A power supply control method, device and medium based on lithium battery energy storage characteristics

Abstract

This invention discloses a power control method, device, and medium based on the energy storage characteristics of lithium batteries, belonging to the field of intelligent power control technology. The method includes: collecting operating state parameters to obtain a clean data stream; performing initial state of charge (SOC) inversion using the SOC to obtain an initial parameter package; performing interval propagation updates based on the SOC and the initial parameter package to obtain SOC estimation results and voltage uncertainty; calculating the available current boundary based on the SOC estimation results and voltage uncertainty to obtain the available power range; comparing a preset target bus voltage value with the measured bus voltage, and using a proportional-integral method to trim the available power range to obtain the final power control command. This invention achieves accurate initialization of SOC estimation by combining the open-circuit voltage-SOC-temperature correspondence with equivalent internal resistance correction for initial SOC inversion and calculating the initial SOC uncertainty.

Description

Technical Field

[0001] This invention relates to the field of intelligent power control technology, and in particular to a power control method, device and medium based on the energy storage characteristics of lithium batteries. Background Technology

[0002] With the large-scale integration of renewable energy and the increasing fluctuations in power load, lithium batteries, due to their high energy density and cycle life, are widely used in energy storage and power management. In recent years, battery management systems have gradually developed a state-of-charge estimation method based on model and data fusion, and have been promoted and applied in scenarios such as grid dispatching and distributed energy control, providing technical support for achieving efficient energy dispatching and stable supply.

[0003] However, existing battery control methods generally rely on a single model or empirical parameters in the state of charge estimation process, which makes it difficult to accurately characterize the uncertainty of battery operation and leads to bias in the estimation results. At the same time, the dynamic correction of voltage residual and hysteresis effect is often ignored in the power control process, which can easily cause plateau drift and power distribution instability, making it difficult to achieve precise control and safety assurance of battery operation. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a power control method based on the energy storage characteristics of lithium batteries to solve the problems of insufficient accuracy in state of charge estimation and lack of dynamic correction mechanism in power control.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a power control method based on the energy storage characteristics of lithium batteries, which includes: collecting operating status parameters, performing time alignment and filtering processing, and obtaining a clean data stream;

[0008] By using clean data stream, the operating state parameters under static and steady conditions are obtained, the initial state of charge is inverted to obtain the initial state of charge, and joint fitting analysis is performed to obtain the initial parameter package.

[0009] Based on the clean data stream and initial parameter package, interval propagation update is performed to obtain the state of charge estimate and voltage uncertainty;

[0010] Based on the state of charge estimation and voltage uncertainty, the available current boundary is calculated to obtain the available power range;

[0011] The preset target value of the bus voltage is compared with the measured bus voltage. The available power range is then clipped using proportional-integral conversion to obtain the final power control command.

[0012] As a preferred embodiment of the power control method based on lithium battery energy storage characteristics described in this invention, the specific steps for acquiring operating status parameters, performing time alignment and filtering processing, and obtaining a clean data stream are as follows:

[0013] Collect the operating status parameters of the lithium battery, perform time alignment and filtering on the operating status parameters, and obtain the filtered operating status parameters;

[0014] The filtered operating status parameters are processed to remove outliers, resulting in a clean data stream.

[0015] As a preferred embodiment of the power control method based on lithium battery energy storage characteristics described in this invention, the steps of obtaining operating state parameters under static and stable conditions through a clean data stream, performing initial state of charge inversion, and obtaining the initial state of charge are as follows:

[0016] The open-circuit voltage, state of charge, and temperature of the battery are matched with the operating status parameters in the clean data stream to obtain the operating status parameters under static and stable conditions.

[0017] The initial state of charge is obtained by inverting the table data of operating state parameters, open circuit voltage, state of charge and temperature under static steady conditions.

[0018] As a preferred embodiment of the power control method based on lithium battery energy storage characteristics described in this invention, the specific steps for performing joint fitting analysis to obtain the initial parameter package are as follows:

[0019] The internal resistance voltage drop correction is calculated using clean data stream to obtain the theoretical voltage. Based on the initial state of charge and clean data stream, joint fitting analysis is performed by minimizing the deviation between the theoretical voltage and the measured voltage to obtain battery performance parameters and hysteresis parameters, and to establish an initial parameter package.

[0020] As a preferred embodiment of the power control method based on lithium battery energy storage characteristics described in this invention, the step of performing interval propagation updates based on clean data streams and initial parameter packets specifically includes the following steps:

[0021] Based on the initial parameter package and the clean data stream, the clean data stream is integrally calculated and combined with the initial parameter package to complete the interval propagation update of the initial state of charge, thereby obtaining the state of charge interval.

[0022] As a preferred embodiment of the power control method based on lithium battery energy storage characteristics described in this invention, the specific steps for obtaining the state of charge estimate and voltage uncertainty are as follows:

[0023] The state of charge interval is combined with the hysteresis parameters in the initial parameter package to perform hysteresis correction on the state of charge interval, and the hysteresis-corrected state of charge interval is obtained.

[0024] The hysteresis-corrected state of charge interval is compared with the clean data stream to calculate the voltage residual. Based on the voltage residual, the hysteresis-corrected state of charge interval is dynamically shrunk to obtain the dynamically adjusted state of charge interval.

[0025] The state of charge (SOC) estimate, including the upper and lower bounds of the SOC, is calculated by dynamically adjusting the SOC interval. The voltage uncertainty is then obtained by looking up table data for open-circuit voltage, SOC, and temperature from the initial parameter package.

[0026] As a preferred embodiment of the power control method based on lithium battery energy storage characteristics described in this invention, the specific steps for calculating the available current boundary and obtaining the available power range based on the state of charge estimation results and voltage uncertainty are as follows:

[0027] Set failure probability constraints, calculate the state of charge estimation and voltage uncertainty based on the failure probability constraints, and obtain the state of charge estimation and voltage uncertainty that satisfy the failure probability constraints;

[0028] The available current boundaries in the charging direction and the available current boundaries in the discharging direction are calculated using the state of charge estimation and voltage uncertainty that satisfy the failure probability constraint.

[0029] By combining the available current boundary with the clean data stream, the power boundaries in the charging and discharging directions are obtained. Then, the interval intersection calculation is performed with the temperature derating power limit and the device rated power to obtain the available power range.

[0030] As a preferred embodiment of the power control method based on lithium battery energy storage characteristics described in this invention, the steps of comparing the preset target bus voltage with the measured bus voltage, and then cropping the available power range through proportional-integral scaling to obtain the final power control command are as follows:

[0031] Set a target value for the bus voltage, calculate the difference between the target value and the measured bus voltage at the same time point, and use the difference between the target value and the measured bus voltage as the bus voltage comparison data.

[0032] The power command is obtained by weighted summation of the bus voltage comparison data using the proportional-integral method.

[0033] The power command is limited by its ramp rate, compared with the available power range, and the power command with the ramp rate limited is pruned to obtain the final power control command.

[0034] In a second aspect, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the power control method based on the energy storage characteristics of a lithium battery as described in the first aspect of the present invention.

[0035] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the power control method based on the energy storage characteristics of a lithium battery as described in the first aspect of the present invention.

[0036] The beneficial effects of this invention are as follows: By collecting lithium battery operating state parameters and performing time alignment, filtering, and outlier removal, a clean data stream is obtained, thus achieving the accuracy and reliability of the input data; by performing initial state of charge inversion based on the open-circuit voltage-state of charge-temperature correspondence combined with equivalent internal resistance correction under static steady-state conditions, and calculating the initial state of charge uncertainty, accurate initialization of state of charge estimation is achieved; by performing interval propagation updates based on the clean data stream and initial parameter package, and combining voltage residuals to dynamically shrink and translate the state of charge interval, real-time interval estimation and uncertainty constraints of the battery state of charge are achieved. Attached Figure Description

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a flowchart of a power control method based on the energy storage characteristics of lithium batteries.

[0039] Figure 2 This is a flowchart illustrating the process of processing operational status parameters into a clean data stream.

[0040] Figure 3 This is a flowchart for the initial state of charge inversion and consistency verification.

[0041] Figure 4 This is a flowchart for state of charge estimation and power control. Detailed Implementation

[0042] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0043] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0044] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0045] Reference Figures 1-4 This is one embodiment of the present invention, which provides a power control method based on the energy storage characteristics of lithium batteries, including the following steps:

[0046] S1. Collect operating status parameters, perform time alignment and filtering, and obtain clean data stream.

[0047] The system collects the operating status parameters of the lithium battery, performs time alignment and filtering on the operating status parameters, obtains the filtered operating status parameters, removes outliers from the filtered operating status parameters, and obtains a clean data stream.

[0048] Furthermore, the individual cell voltage, individual cell temperature, cluster current, bus voltage, and ambient temperature of the lithium battery are collected at a sampling period (e.g., 1 second) to obtain operating status parameters. The collection timestamps of all operating status parameters are recorded and compared with a reference. For sampling points with deviations, linear interpolation is used to correct the operating status parameters to the reference time point, so that different operating status parameters correspond to the same moment under a unified time base. Median filtering is used to remove impulse interference in the operating status parameters, and a first-order low-pass filtering method is used to smooth the operating status parameters, thereby reducing high-frequency noise and preserving the true trend of change. A fixed-length time period (e.g., 30 seconds) that slides forward from the sampling period is used as a sliding window. Statistical analysis is performed on the operating status parameters within the sliding window. Points that deviate too much from the median are identified as outliers and replaced by linear interpolation of adjacent points, thereby eliminating outliers and obtaining a clean data stream.

[0049] S2. Obtain the operating state parameters under static and stable conditions through the clean data stream, perform initial state of charge inversion, obtain the initial state of charge, perform joint fitting analysis, and obtain the initial parameter package.

[0050] By combining the lookup data corresponding to the battery's open-circuit voltage, state of charge, and temperature with the operating status parameters in the cleanroom data stream, the operating status parameters under static and stable conditions can be obtained.

[0051] Furthermore, the open-circuit voltage of the battery is retrieved from a lookup table corresponding to its state of charge and temperature. The lookup table records the characteristic curves of the battery open-circuit voltage under different states of charge and temperatures. Sampling points corresponding to individual cell voltage, individual cell temperature, and cluster current are selected from the clean data stream. The amplitude threshold is obtained through the static stability condition judgment standard provided by the GB standard. The change rate threshold is obtained based on the sampling period and the allowable static stability fluctuation range. The change rate is obtained by calculating the ratio of the difference in cluster current between adjacent sampling points to the sampling period. By performing amplitude threshold detection and change rate threshold detection on the cluster current, time periods in which the current is close to zero and remains within a small range of fluctuation are selected as candidate intervals for static stability conditions.

[0052] Within the candidate range of static and stable operating conditions, the corresponding temperature curve is first located in the lookup table data corresponding to the open circuit voltage, state of charge, and temperature provided at the time of battery delivery based on the individual cell temperature. Then, the individual cell voltage in the operating status parameters under static and stable operating conditions is compared point by point with the open circuit voltage on the temperature curve. If the voltage deviation is less than the allowable deviation range (e.g., ±10 mV), it is confirmed that the condition of the correspondence between open circuit voltage, state of charge, and temperature is met. Within the time period that meets the condition, the corresponding individual cell voltage, individual cell temperature, and cluster current data are extracted as the operating status parameters under static and stable operating conditions.

[0053] The initial state of charge is obtained by inverting the operating state parameters under static and steady conditions with the lookup table data.

[0054] Furthermore, during the static steady-state operating period, the average values ​​of individual cell voltage, individual cell temperature, and cluster current are calculated to obtain representative values ​​of the operating state parameters under static steady-state conditions. The average value of the individual cell voltage is corrected by summing the product of the initial equivalent internal resistance (provided in the battery specifications at the factory) and the average cluster current with the average individual cell voltage. This yields a voltage estimate consistent with the open-circuit condition, where the cluster current is zero and the voltage is determined solely by the state of charge (SOC) and temperature. The adjacent temperature curves are located in the lookup table data corresponding to the SOC and temperature, and monotonically interpolated to obtain the isothermal open-circuit voltage and SOC correspondence for the corresponding individual cell temperature average. A robust weighted inversion method is used to perform initial SOC inversion during the operating state parameter time period under static steady-state conditions, and the initial SOC uncertainty is calculated. The expression is:

[0055] ;

[0056] ;

[0057] ;

[0058] in, This represents the optimal estimate obtained from the initial state of charge inversion. Candidate variables representing the state of charge. Indicates the length of the time window. Indicates the start time of the static steady-state operating condition period. Indicates at time The individual unit voltage, Indicates the state of charge and temperature Open circuit voltage below, Indicates at time Cluster current, This represents the initial value of the equivalent internal resistance. Represents a robust loss function. Represents the temperature weighting function. This represents the prior constraint weight coefficient. This represents the prior value of the state of charge obtained based on the current integral. This represents the average temperature over a time period under static and steady-state operating conditions. This represents the uncertainty of the initial state of charge. Indicates in and Below, the slope of the open-circuit voltage-state-of-charge curve, This represents the standard deviation of voltage measurement noise. Represents the regularization coefficient. This indicates the deviation between the measured voltage and the predicted voltage. Indicates the threshold for a transition.

[0059] It should be noted that, It was obtained by looking up table data corresponding to open-circuit voltage, state of charge, and temperature provided by the battery manufacturer. It is obtained through parameters provided by the manufacturer. It is expressed in Gaussian kernel function form. It is obtained through adaptive adjustment based on real-time calculation of the degree of current integral drift and voltage residual variance. This is used to avoid divergence in uncertainty calculations when the voltage-state-of-charge slope is too small or the voltage noise is too high. Based on the proportional setting of the voltage residual noise level, the predicted voltage is obtained by substituting the candidate state of charge and instantaneous temperature into the open-circuit voltage-state-of-charge-temperature correspondence table data to obtain the open-circuit voltage. Then, the internal resistance voltage drop is calculated by combining the cluster current in the clean data stream and the initial value of the equivalent internal resistance. The difference between the open-circuit voltage and the internal resistance voltage drop is obtained as the predicted voltage. It is determined based on voltage residual statistics. During the static and stable operating period, the standard deviation and median absolute deviation of the residual sequences between the predicted and measured voltages are statistically analyzed. The smaller value is taken when the residual fluctuation level is small. To enhance sensitivity, a larger value is chosen when the residual fluctuation level is large. The value is used to suppress the influence of outliers.

[0060] The voltage estimate consistent with the open-circuit condition is used to perform a one-dimensional monotonically reverse lookup of the correspondence between the isothermal open-circuit voltage and the state of charge. By performing a binary search interval by interval, it is determined whether the optimal estimate of the initial state of charge inversion falls within the corresponding interval within the voltage matching tolerance range (e.g., ±5 mV). Candidate values ​​of the initial state of charge are obtained. A consistency check is performed on the candidate values ​​of the initial state of charge. The consistency check requires that the voltage deviation of the lookup data corresponding to the open-circuit voltage, state of charge and temperature under the same temperature condition does not exceed the consistency tolerance range (e.g., ±10 mV). If the consistency check passes, the initial state of charge is output. If it fails, the initial state of charge is recalculated after extending the time window or increasing the outlier removal intensity within the time period of the operating state parameters under static steady condition.

[0061] Based on the initial state of charge and clean data stream, a joint fitting analysis is performed to obtain battery performance parameters and hysteresis parameters, and an initial parameter package is established.

[0062] Furthermore, a continuous sampling time period in the clean data stream is selected as the fitting window. Within the fitting window, the charge amount in each sampling period is obtained by multiplying the cluster current value in the clean data stream by the sampling period. The total charge amount within the window is obtained by continuous accumulation. Then, the ratio of the total charge amount to the nominal battery capacity is converted to obtain the change in state of charge. At each sampling moment, the initial state of charge is used as the reference value. The state of charge at the previous moment is superimposed with the change in state of charge in the current sampling period to obtain the state of charge at the current moment. Within the fitting window, the state of charge at the current moment and the temperature are used as input parameters. The system retrieves table data corresponding to open-circuit voltage, state of charge, and temperature, and performs interpolation between adjacent data points to obtain a theoretical voltage sequence. It then performs difference calculations between the theoretical voltage sequence and the voltage signal in the clean data stream, point by point, to obtain a voltage deviation sequence. The coulombic efficiency is determined by correcting the deviation between the current state of charge and the voltage measurement result. The voltage measurement result is the voltage signal in the clean data stream. The cluster current direction in the clean data stream is divided: when the cluster current value is greater than zero, the corresponding sampling point is divided into a discharge interval; when the cluster current value is less than zero, the corresponding sampling point is divided into a charging interval. When the cluster current value is near zero, the corresponding sampling points are divided into a steady-state interval. Within the fitting window, charging and discharging interval sample sets are formed. The difference between the predicted voltage and the measured voltage is calculated for each sampling point in both sets to obtain the voltage deviation statistic. Under the same state of charge, the mean voltage deviation for the charging and discharging intervals is statistically analyzed, and the difference is compared. The voltage deviation difference is used as a characterization value of the voltage hysteresis effect. The hysteresis parameter is obtained through nonlinear fitting using the relationship between the characterization value of the voltage hysteresis effect and the current amplitude. The mean, root mean square, and standard deviation of the voltage deviation sequence are calculated. The rationality of battery performance parameters, consisting of nominal capacity, equivalent internal resistance, coulombic efficiency, and hysteresis parameters, is determined. When the deviation statistics fall within the allowable range (e.g., the mean does not exceed ±10 mV, and the root mean square does not exceed 20 mV), the battery performance parameters are considered reasonable. Otherwise, the battery performance parameters are readjusted to gradually reduce the difference between the state of charge and the voltage signal in the clean data stream. The battery performance parameters, including nominal capacity, equivalent internal resistance, and coulombic efficiency, the battery open-circuit voltage-state of charge lookup table data, the temperature derating power limit curve, the equipment rated power, and the hysteresis parameters are uniformly organized to obtain the initial parameter package.

[0063] S3. Based on the clean data stream and initial parameter package, perform interval propagation update to obtain the state of charge estimate and voltage uncertainty.

[0064] Based on the initial parameter package and the clean data stream, the current and time in the clean data stream are integrated and calculated. Combined with the capacity information in the initial parameter package, the interval propagation update of the state of charge is completed to obtain the state of charge interval.

[0065] Furthermore, during interval propagation updates, the initial state of charge (SBC) is first used as the center value, and an initial propagation interval is constructed based on the uncertainty of the initial SBC to ensure that the propagation starting point covers the credible range of the initial estimate. The current value and timestamp at the current sampling moment are obtained from the clean data stream. The sampling period (e.g., 1 second) is obtained through adjacent timestamps. The current value is represented as a current interval based on the upper bound of the current sensor error. The product of the current interval and the sampling period is used as the charge interval. When there are missing sampling points within the sampling period, linear interpolation is used to recover the missing points before performing interval multiplication to ensure that the charge interval covers the true values. Fixed tolerances are set for nominal capacity and coulombic efficiency within the initial parameter package: 5% for nominal capacity and 1% for coulombic efficiency. The nominal capacity interval and coulombic efficiency interval are obtained. The charge interval is normalized according to the nominal capacity interval and coulombic efficiency interval to obtain the upper and lower bounds of the SBC increment. The propagation is conservative and reproducible by taking the upper and lower bounds of the entire combination. The expression is:

[0066] ;

[0067] ;

[0068] in, Indicates the first The state-of-charge interval at the end of each sampling period. Indicates nominal capacity, Indicates Coulomb efficiency. Indicates the first The charge range calculated within each sampling period Denotes the set of all combinations. This represents the upper and lower bounds of the Coulomb efficiency range. Indicates the upper and lower bounds of the nominal capacity range. Indicates the upper and lower bounds of the charge range. Represents taking a set The minimum and maximum values.

[0069] It should be noted that the nominal capacity tolerance of 5% is determined based on the capacity consistency distribution range given by the battery, and the coulombic efficiency tolerance of 1% is determined by combining the accuracy index of the current sensor and the level of side reaction loss of the battery during charging and discharging, thereby ensuring that the interval propagation update covers the reasonable uncertainty in actual operation.

[0070] Based on the influence of current direction on the monotonicity of the state of charge, subtraction is performed on the state of charge interval in the discharge direction, and addition is performed in the charging direction. When the current interval crosses zero, it is decomposed into charging and discharging intervals for propagation, and the minimum outer bound is taken to avoid erroneous contraction caused by directional uncertainty. This is expressed as:

[0071] ;

[0072] ;

[0073] ;

[0074] in, Indicates the first The state-of-charge interval at the end of each sampling period. Indicates the first The range of state of charge increment within each sampling period This represents the increment of the state of charge obtained by converting the negative current sub-interval when the current interval is negative (charging interval). This represents the increment of the state of charge obtained by converting the positive current sub-interval when the current interval is positive (discharge interval). This means that the calculated state of charge interval is limited to the range [0,1]. This represents the minimum bounding interval operator, which takes the minimum lower bound and maximum upper bound of the two input intervals and generates the minimum interval that covers both.

[0075] By combining the state-of-charge interval with the hysteresis parameters in the initial parameter package, hysteresis correction of the state-of-charge interval is performed to suppress plateau drift, and the hysteresis-corrected state-of-charge interval is obtained.

[0076] In the initial parameter package, a fixed value (e.g., the median value within the statistical range of historical operating data) is configured as the decay time constant. The hysteresis state is set to 0 at the initial moment. In each sampling period, monotonically accumulated current is performed according to the current direction in the clean data stream (increases according to the charging coefficient in the charging direction and decreases according to the discharging coefficient in the discharging direction). At the end of each sampling period, an exponential decay update is performed according to the decay time constant to obtain the hysteresis state. Based on the deviation difference between the voltage in the charging interval and the voltage in the discharging interval under the same charge state window in the clean data stream, and combined with the current amplitude, a weighted ratio is calculated to obtain the hysteresis sensitivity coefficient to voltage.

[0077] Furthermore, the hysteresis state is updated within a sampling period based on the current direction in the clean data stream. When the current is in the charging direction, it is accumulated according to the charging coefficient; when the current is in the discharging direction, it is accumulated according to the discharging coefficient. Simultaneously, the hysteresis state of the previous moment is exponentially decayed according to the decay time constant to obtain the hysteresis state at the current sampling moment. The product of the hysteresis state and the hysteresis sensitivity coefficient to voltage is taken as the voltage hysteresis range. The voltage hysteresis interval is mapped to the state-of-charge correction interval through the slope interval. The expression is:

[0078] ;

[0079] in, This represents the upper bound of the slope of the open-circuit voltage-state-of-charge curve. This represents the lower bound of the slope of the open-circuit voltage-state-of-charge curve.

[0080] It should be noted that the open-circuit voltage-state-of-charge curve was obtained from the battery's lookup table data.

[0081] The state of charge interval is monotonically shifted according to the direction of the current, and the charging direction shifts the state of charge interval upward. The discharge direction shifts the state of charge range downwards. When the current fluctuates in the near-zero range, two translation results are calculated and the minimum outer bound is taken. Then, the minimum outer bound is truncated at the [0,1] boundary and slightly exponentially smoothed with the interval of the previous time step to improve continuity, so as to obtain the charged state interval after hysteresis correction.

[0082] The hysteresis-corrected state of charge interval is compared with the voltage in the clean data stream to calculate the voltage residual. Based on the voltage residual, the hysteresis-corrected state of charge interval is dynamically shrunk to obtain the dynamically adjusted state of charge interval.

[0083] Furthermore, the hysteresis-corrected state-of-charge interval is mapped point by point to the open-circuit voltage interval. Based on the current direction in the clean data stream and the equivalent internal resistance in the initial parameter package, the voltage is compensated for internal resistance to obtain the predicted voltage interval. When the predicted voltage interval is completely higher than the voltage signal in the clean data stream (after deducting the upper limit of measurement uncertainty), the voltage residual is determined to be negative. When the predicted voltage interval is completely lower than the voltage signal in the clean data stream (after deducting the upper limit of measurement uncertainty), the voltage residual is determined to be positive. When the two overlap within the measurement uncertainty range, the voltage residual is determined to be acceptable. A residual tolerance band (e.g., ±15 mV) is set. When the voltage residual interval is completely within the residual tolerance band, it is not contracted, and only passability is recorded.

[0084] It should be noted that the measurement uncertainty is determined based on the measurement accuracy given in the sensor specifications.

[0085] When the voltage residual interval is positive, it indicates that the voltage corresponding to the hysteresis-corrected state of charge interval is too low, and it should be shifted and narrowed towards a higher state of charge. When the voltage residual interval is negative, it indicates that the voltage corresponding to the hysteresis-corrected state of charge interval is too high, and it should be shifted and narrowed towards a lower state of charge. The voltage residual interval is converted into the upper and lower bounds of the state of charge correction amount using the open-circuit voltage-state of charge slope upper and lower bounds. When the residual is small, the upper bound of the slope is used for conversion; when the residual is large, the lower bound of the slope is used. A maximum single-step shrinkage ratio (e.g., not exceeding 30% of the current interval width) and a minimum interval width threshold (e.g., 0.5% state of charge) are set to prevent jitter or excessive convergence. In the decision direction, the hysteresis-corrected state of charge interval is adjusted... The upper and lower bounds of the state of charge correction amount are shifted in the body direction to obtain the translated candidate interval. Then, the intersection of the translated candidate interval and the hysteresis-corrected state of charge interval is calculated. If the intersection is not empty, the intersection is the candidate shrinking interval. If the intersection is empty, the interval with the minimum width centered at the midpoint of the two intervals is taken as the candidate shrinking interval. Finally, the candidate shrinking interval is truncated at the [0,1] boundary to obtain a candidate shrinking interval that is closer to the measurement. In order to suppress the interval jitter caused by sampling jitter and noise, the upper and lower bounds of the candidate shrinking interval and the hysteresis-corrected state of charge interval are lightly exponentially smoothed (e.g., the weight of the previous time step is 0.9 and the weight of the current result is 0.1). After smoothing, the [0,1] truncation is performed again to obtain the dynamically adjusted state of charge interval.

[0086] It should be noted that the weight of the previous time step was set to 0.9, and the weight of the current result was set to 0.1. This ratio corresponds to a time constant of about 10 sampling periods, which means that new data only affects the result by 10% each time. This is consistent with the dynamic characteristics of battery state estimation and can achieve a balance between stability and response speed.

[0087] The estimated state of charge (SOC) results, including the upper and lower bounds of the SOC, are obtained by calculating the SOC range after dynamic adjustment. The voltage uncertainty is then obtained by combining the voltage-SOC lookup table relationship in the initial parameter package.

[0088] Furthermore, the dynamically adjusted state of charge (SOC) interval is used as the SOC estimation result at the current moment. At the temperature level corresponding to the unit temperature in the clean data stream, linear interpolation is performed on both ends of the dynamically adjusted SOC interval using the voltage-SOC lookup table relationship in the initial parameter package to obtain the open-circuit voltage corresponding to the lower boundary and the upper boundary of the SOC interval. Based on the monotonicity of the lookup table relationship, the smaller of the two voltages is taken as the lower boundary of the voltage interval, and the larger of the two voltages is taken as the upper boundary of the voltage interval to obtain the isothermal open-circuit voltage interval. The half-width of the isothermal open-circuit voltage interval is taken as the voltage uncertainty component obtained by mapping the SOC uncertainty. The upper limit of the voltage uncertainty is obtained through the sensor specification. The sum of the voltage uncertainty component and the upper limit of the voltage uncertainty is taken as the voltage uncertainty at the current moment.

[0089] S4. Based on the state of charge estimation and voltage uncertainty, calculate the available current boundary and obtain the available power range.

[0090] Based on industry functional safety standards, reliability requirements of application scenarios, and statistical analysis of collected data, failure probability constraints are set. One-sided 1% is selected as the confidence multiplier, which ensures that the state of charge estimation result and voltage uncertainty meet safety requirements at a 99% confidence level. The product of voltage uncertainty and confidence multiplier is used as the voltage safety margin, and the product of half-width of state of charge estimation result and confidence multiplier is used as the state of charge safety margin. In the charging direction, the sum of the upper bound of state of charge estimation result and state of charge safety margin is used as the protective state of charge for charging, and in the discharging direction, the difference between the lower bound of state of charge estimation result and state of charge safety margin is used as the conservative state of charge for discharging.

[0091] By using the voltage-state-of-charge lookup table relationship in the initial parameter package, linear interpolation is performed using the conservative state of charge for charging and the conservative state of charge for discharging respectively to obtain the conservative open-circuit voltage for charging and discharging. The voltage safety margin is used as the tightened allowable voltage threshold. In the charging direction, the difference between the upper limit of the single-cell voltage and the voltage safety margin is used as the upper limit of the allowable voltage in the charging direction. In the discharging direction, the lower limit of the single-cell voltage and the voltage safety margin are used as the lower limit of the allowable voltage in the discharging direction. In the charging direction, the maximum charging current amplitude not exceeding the upper limit of the allowable voltage in the charging direction is selected by using the equivalent internal resistance and the conservative open-circuit voltage for charging in the initial parameter package. In the discharging direction, the maximum discharging current amplitude not lower than the lower limit of the allowable voltage in the discharging direction is selected by using the equivalent internal resistance and the conservative open-circuit voltage for discharging in the initial parameter package. The candidate values ​​of the charging current boundary under voltage constraints and the candidate values ​​of the discharging current boundary under voltage constraints are obtained.

[0092] Within the sampling period, the product of current and time is converted into the state of charge increment using the nominal capacity and coulombic efficiency in the initial parameter package. In the charging direction, the current is limited to ensure that the conservative state of charge plus the state of charge increment does not exceed 1. In the discharging direction, the current is limited to ensure that the conservative state of charge minus the state of charge increment is not less than 0. Thus, candidate values ​​for the charging current boundary under state of charge constraints and candidate values ​​for the discharging current boundary under state of charge constraints are obtained respectively. The smaller of the amplitudes of the candidate values ​​for the charging current boundary under voltage constraints and the candidate values ​​for the charging current boundary under state of charge constraints is taken as the available current boundary in the charging direction. Similarly, the smaller of the amplitudes of the candidate values ​​for the discharging current boundary under voltage constraints and the candidate values ​​for the discharging current boundary under state of charge constraints is taken as the available current boundary in the discharging direction.

[0093] Based on the bus voltage in the clean data stream, the arithmetic mean of three sampling points is selected as the measured bus voltage. The temperature derating power limit curve is read from the initial parameter package, and the individual unit temperature in the clean data stream is used as input to look up the table to obtain the temperature derating power limit corresponding to the current temperature. The product of the available current boundary in the charging direction and the measured bus voltage is used as the candidate value of the power boundary in the charging direction, and the product of the available current boundary in the discharging direction and the measured bus voltage is used as the candidate value of the power boundary in the discharging direction. The absolute value of the power boundary candidate value is compared with the temperature derating power limit, and the smaller one, which does not exceed the temperature derating power limit, is taken as the candidate value of the power boundary under temperature constraint. Limiting processing is performed on the charging direction and the discharging direction respectively. The candidate value of the power boundary under temperature constraint is compared with the rated power of the equipment, and the smaller one, which does not exceed the rated power of the equipment, is taken as the final power boundary in that direction. The charging direction is recorded as the power boundary in the charging direction, and the discharging direction is recorded as the power boundary in the discharging direction. The power boundary in the charging direction is used as the lower boundary of the available power range, and the power boundary in the discharging direction is used as the upper boundary of the available power range, thus forming the available power range at the current moment.

[0094] S5. Compare the preset target value of the bus voltage with the measured bus voltage, and use proportional-integral conversion to cut the available power range to obtain the final power control command.

[0095] The target value of the bus voltage is set at the fixed value of the rated DC bus voltage of the energy storage device. The difference between the target value and the measured bus voltage is taken as the voltage deviation. The product of the voltage deviation and the proportional coefficient is taken as the proportional quantity. The voltage deviation is accumulated over the sampling time to obtain the cumulative sum. The product of the cumulative sum and the integral coefficient is taken as the integral term state. The sum of the proportional quantity and the integral term state is taken as the uncut power command. A ramp rate limit is applied to the uncut power command to ensure that the power change in adjacent sampling periods does not exceed the preset rate limit, thus obtaining the rate-limited power command. The rate-limited power command is compared with the available power range. If the uncut power command is within the available power range, it is directly taken as a candidate output. If it exceeds the upper limit of the available power range, it is truncated to the upper limit of the available power range. If it is below the lower limit of the available power range, it is raised to the lower limit of the available power range. The truncated value is recorded as the cut power command. The cut power command is taken as the final power control command.

[0096] It should be noted that the proportional coefficient is obtained by gradually changing the target value of the bus voltage (e.g., increasing it by 1V or 2V) during the commissioning of the energy storage device, and recording the required power adjustment (e.g., +10kW or +20kW) under this voltage deviation. The proportional coefficient is obtained by the ratio of the voltage deviation to the power change. The integral coefficient is adjusted during the commissioning phase of the energy storage device through multiple sets of step signal response processes to ensure that the convergence speed and overshoot of the bus voltage near the target value meet the design requirements (e.g., when the convergence speed is too slow, the integral coefficient is appropriately increased; when oscillation occurs, the integral coefficient is appropriately decreased). The preset rate limit is determined by combining the maximum allowable current change rate of the battery cell with the DC bus voltage and is corrected by combining the operating data to ensure that excessive power surges are avoided while meeting the response speed requirements.

[0097] This embodiment also provides a computer device applicable to the power control method based on the energy storage characteristics of lithium batteries, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the power control method based on the energy storage characteristics of lithium batteries as proposed in the above embodiment.

[0098] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0099] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the power control method based on the energy storage characteristics of a lithium battery as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0100] In summary, this invention achieves accuracy and reliability of input data by: acquiring lithium battery operating state parameters and performing time alignment, filtering, and outlier removal to obtain a clean data stream; performing initial state of charge inversion based on the open-circuit voltage-state of charge-temperature correspondence combined with equivalent internal resistance correction under static steady-state conditions, and calculating the initial state of charge uncertainty; and performing interval propagation updates based on the clean data stream and initial parameter packets, combined with dynamic contraction and translation of the state of charge interval using voltage residuals, thereby achieving real-time interval estimation and uncertainty constraint of the battery's state of charge.

[0101] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A power control method based on the energy storage characteristics of lithium batteries, characterized in that: include, Collect operational status parameters, perform time alignment and filtering, and obtain a clean data stream; By using clean data stream, the operating state parameters under static and steady conditions are obtained, the initial state of charge is inverted to obtain the initial state of charge, and joint fitting analysis is performed to obtain the initial parameter package. Based on the clean data stream and initial parameter package, interval propagation update is performed to obtain the state of charge estimate and voltage uncertainty; Based on the state of charge estimation and voltage uncertainty, the available current boundary is calculated to obtain the available power range; The preset target value of the bus voltage is compared with the measured bus voltage. The available power range is then cut off through proportional-integral conversion to obtain the final power control command. The interval propagation update includes: constructing an initial state of charge interval, determining the charge interval by combining the current value, sampling period and upper bound of current error, performing normalization processing under the constraints of nominal capacity interval and coulomb efficiency interval, and updating the state of charge interval by taking the upper and lower bounds of the full combination. The specific steps for obtaining the state of charge estimate and voltage uncertainty are as follows: combining the state of charge interval with the hysteresis parameter in the initial parameter package, performing hysteresis correction on the state of charge interval, and obtaining the hysteresis-corrected state of charge interval. The hysteresis-corrected state of charge interval is compared with the clean data stream to calculate the voltage residual. Based on the voltage residual, the hysteresis-corrected state of charge interval is dynamically shrunk to obtain the dynamically adjusted state of charge interval. The state of charge (SOC) estimate, including the upper and lower bounds of the SOC interval, is calculated by dynamically adjusting the SOC interval. The voltage uncertainty is obtained by looking up the open-circuit voltage, SOC, and temperature data in the initial parameter package. The steps for calculating the available current boundary and obtaining the available power range based on the state of charge estimation results and voltage uncertainty are as follows: setting failure probability constraints, calculating the state of charge estimation and voltage uncertainty through failure probability constraints, and obtaining the state of charge estimation and voltage uncertainty that satisfy the failure probability constraints. The available current boundaries in the charging direction and the available current boundaries in the discharging direction are calculated using the state of charge estimation and voltage uncertainty that satisfy the failure probability constraint. By combining the available current boundary with the clean data stream, the power boundaries in the charging and discharging directions are obtained. Then, the interval intersection calculation is performed with the temperature derating power limit and the device rated power to obtain the available power range.

2. The power control method based on the energy storage characteristics of lithium batteries as described in claim 1, characterized in that: The collected operational status parameters are time-aligned and filtered to obtain a clean data stream. The specific steps are as follows: Collect the operating status parameters of the lithium battery, perform time alignment and filtering on the operating status parameters, and obtain the filtered operating status parameters; The filtered operating status parameters are processed to remove outliers, resulting in a clean data stream.

3. The power control method based on the energy storage characteristics of lithium batteries as described in claim 2, characterized in that: The process involves acquiring operating state parameters under static and stable conditions via a clean data stream, performing initial state of charge inversion, and obtaining the initial state of charge. The specific steps are as follows: The open-circuit voltage, state of charge, and temperature of the battery are matched with the operating status parameters in the clean data stream to obtain the operating status parameters under static and stable conditions. The initial state of charge is obtained by inverting the table data of operating state parameters, open circuit voltage, state of charge and temperature under static steady conditions.

4. The power control method based on the energy storage characteristics of lithium batteries as described in claim 3, characterized in that: The steps for performing joint fitting analysis and obtaining the initial parameter package are as follows: The internal resistance voltage drop correction is calculated using clean data stream to obtain the theoretical voltage. Based on the initial state of charge and clean data stream, joint fitting analysis is performed by minimizing the deviation between the theoretical voltage and the measured voltage to obtain battery performance parameters and hysteresis parameters, and to establish an initial parameter package.

5. The power control method based on the energy storage characteristics of lithium batteries as described in claim 4, characterized in that: The interval propagation update based on the clean data stream and initial parameter packet involves the following steps: Based on the initial parameter package and the clean data stream, the clean data stream is integrally calculated and combined with the initial parameter package to complete the interval propagation update of the initial state of charge, thereby obtaining the state of charge interval.

6. The power control method based on the energy storage characteristics of lithium batteries as described in claim 5, characterized in that: The steps involve comparing the preset target bus voltage with the measured bus voltage, and then using proportional-integral scaling to cut off the available power range to obtain the final power control command. Set a target value for the bus voltage, calculate the difference between the target value and the measured bus voltage at the same time point, and use the difference between the target value and the measured bus voltage as the bus voltage comparison data. The power command is obtained by weighted summation of the bus voltage comparison data using the proportional-integral method. The power command is limited by its ramp rate, compared with the available power range, and the power command with the ramp rate limited is pruned to obtain the final power control command.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the power control method based on the energy storage characteristics of lithium batteries as described in any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the power control method based on the energy storage characteristics of lithium batteries as described in any one of claims 1 to 6.

Images