corresponding local response coefficients at the center of each window. These local response coefficients are arranged in order of increasing charge to form a local response coefficient sequence. This sequence describes the sensitivity of the droop slope to changes in charge within different charge ranges. The local response coefficient sequence reflects the instantaneous local rate of change of the droop slope with changes in charge. However, due to the significant nonlinear characteristics of the internal resistance of the battery energy storage module, noise and fluctuations exist. Therefore, in order to obtain a smooth and globally continuous droop slope response curve, this embodiment can use fitting methods such as polynomial fitting, spline interpolation, or kernel-based smoothing regression to obtain the local response coefficient. The system performs nonlinear trend fitting on the local response coefficient sequence. For example, in this embodiment, the charge is used as the independent variable and the local response coefficient is used as the dependent variable. A piecewise cubic spline function is used to approximate the overall trend of the sequence. During the fitting process, the system balances the fitting accuracy and curve smoothness to avoid overfitting noise. After the fitting is completed, the rate of change of the droop slope with the charge is equal to a certain nonlinear function with respect to the charge. Then, the system integrates this nonlinear function (i.e., integrates from the initial charge to the current charge) to obtain the cumulative effect of the droop slope itself with the change of charge, i.e., the droop slope response trend curve. This droop slope response trend curve directly gives the droop slope that adapts to the current internal resistance characteristics of the battery under different charge levels.

[0043] The electrochemical polarization phenomenon inside the battery exhibits a significant hysteresis effect. When the discharge current or charge changes, the internal resistance and terminal voltage do not respond immediately but require a relaxation process to reach a new steady state. The length of the polarization steady-state window (i.e., the time span) directly reflects the strength of the hysteresis effect. Meanwhile, the droop slope response trend curve is obtained by fitting the steady-state data point cloud, describing the ideal relationship between the droop slope and charge in a fully steady state. However, during actual black-start operation, the charge changes in real time, and due to the polarization hysteresis effect, the current internal resistance (and thus the ideal droop slope) actually depends on the charge change history over a past period, rather than solely on the current charge. To reflect this physical reality, this embodiment uses the hysteresis effect characteristic to perform phase compensation correction on the trend curve. Therefore, this embodiment extracts the polarization relaxation response time corresponding to the polarization steady-state window and defines this polarization relaxation response time as the hysteresis effect of electrochemical polarization inside the battery during black-start. The longer the polarization relaxation response time, the more pronounced the battery polarization hysteresis effect, and the greater the time delay in the response of internal resistance to changes in charge. Conversely, the shorter the polarization relaxation response time, the faster the response of internal resistance to changes in charge. To offset the control delay caused by polarization hysteresis, this embodiment performs phase compensation correction on the droop slope response trend curve based on the hysteresis effect characteristics. For the hysteresis effect of positive delay characteristics, the droop slope response trend curve is shifted in the opposite direction of the charge state change direction. The shift amount is determined comprehensively based on the intensity of the hysteresis effect and the rate of charge state change. For the hysteresis effect of negative lead characteristics, the curve is shifted in the direction of charge state change. The shifted curve reflects the actual response characteristics of the droop slope after considering the electrochemical polarization delay. On this corrected curve, the corresponding droop slope value is queried based on the current real-time charge data, which is the real-time estimated value of the droop slope. This estimated value compensates for the hysteresis effect of the battery's internal polarization process, enabling the control command to remain synchronized with the actual physical process in time.

[0044] In another embodiment, to offset the control delay caused by polarization hysteresis, this embodiment monitors the rate of change of charge over time (i.e., discharge rate) in real time and constructs a filter with inertial delay based on the polarization relaxation response duration. The real-time charge at the current moment is processed by the filter to obtain the effective charge, which takes into account the equivalent charge state considering polarization inertia. Then, the effective charge is substituted into the droop slope response trend curve, and the calculated droop slope value is the real-time estimate of the droop slope considering the hysteresis effect. When the charge changes slowly, the hysteresis effect is not obvious, and the real-time estimate of the droop slope is close to the steady-state value. When the charge changes rapidly, the real-time estimate of the droop slope will lag behind the steady-state curve to avoid oscillation caused by advance adjustment.

[0045] Finally, this embodiment divides the effective range of real-time charge data into several continuous intervals. Within each interval, the correspondence between the real-time estimated droop slope and the real-time charge data is discretized and stored to form a linkage mapping lookup table. The data points in the linkage mapping lookup table are arranged in ascending order of charge to ensure data orderliness. For charge values ​​not present in the linkage mapping lookup table, this embodiment can use linear interpolation or spline interpolation methods to calculate the corresponding real-time estimated droop slope, ensuring the continuity and completeness of the linkage mapping relationship. After the linkage mapping relationship is established, during black-start operation, this embodiment can collect the charge data of each battery energy storage module in real time. By querying the linkage mapping lookup table and performing necessary interpolation calculations, the corresponding real-time estimated droop slope can be quickly obtained. This linkage mapping relationship realizes real-time and accurate mapping between charge state and droop slope parameter, providing key control parameters for subsequent internal resistance nonlinear drift compensation and power equalization distribution.

[0046] S5. Based on the aforementioned linkage mapping relationship, dynamically compensate for the nonlinear drift of the internal resistance of each battery energy storage module to obtain the power output ratio of each battery energy storage module in the black start operation mode.

[0047] In some embodiments, the step of dynamically compensating for the nonlinear drift of the internal resistance of each battery energy storage module according to the linkage mapping relationship to obtain the power output ratio of each battery energy storage module in the black start operation mode includes: The current charge data of each battery energy storage module is collected based on the real-time operating status of the flexible boost energy storage system after it enters the black start operation mode. By using the aforementioned linkage mapping relationship to query the real-time estimated value of the droop slope corresponding to the current charge data, the current droop slope value of each battery energy storage module can be obtained. The reference value of the internal resistance of each battery energy storage module is determined based on the internal resistance sampling stability range, and the real-time transient internal resistance value of each battery energy storage module is continuously collected during the black start operation. Based on the real-time transient internal resistance value and the internal resistance reference value, the nonlinear drift compensation amount of the internal resistance of each battery energy storage module during the black start operation process is obtained. Based on the internal resistance nonlinear drift compensation amount, the current droop slope value is reversed and corrected to obtain the droop slope control value after compensation and correction. Based on the rated power capacity of each battery energy storage module and the droop slope control value, the module equivalent droop coefficient of each battery energy storage module under the current operating state is obtained. Using the equivalent droop coefficient of the module as a weighting factor, the total load demand power of the flexible boost energy storage system in black start operation mode is weighted and allocated using the weighting factor to obtain the power output ratio of each battery energy storage module in black start operation mode.

[0048] Specifically, in this embodiment, the current charge data of each battery energy storage module is collected through the battery management system. Since the load demand may change at any time during the black start process and the discharge rate of each battery energy storage module is different, this embodiment needs to continuously and synchronously collect the current charge data of all battery energy storage modules to ensure that subsequent calculations are based on the same time segment. For each battery energy storage module, this embodiment uses its current charge data as a query key value and substitutes it into the linkage mapping relationship to obtain the corresponding real-time estimate of the droop slope. The real-time estimate of the droop slope obtained is used as the current droop slope value of each battery energy storage module under the current charge state. Since the linkage mapping relationship has been corrected by previous steady-state data point cloud fitting and phase compensation, This embodiment fully considers the nonlinear change of internal resistance with charge and the polarization hysteresis effect. Therefore, the current droop slope value reflects the real-time response characteristics of the droop slope after considering the electrochemical polarization hysteresis effect, which can provide basic parameters for subsequent internal resistance drift compensation. At the same time, this embodiment performs statistical processing on the transient internal resistance values ​​of each battery energy storage module within the internal resistance sampling stability interval to obtain the internal resistance benchmark value of each battery energy storage module. For example, this embodiment can obtain the internal resistance benchmark value of each battery energy storage module by calculating the average or median value of all transient internal resistance values ​​within the internal resistance sampling stability interval. The internal resistance benchmark value represents the standard internal resistance of the battery energy storage module under stable operating conditions without significant polarization disturbance, and serves as the reference zero point for subsequent compensation.

[0049] During the black-start operation, this embodiment continues to employ a dynamic sampling strategy to continuously collect the instantaneous terminal voltage and discharge current values ​​of each battery energy storage module. The transient internal resistance real-time value is calculated by the ratio of the terminal voltage change to the discharge current change at adjacent sampling moments. For each battery energy storage module, within each control cycle, the system subtracts the internal resistance reference value of that module from the currently collected transient internal resistance real-time value to obtain the internal resistance nonlinear drift compensation amount. This internal resistance nonlinear drift compensation amount reflects the degree of deviation of the internal resistance parameter of each module from the stable reference state. If the internal resistance nonlinear drift compensation amount is positive, it indicates that the current internal resistance is greater than the reference value. A negative value (e.g., due to increased temperature or accelerated aging) indicates an increasing trend in internal resistance. If the internal resistance nonlinear drift compensation is negative, it indicates that the current internal resistance is less than the reference value (e.g., due to a temporary weakening of the polarization effect), indicating a decreasing trend in internal resistance. Since the internal resistance often exhibits nonlinear characteristics with changes in discharge depth, temperature, aging, etc., the difference between the real-time value of the transient internal resistance and the reference value of the internal resistance of the battery energy storage module is essentially a quantification of this nonlinear drift. The system uses this difference as the internal resistance nonlinear drift compensation. The larger the absolute value of the internal resistance nonlinear drift compensation, the further the internal resistance deviates from the reference, and the greater the correction magnitude of the droop slope needs to be.

[0050] This embodiment utilizes the internal resistance nonlinear drift compensation (real-time deviation of internal resistance) to correct the current droop slope value, ensuring that the corrected droop slope can offset the power distribution deviation caused by internal resistance drift. The corrected droop slope control value is equal to the current droop slope value plus or minus the internal resistance nonlinear drift compensation multiplied by a preset compensation coefficient. When the transient real-time internal resistance value is higher than the internal resistance reference value, it indicates that the module's internal resistance is too high, resulting in less current under the same voltage droop characteristics. To achieve balanced power distribution, the droop slope needs to be appropriately increased (i.e., reducing the virtual resistance value of the droop resistor, or adjusting it in the opposite direction according to the control strategy). The system calculates the correction amount based on the pre-calibrated compensation coefficient and adds it to the current... Based on the droop slope value, the final droop slope control value used for control is obtained. Then, in this embodiment, the equivalent droop coefficient of the module is calculated based on the rated power capacity of each battery energy storage module and the compensated and corrected droop slope control value. The specific calculation method of the equivalent droop coefficient of the module includes multiplying the rated power capacity and the compensated and corrected droop slope control value and then normalizing it, or dividing the rated power capacity by the droop slope control value and then normalizing it. The specific calculation method needs to be determined according to the type of droop control strategy. Under the same common bus voltage, the larger the equivalent droop coefficient of the battery energy storage module, the higher the power allocation weight that the module should bear in the current state; conversely, the smaller the equivalent droop coefficient, the lower the droop coefficient.

[0051] This embodiment uses the equivalent droop coefficient of each battery energy storage module as a weighting factor for weighted allocation, enabling optimal power distribution among modules based on their capacity and internal resistance. The specific allocation process involves first calculating the sum of the equivalent droop coefficients of all modules. Then, for each battery energy storage module, its equivalent droop coefficient is divided by the sum of the equivalent droop coefficients of all modules to obtain its power allocation weight (this weight is between zero and one, and the sum of the weights of all battery energy storage modules is one). Finally, this embodiment multiplies the power allocation weight by the total load demand to obtain the power output value that the module should bear. Dividing this power output value by the rated power capacity of the battery energy storage module yields the power output ratio. Since the equivalent droop coefficient already includes... To compensate for uneven internal resistance, modules with high internal resistance will automatically allocate less power (or allocate more reasonable power based on droop characteristics), while modules with low internal resistance and good condition will bear more power. This achieves balanced power sharing among modules in dynamic changes, avoiding overload or underload. In this embodiment, the power output ratio is directly used to control the power electronic converters (such as DC / DC or DC / AC converters) of each battery energy storage module, adjusting the actual output power of each module. This allows each module to work collaboratively according to its current state of charge, internal resistance characteristics, and capacity during system black start, avoiding overload of some modules and underload of others, thus achieving true balanced power sharing. In the black start operation mode, the total load demand power is the sum of the power required by all load devices.

[0052] This invention provides a black-start power equalization control method for energy storage. The method includes: when the flexible boost energy storage system enters the black-start preparation state, collecting real-time charge data, terminal voltage drop rate, and polarization relaxation response time of each battery energy storage module in the flexible boost energy storage system to obtain discharge depth segment distribution data; identifying the internal resistance abrupt change inflection point of the battery energy storage module based on the discharge depth segment distribution data, and determining the polarization steady-state window based on the discharge depth position corresponding to the internal resistance abrupt change inflection point; acquiring transient internal resistance sampling data during the discharge process of each battery energy storage module, and performing droop slope oscillation risk detection based on the transient internal resistance sampling data and the polarization steady-state window to determine the internal resistance sampling stability interval; after the flexible boost energy storage system enters the black-start operation mode, establishing a linkage mapping relationship between the real-time droop slope value and the real-time charge data based on the internal resistance sampling stability interval; and dynamically compensating for the nonlinear drift of the internal resistance of each battery energy storage module according to the linkage mapping relationship to obtain the power output ratio of each battery energy storage module in the black-start operation mode. Compared with existing technologies, this method establishes a real-time linkage mapping between droop slope and state of charge by dynamically identifying the characteristics of sudden changes in battery internal resistance and polarization steady-state window, and dynamically compensates and corrects nonlinear drift of internal resistance. This allows the droop slope control value to track the actual electrochemical characteristics of the battery in real time, thereby accurately calculating the power output ratio of each module in black-start operation mode. This effectively avoids the risk of power distribution imbalance and equipment overload caused by instantaneous impedance fluctuations, and improves the power balancing accuracy and operational stability of multi-module parallel systems under black-start and emergency conditions.

[0053] It should be noted that the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0054] In one embodiment, such as Figure 2 As shown, this embodiment of the invention provides an energy storage black start power balancing control system, applied to a flexible boost energy storage system composed of multiple battery energy storage modules connected in parallel. The energy storage black start power balancing control system includes: The data acquisition module 101 is used to acquire real-time charge data, terminal voltage drop rate and polarization relaxation response time of each battery energy storage module from the flexible boost energy storage system when the flexible boost energy storage system enters the black start preparation state, so as to obtain the discharge depth segment distribution data. The steady-state analysis module 102 is used to identify the internal resistance changeover point of the battery energy storage module based on the discharge depth segment distribution data, and to determine the polarization steady-state window based on the discharge depth position corresponding to the internal resistance changeover point. The oscillation detection module 103 is used to acquire transient internal resistance sampling data during the discharge process of each battery energy storage module, and to perform droop slope oscillation risk detection based on the transient internal resistance sampling data and the polarization steady state window to determine the internal resistance sampling stable range. Linkage mapping module 104 is used to establish a linkage mapping relationship between the real-time value of the droop slope and the real-time charge data based on the internal resistance sampling stability range after the flexible boost energy storage system enters the black start operation mode. The power balancing module 105 is used to dynamically compensate for the nonlinear drift of the internal resistance of each battery energy storage module according to the linkage mapping relationship, so as to obtain the power output ratio of each battery energy storage module in the black start operation mode.

[0055] For specific limitations regarding the energy storage black-start power balancing control system, please refer to the above-described limitations regarding the energy storage black-start power balancing control method, which will not be repeated here. Those skilled in the art will recognize that the various modules and steps described in conjunction with the embodiments disclosed in this application can be implemented in hardware, software, or a combination of both. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0056] This invention provides a black-start power balancing control system for energy storage. The system uses a data acquisition module to collect real-time charge data, terminal voltage drop rate, and polarization relaxation response time of each battery energy storage module when the flexible boost energy storage system enters the black-start preparation state, obtaining discharge depth segment distribution data. A steady-state analysis module identifies the internal resistance abrupt change inflection points of the battery energy storage modules based on the discharge depth segment distribution data, and determines the polarization steady-state window based on the discharge depth position corresponding to the internal resistance abrupt change inflection point. An oscillation detection module detects the internal resistance abrupt change inflection points in each battery module. During the discharge process of the energy storage module, transient internal resistance sampling data is acquired, and droop slope oscillation risk detection is performed based on the transient internal resistance sampling data and the polarization steady-state window to determine the internal resistance sampling stability range. After the flexible boost energy storage system enters the black-start operation mode, the linkage mapping module establishes a linkage mapping relationship between the real-time value of the droop slope and the real-time charge data based on the internal resistance sampling stability range. The power balancing module dynamically compensates for the nonlinear drift of the internal resistance of each battery energy storage module according to the linkage mapping relationship to obtain the power output ratio of each battery energy storage module in the black-start operation mode. Compared with existing technologies, this system establishes a real-time linkage mapping between droop slope and state of charge by dynamically identifying the characteristics of sudden changes in battery internal resistance and polarization steady-state window, and dynamically compensates for nonlinear drift in internal resistance. This allows the droop slope control value to track the actual electrochemical characteristics of the battery in real time, thereby accurately calculating the power output ratio of each module in black-start operation mode. This effectively avoids the risk of power distribution imbalance and equipment overload caused by instantaneous impedance fluctuations, and improves the power balancing accuracy and operational stability of multi-module parallel systems under black-start and emergency conditions.

[0057] The embodiments described above are merely preferred embodiments of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various improvements and substitutions without departing from the technical principles of this invention, and these improvements and substitutions should also be considered within the scope of protection of this application. Therefore, the scope of protection of this patent application should be determined by the scope of the claims.

Claims

1. A method for equalizing black-start power control in energy storage, characterized in that, The energy storage black-start power balancing control method, applicable to a flexible boost energy storage system composed of multiple battery energy storage modules connected in parallel, includes the following steps: When the flexible boost energy storage system enters the black start preparation state, real-time charge data, terminal voltage drop rate and polarization relaxation response time of each battery energy storage module are collected from the flexible boost energy storage system to obtain discharge depth segment distribution data. Identify the abrupt change in the internal resistance of the battery storage module based on the discharge depth segment distribution data, and determine the polarization steady-state window based on the discharge depth position corresponding to the abrupt change in internal resistance, including: Extract the discharge depth location corresponding to the internal resistance abrupt change inflection point from the discharge depth segment distribution data to obtain the discharge depth inflection point. Based on the polarization relaxation response time corresponding to the discharge depth inflection point and the preset steady-state threshold, the range of continuous discharge depth segments is selected to obtain the initial time interval in which the polarization state tends to be stable. The initial time interval is time-constrained and corrected using the sampling timestamp of the internal resistance abrupt change in the inflection point to obtain the polarization steady-state window. During the discharge process of each battery energy storage module, transient internal resistance sampling data is acquired, and the risk of droop slope oscillation is detected based on the transient internal resistance sampling data and the polarization steady-state window to determine the internal resistance sampling stability range. After the flexible boost energy storage system enters the black start operation mode, a linkage mapping relationship between the real-time value of the droop slope and the real-time charge data is established based on the internal resistance sampling stability range. Based on the aforementioned linkage mapping relationship, dynamic compensation is performed on the nonlinear drift of the internal resistance of each battery energy storage module to obtain the power output ratio of each battery energy storage module in the black start operation mode, including: The current charge data of each battery energy storage module is collected based on the real-time operating status of the flexible boost energy storage system after it enters the black start operation mode. By using the aforementioned linkage mapping relationship to query the real-time estimated value of the droop slope corresponding to the current charge data, the current droop slope value of each battery energy storage module can be obtained. The reference value of the internal resistance of each battery energy storage module is determined based on the internal resistance sampling stability range, and the real-time transient internal resistance value of each battery energy storage module is continuously collected during the black start operation. Based on the real-time transient internal resistance value and the internal resistance reference value, the nonlinear drift compensation amount of the internal resistance of each battery energy storage module during the black start operation process is obtained. Based on the internal resistance nonlinear drift compensation amount, the current droop slope value is reversed and corrected to obtain the droop slope control value after compensation and correction. Based on the rated power capacity of each battery energy storage module and the droop slope control value, the module equivalent droop coefficient of each battery energy storage module under the current operating state is obtained. Using the equivalent droop coefficient of the module as a weighting factor, the total load demand power of the flexible boost energy storage system in black start operation mode is weighted and allocated using the weighting factor to obtain the power output ratio of each battery energy storage module in black start operation mode.

2. The energy storage black start power equalization control method as described in claim 1, characterized in that, The step of acquiring real-time charge data, terminal voltage drop rate, and polarization relaxation response time of each battery energy storage module from the flexible boost energy storage system to obtain discharge depth segment distribution data includes: Collect real-time charge data and terminal voltage data of each battery energy storage module, and divide each battery energy storage module into several discharge depth segments based on the real-time charge data; The rate of voltage drop of each battery energy storage module is obtained by the ratio of the difference in terminal voltage data within adjacent sampling periods to the sampling period duration. The discharge state of each battery energy storage module is marked according to the terminal voltage drop rate, and based on the discharge state, the time taken for the terminal voltage to recover from the fluctuating state to the stable state during the discharge process is monitored to obtain the polarization relaxation response time. The real-time charge data, the terminal voltage drop rate, and the polarization relaxation response duration are time-aligned and matched according to a unified time base to obtain multi-dimensional battery module state parameters. The multi-dimensional battery module state parameters are mapped to the discharge depth segment of each battery energy storage module to obtain discharge depth segment distribution data.

3. The energy storage black start power equalization control method as described in claim 2, characterized in that, The step of identifying the inflection point of internal resistance change in the battery energy storage module based on the discharge depth segment distribution data includes: Based on the discharge depth segment distribution data, the terminal voltage drop rate of each battery energy storage module within the same discharge depth segment is compared cycle by cycle to obtain rate deviation data; Based on the rate deviation data and the preset deviation threshold, the inflection point of the internal resistance change of the battery energy storage module is identified.

4. The energy storage black start power equalization control method as described in claim 2, characterized in that, The steps for obtaining transient internal resistance sampling data during the discharge process of each battery energy storage module include: Based on the distribution data of the discharge depth segments, extract the real-time data change trend of the battery energy storage module during the discharge process in each discharge depth segment, and determine the gradient of the charge change. Based on the polarization relaxation response time, the characteristics of the polarization effect dissipation time of the battery energy storage module in each discharge depth segment are determined. The dynamic sampling time sequence is determined based on the charge change gradient and the polarization effect dissipation time characteristics, and the instantaneous terminal voltage and instantaneous discharge current values ​​are synchronously collected during the discharge process of each battery energy storage module according to the dynamic sampling time sequence. The difference between the instantaneous terminal voltage values ​​at adjacent sampling times is calculated to obtain the change in terminal voltage, and the difference between the instantaneous discharge current values ​​at adjacent sampling times is calculated to obtain the change in discharge current. The transient internal resistance value is obtained based on the ratio between the change in terminal voltage and the change in discharge current. The transient internal resistance value and its corresponding discharge depth segment are arranged in chronological order to form transient internal resistance sampling data.

5. The energy storage black start power equalization control method as described in claim 1, characterized in that, The step of detecting droop slope oscillation risk based on the transient internal resistance sampling data and the polarization steady-state window, and determining the internal resistance sampling stability interval, includes: The transient internal resistance sampling data is sequentially mapped to the polarization steady-state window according to the sampling time to determine the fluctuation morphology characteristics at the edge of the polarization steady-state window; Based on the wave morphology characteristics, the droop slope oscillation risk energy density at each sampling point within the polarization steady-state window is calculated to obtain the continuous probability distribution of oscillation risk. Morphological clustering analysis was performed on the continuous probability distribution of the oscillation risk to obtain a stable candidate set of internal resistance sampling. Boundary envelope fitting is performed on the internal resistance sampling stable candidate set to obtain the internal resistance sampling stable interval.

6. The energy storage black start power equalization control method as described in claim 1, characterized in that, The step of establishing a linkage mapping relationship between the real-time value of the droop slope and the real-time charge data based on the internal resistance sampling stability interval includes: Based on the sampling time corresponding to the internal resistance sampling stability interval, the real-time charge data of each sampling time is extracted from the distribution data of the discharge depth segment to obtain the steady-state data point cloud; The steady-state data point cloud is traversed using a sliding window, and the local response coefficient sequence at the center of the sliding window with the state of charge is obtained by using a linear weighted least squares fitting method. The local response coefficient sequence is fitted with a nonlinear trend to obtain the droop slope response trend curve as the charge changes. Based on the polarization relaxation response time corresponding to the polarization steady-state window, the hysteresis effect characteristics of the internal electrochemical polarization of the battery during the black start process are determined. Based on the hysteresis effect characteristics, the droop slope response trend curve is phase-compensated and corrected to obtain a real-time estimate of the droop slope. Based on the relationship between the real-time estimated droop slope and the steady-state point cloud data, a linkage mapping relationship is established with real-time charge data as input and the real-time estimated droop slope as output.

7. The energy storage black start power equalization control method as described in claim 1, characterized in that: The nonlinear drift compensation amount of the internal resistance is the difference between the real-time value of the transient internal resistance and the reference value of the internal resistance.

8. A black-start power balancing control system for energy storage, characterized in that, The energy storage black-start power balancing control system, applicable to a flexible boost energy storage system composed of multiple battery energy storage modules connected in parallel, includes: The data acquisition module is used to collect real-time charge data, terminal voltage drop rate and polarization relaxation response time of each battery energy storage module from the flexible boost energy storage system when the flexible boost energy storage system enters the black start preparation state, so as to obtain the discharge depth segment distribution data. The steady-state analysis module is used to identify the abrupt change in the internal resistance of the battery energy storage module based on the discharge depth segment distribution data, and to determine the polarization steady-state window based on the discharge depth position corresponding to the abrupt change in internal resistance. This includes: Extract the discharge depth location corresponding to the internal resistance abrupt change inflection point from the discharge depth segment distribution data to obtain the discharge depth inflection point. Based on the polarization relaxation response time corresponding to the discharge depth inflection point and the preset steady-state threshold, the range of continuous discharge depth segments is selected to obtain the initial time interval in which the polarization state tends to be stable. The initial time interval is time-constrained and corrected using the sampling timestamp of the internal resistance abrupt change in the inflection point to obtain the polarization steady-state window. The oscillation detection module is used to acquire transient internal resistance sampling data during the discharge process of each battery energy storage module, and to perform droop slope oscillation risk detection based on the transient internal resistance sampling data and the polarization steady state window to determine the internal resistance sampling stable range. The linkage mapping module is used to establish a linkage mapping relationship between the real-time value of the droop slope and the real-time charge data based on the internal resistance sampling stability range after the flexible boost energy storage system enters the black start operation mode. A power balancing module is used to dynamically compensate for the nonlinear drift of the internal resistance of each battery energy storage module according to the aforementioned linkage mapping relationship, and to obtain the power output ratio of each battery energy storage module in the black start operation mode, including: The current charge data of each battery energy storage module is collected based on the real-time operating status of the flexible boost energy storage system after it enters the black start operation mode. By using the aforementioned linkage mapping relationship to query the real-time estimated value of the droop slope corresponding to the current charge data, the current droop slope value of each battery energy storage module can be obtained. The reference value of the internal resistance of each battery energy storage module is determined based on the internal resistance sampling stability range, and the real-time transient internal resistance value of each battery energy storage module is continuously collected during the black start operation. Based on the real-time transient internal resistance value and the internal resistance reference value, the nonlinear drift compensation amount of the internal resistance of each battery energy storage module during the black start operation process is obtained. Based on the internal resistance nonlinear drift compensation amount, the current droop slope value is reversed and corrected to obtain the droop slope control value after compensation and correction. Based on the rated power capacity of each battery energy storage module and the droop slope control value, the module equivalent droop coefficient of each battery energy storage module under the current operating state is obtained. Using the equivalent droop coefficient of the module as a weighting factor, the total load demand power of the flexible boost energy storage system in black start operation mode is weighted and allocated using the weighting factor to obtain the power output ratio of each battery energy storage module in black start operation mode.

Images